United States
         Environmental Protection
            Region 5
            77 West Jackson Blvd.
            Chicago, Illinois 60604
EPA 905/R-92/003
March, 1992
1991 Midwest Pollution Control
Biologists Meeting

Environmental Indicators:
Measurement and
Assessment Endpoints
Lincolnwood, Illinois
March 19-22,1991

                                                               EPA 905/R-92/003
von !HM YfMSfi QF *it*ig 1991 MIDfVEST


Held in


MARCH 19-22, 1991

Edited by:

Thomas P. Simon
U.S. Environmental Protection Agency, Region V
Water Quality Branch
77 West Jackson, WQS-16J
Chicago, IL  60604

Wayne S. Davis
U.S. Environmental Protection Agency
Monitoring and Quality Assurance Branch
77 West Jackson, SQ-14J
Chicago, IL  60604

Sponsored by:

U.S. Environmental Protection Agency, Region V
Instream Biocriteria and Ecological Assessments  Committee
Chicago, IL  60604
                                                  U.S. r-~P' '"\*^>.?.-';?l rro*


This document and it's contents do not necessarily reflect the position or
opinions of the U.S. Environmental Protection Agency.   This document is
intended to facilitate information exchange between professional pollution
control biologists in the midwest and nationally.  Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.

When citing individual papers within this document:

Kohlhepp, G.W. and R.A. Hellenthal. 1992. The effects of sediment deposition
on insect populations and production in a northern Indiana stream, pp. 73-84.
In T.P. Simon and W.S. Davis (editors). Proceedings of the 1991 Midwest
Pollution Control Biologists Meeting: Environmental Indicators: Measurement
and Assessment Endpoints. U.S. Environmental Protection Agency, Region V,
Environmental Sciences Division, Chicago, IL.  EPA 905/R-92/003.

When citing this document:

T.P. Simon and W.S. Davis (editors). 1992. Proceedings of the 1991 Midwest
Pollution Control Biologists Meeting: Environmental Indicators: Measurement
and Assessment Endpoints.  U.S. Environmental Protection Agency, Region V,
Environmental Sciences Division, Chicago, IL.  EPA 905/R-92/003.
Cover: Cover design and illustration by Elaine D. Synder of EA Engineering,
Science and Technology, Inc. Depicted is a fathead minnow, a bluegill, a
gammarid amphipod, and an emphemerellid mayfly superimposed on a drop of
water.  This design was developed for USEPA's Rapid Bioassessment Program,
Assessment and Watershed Protection Division, Office of Water, Washington,


After several years of hosting the Biocriteria and Ecological Assessment
Conferences in Region V, it has become a regular event our Agency can be proud
of.  The theme of this years meeting "Environmental Indicators:  Measurement
and Assessment Endpoints" is another example that Region V pollution control
biologists are in the forefront of their field bringing together innovative
approaches for addressing national issues relating to water quality.  In
addition to the keynote paper by Dr.  James Karr, a total of 22 papers were
presented at this years meeting.  Of significance was the two workshops
including the Oligoechaete taxonomy Workshop (Dr. Donald Klemm)  and Sediment
Toxicity Testing Workshop (Marsha Nelson and Jim Coyle).  The first meeting of
the Regional-State Biocriteria Workgroup (chaired by Thomas Simon and Chris
Yoder) met and discussed issues pertaining to lake and large river biocriteria

The editors of this years meeting wish to thank the Region V, Instream
Biocriteria and Ecological Assessment Committee and Region V management for
supporting this meeting.

                               TABLE OF CONTENTS
Karr and Kerans
Burton, Mullen, and
Bggert, Burton, and
Klemra and Hiltunen

Lewis and Smith

Components of Biological Integrity: their
Def itition and Use in Development of an
Invertebrate IBI                                    i

Effects of Extremely Low Frequency (ELF)
Electromagnetic Fields on the Diatom Community
of the Ford River, Michigan                        17

A Comaprison of RIA and BAd Analysis for
Detecting Pollution Effects on Stream Benthic
Algal Communities                                  26

The Freshwater Annelida (Polychaeta,  Naidid, and
Tubificid Olichaeta, and Hirundinea)  of the Great
Takes Region—an Overview                          35

A Comparison of Macroinvertebrates collected from
Bottom Sediments in three Lake Erie Estuaries      49
A Comparison of the results of a Volunteer Stream
Quality Monitoring Program and the Ohio EPA's
Biological Indices
Kohlhepp and Hellenthal The Effects of Sediment Deposition on Insect
                        Populations and Production in a Northern Indiana
Gammon and Gammon
Charles and Searle
Hoyt and Adbul-Rahim
Agricultural Impacts on the Fishes of the Eel
River, Indiana

Selenastrum Algal Growth Test: Culturing and
Test Protocol at the Illinois EPA

Effects of Acute Sublethal Levels of pH on the
Feeding Behavior of Juvenile Fathead Minnows

             Components of Biological Integrity: Their Definition and
                     Use In Development of An Invertebrate IBI

James R. Karr1 and Billie L. Kerans1
Department of Biology
Virginia Polytechnic Institute
 and State University
Blacksburg,  VA 24061-0406

Protection of quality water resources is critical to the maintenance of our way of life. Recent threats,
such as the drought in California, fish consumption advisories, and contamination of beaches, are
illustrative of the extent of abuse of water resources. These widespread declines in the quality of
water  resources  have  altered  societal  perceptions of and goafs for the management  of  those
resources. Growing interest in biological assessment in the last decade is in sharp contrast to the
status quo of earlier decades. In this paper, we briefly review the evolution of water law and outline
the conceptual foundations of ambient  biological monitoring.  We illustrate the use  of  those
foundations as we outline our efforts to develop a methodology for use of invertebrates in assessing
biological integrity.
Water Law
Early in this century,  streams and  lakes were
viewed as sources of water or as locations for
the discharge  of  societal wastes.  Eventually,
concern was expressed for the role of water
pollution   in  the  spread  of  human  health
problems; microbial contamination and oxygen-
demanding wastes were early concerns and the
threat of toxic contamination continues to

As  problems with and  perceptions of water
resources  have  changed,  water  law  has
changed as well. The first major water legisla-
tion in the United States was passed in  1889 to
control oil  pollution  and protect  navigation.
Throughout the early part of this century and
into the 1940's, legislative actions dealing with
water resources tended to be relatively weak
and provided little or no money. By the 1950's
and  1960's legislation  was  tougher,  but  it
concentrated on the development of construc-
tion grant programs to treat domestic effluent.

Passage of the Water Quality Act Amendments
of 1972  (Public Law 92-500) brought a new
approach with incorporation of the stated goal
of protecting the f ishable and swimmable status
of these resources through control of point and
non-point  sources of pollution.   Timetables,
deadlines by which society had to respond  to
and protect water resources, were developed.
For the first time, the phrase "biotic integrity"
came  into  clean water legislation with the
charge  "to  restore and  maintain physical,
chemical, and biological integrity of the nation's
waters."  But  even  that  innovation did not
stimulate  a  very  broad  perspective.  The
dominant approach  continued to be control  of
chemical contaminants.  Over the past decade
the  failure  of  that  approach  has  become
obvious. By the 1980's, new phrases such  as
"anti-degradation"   and   "ambient   biotic
assessment"  were  added  to  the  lexicon  of
water resource professionals. Calls for adoption
of biological criteria became common in the late
1980's(Karr 1991).

Overall,  early   legislative trends can   be
characterized by several common themes:  an
inordinate concentration on chemical contam-
ination; funding  for technology development
and  construction  of wastewater treatment
       Current address is Institute for Environmental Studies, Engineering Annex FM-12,
       University of Washington, Seattle, WA 98195

Karr and Kerans
facilities; and increased enforcement  efforts
directed  towards  control   of  point-source
pollution. Non-point sources of pollution were
not  seriously  considered  in  water  quality
legislation until  1972.  Even  then,  efforts to
control non-point sources were relatively weak,
primarily because they  utilized point-source
approaches and conceptual frameworks that are
inappropriate for treatment of non-point source
contamination (Karr 1990s).

Pollution Defined
The  narrow   contaminant  perspective
misconstrues the intent of the Clean Water Act.
Pollution is defined in the 1987 Act as human
"alteration of the chemical, physical, biological,
or  radiological   integrity  of  water." That
definition clearly goes  beyond  treatment of
chemical contamination to a broader conception
of factors responsible for degradation of water
resources. An integral component of any effort
to use that broad conception of pollution is the
evaluation of the resource's ability to sustain a
balanced  biological  community. If a  water
resource is degraded  to the point that it does
not support a healthy biological community, it
very  likely  will not support one  or more
beneficial uses.

Assessing Water Quality
Toxicity testing  and  chemical evaluations of
water samples have long been the mainstay of
water  resource evaluation.  Each provides
valuable information  but, when conducted in
the absence of ambient biological  monitoring,
they  do not provide  sufficient information to
protect water  resources. In  a  recent  study,
water chemistry data failed to detect 50% of
the impairment in Ohio surface waters that was
detected   with  integrated   biological   and
chemical monitoring (Rankin et al. 1990). Thus,
increased use of ambient biological monitoring
is  essential  for the  protection  of  water
resources. Why has use of  direct biological
evaluations been so limited?

First, early efforts to maintain the quality of
water resources were narrowly conceived and
planned (Karr 1991). Water pollution  control
engineers   dominated  the  agenda because
chemical contamination  was viewed  as the
problem. Water resource leadership was not
familiar with,  nor  did   it  understand,  the
ecological  dynamics that  are  important  in
influencing the effects of toxic compounds  or
other chemical contaminants. Further, they did
not  appreciate  that  degradation  of  water
resources may be caused by factors other than
chemical   contamination.   Ecologists   and
biologists must share the blame for inadequate
incorporation of biological insights  into water
resource management. Most ecologists  and
biologists  were either unable or unwilling  to
translate the  foundations of their ecological
knowledge  into   useable methodologies  to
evaluate the quality of water resources. The
lack of a defensible conceptual definition  of
biological integrity also limited progress  and use
of biological monitoring.  The phrase biological
integrity was used in the  Clean Water  Act (PL
92-500),  but development  of  a conceptual
foundation was not vigorously  pursued. The
lack of standardized field methods to sample
the biological community, to analyze the results
of  sampling,  and the tack of procedures  to
synthesize that information into assessments of
the conditions of a water resource also limited
the utility and, thus, use of biological monitor-
ing. Finally, misconceptions about the costs of
biological monitoring perpetuated the idea that
biological monitoring was too expensive.

The development of a broader  perspective to
ambient biological  monitoring  is  critical  to
protection of water resources. Two concepts
are  important  in  the development  of this
broader perspective:

    biological  integrity -  "the  capability  of
    supporting  and  maintaining a  balanced,
    integrated,   adaptive  community   of
    organisms having a species composition and
    functional organization comparable to that of
    natural habitat in the region"  (Frey 1975,
    Karr and Dudley 1981).

                                                          Components of Biological Integrity
   ecological  health  -  "...   a  biological
   system...can be considered healthy when its
   inherent potential is realized, its condition is
   stable, its  capacity  for self  repair  when
   perturbed is preserved, and minimal external
   support for management is needed " (Karr et
   al. 1986).

Evidence that biological integrity and ecological
health are seriously threatened is widespread.
Forty percent of the molluscs of the Ohio River
drainage were listed as rare,  endangered, or
extinct by 1970 (Stansbery 1970). Two-thirds
of the fishes of the Illinois River and 43% of
Maumee River fish species have  declined in
abundance substantially or have disappeared in
the last  century  (Karr et al.  1985).  For  the
fishes of California, sixty-four per cent are  in a
list that ranges from  extinct (6%) to declining
populations (22% - Moyle and Williams 1989);
only 36% of species have stable populations. In
North America, 364 fishes warrant protection
because  of their rarity  (Williams  et al.  1989).
Despite massive expenditures to improve water
quality,  none of 251 fishes listed as  rare in
1979 could be removed from the list in 1989
because   of  successful  recovery   efforts
(Williams et al. 1989).

As  these  examples   demonstrate,   water
resources are  still not adequately protected.
Inadequately treated problems include  toxics,
non-point sources, habitat  destruction,  and
altered stream flows. The limited use of biologi-
cal factors in  evaluating the  quality of water
resources  perpetuates  these problems  and
results in continuing declines in the health, the
biological integrity, of water resource systems.

The Constituents of Biological Integrity
Weaknesses  of  most  past   approaches  of
biological  monitoring  include  1)  a   narrow
conception  of the  factors   responsible  for
degradation and 2) a limited perspective on the
components of biological integrity. Over  the
past  decade,   my colleagues  and  I have
identified five primary classes of  variables that
humans impact that result in the degradation of
water resources:

   1. Water quality  -  temperature,  turbidity,
     dissolved oxygen, organic and inorganic
     chemicals,  heavy   metals,   toxic
     substances, etc.

   2. Habitat structure - substrate type, water
     depth and  current velocity, spatial  and
     temporal complexity of physical habitat.

   3. Flow regime - water volume, temporal
     distribution of flows.

   4. Energy  source  -  type,  amount,  and
     particle size of organic material entering
     stream,  seasonal   pattern  of  energy

   5. Biotic   interactions   -  competition,
     predation,  disease, parasitism.

Karr et al. (1986, see also Karr  1991) provide
a more  detailed  analysis of these factors and
how human actions impact the quality of water
resources.  Among the five, water quality has
been the primary subject of efforts by USEPA
and equivalent state agencies. The U. S.  Fish
and Wildlife Service and state fish and game
agencies   have  treated   physical   habitat
degradation.   In  recent  years,  those same
agencies evaluated altered flow regimes with
the instream-flow methodology. Few have dealt
with alteration of energy  sources that drive
stream  biology,  and  most impacts on biotic
interactions  have   come   from  efforts  to
introduce exotics and/or through harvesting of
top predators. Overall,  the  determinants of
water   resource  quality  from  a  biological
perspective are complex, and the simplistic EPA
approach  of   making   water  cleaner   is
inadequate.  We  must  evaluate  all  water
resource degradation  to  identify the factors
responsible for degradation and  then treat the
problem in the most cost-effective and efficient
manner. Ambient biological monitoring offers
unique  opportunities  to  detect,  analyze, and
plan treatment of degraded resources.

Karr and Kerans
Table 1. Components of biological integrity.
      Genes within Populations
      Populations within Species
      Species within Communities/Ecosystems
      C/E within Landscapes
      Landscapes within Biosphere

      Nutrient Cycling
      Water cycling
The components of biological integrity are also
narrowly conceived by most individuals and
agencies  charged  with  protecting  water
resources (Karr  1990, in press). Two major
aspects of biological systems - elements and
processes - must be protected (Table 1). The
most commonly cited aspects on the elements
side are the species of plants and animals in
aquatic  communities.   Additional  critical
components of the elements include the genetic
diversity  within   those  species   and   the
assemblages (communities,  ecosystems, and
landscapes) upon which those species depend.
At  the  level  of  processes,  a  myriad  of
interactions ranging from energy  flow and
nutrient dynamics to evolution and  speciation
are  critical to  the  maintenance   of  biotic
integrity. Given sufficient technology, we could
maintain species in zoos and genetic diversity in
gene banks but  in the  absence of complex
species assemblages and the processes that
keep them in existence, we  are not protecting
biotic  integrity.   The  advantages  of  this
approach  to protecting the  quality  of  water
resources are diverse (Table  2).
Table 2.  Advantages of ambient biological

   1.  Broadly based ecologically
   2.  Provides   biologically   meaningful
   3.  Flexible for special needs
   4.  Sensitive to a broad range of degradation
   5.  Integrates cumulative impacts from point
      source, non-point source, flow alteration,
      and other  diverse  impacts of  human
   6.  Integrates and evaluates the full range of
      classes of impacts (water quality, habitat
      structure, etc.) on biotic systems
   7.  Direct evaluation of resource condition
   8.  Easy to relate to general public
   9.  Overcomes   many   weaknesses   of
      individual   parameter   by   parameter
  10.  Can  assess  incremental  degrees  and
      types  of degradation, not just above or
      below some threshold
  11.  Can be used to assess resource trends in
      space or time
Assessing Biotic Integrity
Critical   components  of  a  comprehensive
approach to the protection of biotic  integrity
include evaluation of ecological attributes from
the individual to the assemblage level. Further,
an evaluation must be made with respect to the
expectation for a relatively undisturbed natural
habitat  for that region,  "regional  reference
site(s)."  Within this framework,  an efficient,
accurate assessment of the status of the water
resource is possible using biological monitoring.
Further,  an  assessment  is  likely to detect
degradation,   regardless  of   the  factor
responsible for that  degradation.  Biological
monitoring is at a threshold in the ways that it
can  be  used  and  in   the  potential  for
development of methodologies and indexes that
can provide useful answers to water resource
problems.   One   of  the  most   important
contributions of the recent growth in interest in

                                                          Components of Biological Integrity
biological monitoring has been recognition of
the need to set standards as a function of local
and regional expectations. Indeed, that should
have  been done for chemical and  physical
criteria as well. For example, total phosphorus
standards should vary regionally and according
to primary  use among Minnesota  lakes with
values ranging from less than 15  to  90 ug/l
(Hieskary et al. 1987).

Many examples of  use of ambient biological
monitoring  have been documented  in the past
decade  IKarr  et al. 1986, Ohio  EPA 1988,
Steedman 1988, Simon et al.  1988, Davis and
Simon 1989, Davis  1990). Ohio EPA has been
the most innovative  and comprehensive in their
development  and use of biological  monitoring
(Ohio EPA  1988, 1990, Rankin et al. 1990,
Ohio EPA 1991, Thoma 1991) but many other
states are rapidly  developing sophisticated
approaches as well  (e.g. Michigan,  Wisconsin,
Nebraska, Illinois, etc. - these proceedings). For
example, in the Scioto River near Columbus,
Ohio, a complex of water resource problems
representative of many areas in the U. S. can
be seen. Monitoring of the biota of the river
over the last decade  has  shown  substantial
improvement  in   biological  integrity   in
association with improvements in wastewater
treatment plants (Fig. 1). However, because of
the widespread degradation due to untreated
factors (habitat degradation, non-point source
pollution,  input   from  combined   sewer
overflow), the biotic communities of the Scioto
River adjacent to Columbus remain well below
what might be expected in that region.

Successful efforts to protect water resources
using biological monitoring have incorporated
the  following characteristics  of  biological
systems:  1)  their dynamics  at a variety  of
relevant  spatial  and temporal scales and  2)
appropriate   metrics   at   three   levels:   a)
ecosystem    (productivity,   decomposition,
nutrient  cycling,   atmosphere/biosphere/
geosphere    interactions);   b)   population/
community   (community  structure,   species
richness,   species   interactions,  functional
groupings, population structure); and c) health
of individual organisms.

We must be innovative  in incorporating these
into water resource evaluations. Some can be
incorporated directly and easily (e.g., population
size, species richness) while others  are more
difficult or expensive to  measure directly (Karr
1991). For example, the  total productivity of an
ecosystem  is  very difficult to measure.  We
should seek ways to measure productivity,  or a
surrogate of productivity  that  is indirect  but
reliable. Alternatively, we  might develop more
cost effective ways to measure  productivity by
improvements in technology.

Since early in this century, beginning with the
work  of Kolkwitz and  Marsson (1907)  and
Forbes and Richardson (1928), Forbes (1919)
and continuing  to the present, biologists have
noted   a   number  of   biological   patterns
associated  with  increased  human  influence
within a watershed: the  number of species
declines, a  small  group of intolerant species
disappear quickly, trophic specialists  decline
while trophic   generalists increase.  Effective
biological   monitoring   can structure  these
general observations to define hypotheses (e.g.
Table  3 for hypotheses  implicit  in  IBI)  and
predictions about expected pattern in aquatic
biota  under varying levels of human  influence.
If after evaluating each hypothesis, we  find
general  broad   correlations,   relationships
between   human  disturbances  and  these
attributes  of   the  community,  these  then
become assumptions. That is, we assume that,
on  average,  these  relationships  accurately
reflect the  influence of human  activities on
natural  communities. Thus, we have a fairly
robust   inference   about  the   extent   of
degradation in biological integrity at a site.

Indexes of  biotic  integrity, such  as IBI,  are,
thus,  a quantitative expression  of a number of
known  relationships  between  human distur-
bance and the characteristics of the resident
biota.  These   indexes  have  four  important
properties.  First, the accumulated information

Karr and Kerans
                             SCIOTO  RIVER,  OHIO
                               130       12O        HO
                                      River  Mile
Figure 1. Longitudinal trend in IBI for the Scioto River, Ohio in and downstream from Columbus Ohio,
1979 and 1987. CSO = Combined sewer overflow; WWTP = Wastewater treatment plant inflow;
WWH =  Warmwater habitat; EWH = Excellent warmwater habitat. Stream flow is from left to right.
(From Yoder 1989).
Table  3.    Hypotheses/assumptions  about
biological patterns associated with increasing
human effects on stream biota (modified from
Fauschetal., 1990).

   1. Number of native species  and  those of
      specific taxa on habitat guilds declines
   2. Number of intolerant species declines
   3. Proportion  of   individuals  that  are
      members of tolerant species increases
   4. Proportion of trophic specialists such as
      insectivores or top carnivores declines
   5. Proportion   of  trophic   generalists,
      especially omnivores, increases
   6. Fish abundance generally declines
   7. Proportion of individuals in reproductive
      guilds requiring silt-free coarse spawning
      substrate declines
   8. Incidence of hybrids increases
   9. Incidence of externally evident disease,
      parasites, and morphological anomalies
  10. Proportion   of   individuals   that   are
      members of introduced species increases
provides greater resolving power for the overall
index than for  each metric. The many  com-
ponents of biotic integrity (elements and pro-
cesses) and the complexity of ecological sys-
tems, limits the  likelihood that any single metric
can be used to  assess all forms of degradation
and be sensitive across the full range of degra-
dation. The magnitude of variation involved in
assessments using only a single metric derives
from natural variation and sampling error. As a
result, no single metric is absolutely reliable in
its ability to predict (with narrow precision) the
state of biological  integrity. A suite  of metrics
is better to insure more or less independent
evaluations of site quality. Although  site status
is only generally known based on each indivi-
dual metric (Fig. 2 upper), the addition of other
metrics improves the resolving power of the
approach; strong inferences can be made when
multiple metrics are used (Fig. 2 lower). That is,
each metric  has  a  level  of precision below
100% (perhaps 70-80%), but combining many
metrics with that level of accuracy and across
a variety of attributes of the biota, narrows the
range and  improves  the  precision of  the
estimate of biological integrity at a site.

                                                          Components of Biological Integrity


Figure   2.     Conceptual  depiction  of  the
relationship   between  a  single   ecological
attribute (IBI  metric-upper panel)  and two
ecological attributes  (lower  panel)and  biotic
integrity. Note that simultaneous use of two
metrics narrows the identified biotic integrity
level (dark horizontal bar) relative to use of a
single attribute (metric) (from Karr in press).
Second, the sensitivity of each metric varies
with position along a gradient from undisturbed
sites (high biotic integrity) to disturbed sites
(low biotic  integrity).  Metrics such  as total
number of species seem to decline monoton-
ically across the full range of degradation (Fig.
3).  Intolerant species disappear  before  degra-
dation has proceeded very far while the number
of anomalies changes little  until  the area  is
severely degraded.  In  contrast,  proportion  of
carnivores  declines slowly  with mild human
impacts  and declines  rapidly at intermediate
stages (Fig. 3). Carnivores disappear from a
range of heavily degraded sites.  Redundancies
exist among metrics and relative  sensitivities
vary across the range of biotic integrity.

Third,  biological monitoring  as used  in  IBI
acknowledges   and  accounts   for   natural
geographic variation. Historically, that has not
been done with physical/chemical parameters
despite the reality of natural  variation in those
attributes.  Accurate assessment of biological
integrity requires fine tuning as  one  moves
regionally. In fact, that is a major problem with
historical   chemical   monitoring   where
expectations were not adjusted regionally. For
many chemical attributes, failure to account for
regional natural variation in contaminant levels
is a serious error.

Fourth,  evaluations   attempted  using  the
multimetric  approach yield either  narrative  or
numerical results (or both) to satisfy regulatory

Developing an Invertebrate IBI
The use of  invertebrates to assess specific
anthropogenic impacts on stream biota has a
long history (e.g.,  Chutter  1972, Hilsenhoff
1977,  Winner  et al.  1980,  Rosenberg et  al.
1986). However, comprehensive  attempts  to
evaluate  stream   biotic   integrity  using
invertebrates have been attempted only recently
(e.g., Hilsenhoff 1982, 1987, 1988, Ohio EPA
1988, Lenat 1988, Lang et al. 1989, Plafkin et
al. 1989). Early in the  1980's Ohio EPA began
to  adopt the  concepts  involved in IBI  for
evaluations  using   benthic  invertebrate
communities.  They  developed  a ten  metric

Karr and Kerans
index (Invertebrate Community Index, ICI) that
parallels the original IBI (Ohio EPA 1988). We
applaud these approaches but feel that none
combines metrics that evaluate both elements
and  processes  of  biotic  integrity.  Further,
multimetric   indexes  have  not   involved
evaluation of the robustness of the individual
metrics. Thus, we outline our ongoing effort to
develop a comprehensive invertebrate index for
streams of the Tennessee Valley. We discuss 1)
formulation  of  invertebrate  metrics  and  2)
evaluation of the ability of individual metrics to
determine biological integrity.

Our first task was to develop, § priori, metrics
that  characterize  important  elements   and
processes occurring in streams. We also intend
to represent  the full biological hierarchy  from
individual to  community levels. Taxa richness
(e.g.,  number   of   Plecoptera   taxa)   and
community  composition  (e.g., proportion of
tolerant organisms) metrics are the most widely
used and highly developed  metrics in existing
invertebrate  indexes  (e.g.,  Ohio EPA  1988,
Plafkin et  al.   1989).  Metrics  designed  to
measure ecological  processes and community
function  (e.g.,  proportion   of  grazers  as  a
"surrogate" measure of periphyton production)
have   not  been  widely   used  nor   fully
investigated. Our goal is to investigate  both
types  of   metrics   using   stream   benthic
invertebrate   databases   provided  by   the
Tennessee Valley Authority.

To develop metrics associated with ecological
processes and community function we placed
organisms (usually genera) into biotic categories
describing  trophic  status,  functional group
classification, feeding mechanism, and  habit
(Table 4). Inclusion of trophic category metrics
is usually  not  done, because most benthic
biologists   prefer  categorization  of  stream
organisms   by  functional-feeding  group
(Cummins 1973, Merritt and Cummins 1984).
Our approach allows us to investigate how the
dominance of grazer-scraper or omnivore guilds,
for example, changes across sites. Inclusion of
the habit biotic category allows us to examine
patterns of loss  of taxa in particular habitats.
For example, sprawlers are usually associated
with mineral substrate, while climbers are often
associated   with  submerged  or  emergent

Using the philosophy of the IBI, we developed
28  metrics  in three distinct categories; taxa
richness and community  composition,  trophic
and functional group  composition, and abun-
dance (Table 5).  Taxa richness and community
composition metrics are often used in biological
monitoring   (Ohio  EPA 1988,  Plafkin  et  al.
1989). These include metrics  like  total taxa
richness (Metric 1, Table  5), richness of  in-
tolerant insect orders (5, 6, 8), and the percent
contribution of individuals in tolerant  groups
(15, 16) to the total community. As in IBI, taxa
richness  of  intolerant groups  often reflects
levels of degradation (e.g., Lenat 1988).

Several taxa richness and community composi-
tion metrics relate to  molluscs, an especially
rich and sensitive group  in the Tennessee
Valley. Three involve native snail and long-lived
mussel taxa (2, 3,  4). The mussel fauna of the
Valley was thought to be the most diverse in
the country and  is   declining  (Isom   1969,
Ahlstedt 1983, Starnes and Bogan 1988). Con-
sequently, mussel taxa richness should  reflect
levels of biotic integrity. We also included two
metrics  that measure   the  proportion   of
Corbicula fluminea in the community (13, 14).
The exotic  Corbicula  invaded the Tennessee
River and its tributaries. Although there is some
question as to the tolerance level  of Corbicula.
it certainly  appears to be an "opportunistic"
species (Prezant and  Chalermwat 1984), and
we hypothesize that it might be able to invade
communities where other groups, especially
native mussels, have declined.

Some  metrics  in the   taxa richness  and
community  composition  category involve the
number of  taxa in specific  habit categories.
Skaters,  planktonic  organisms,  divers,  and
swimmers  spend  most  of their time  in the
water-column (9).  Sediment-surface taxa (10)

                                                        Components of Biological Integrity

                   \  ANOMALIES
                                            ^INTOLERANT  SPP
                               BIOTIC  INTEGRITY
Figure 3.  Conceptual depiction of the range of sensitivity of four IBI metrics across the gradient from
low to high biotic integrity.
include clingers and sprawlers, whose lifestyles
place  them  primarily on  benthic  substrates.
Climbing taxa (11) spend much of their time on
submerged or emergent vegetation or debris,
while burrowers (12) live within the substrate.
We hypothesize that declining numbers of taxa
in  the sediment-surface,  water-column, and
climbing guilds should  reflect degradation  of
specific habitat types,  while increasing taxa
richness in the burrowing guild should indicate
degradation.   For  instance,  declining  taxa
richness in the group comprising the sediment-
surface taxa should reflect degradation of the
mineral substrate perhaps due to sedimentation.

Our second set of metrics includes two broad
groups, trophic and functional-feeding group
categories   (Table   5).   Although  trophic
categories   are   rarely  used   in   benthic
invertebrate studies,  we include  both trophic
and functional group categories to increase the
possibility of detecting change in  the resource
base of the community. Using  trophic status,
we hypothesize that we can determine how the
detritus food  base  of  the  community, for
example, changes by monitoring  organisms in
the detritus-feeding guild  (18). We also can
determine how collector-filterers (23) or grazer-
scrapers (24;  and their  underlying resource
base) change across sites. Finally, we included
a metric, percent of individuals in the sample
that are strictly predatory  (26, consume only
other animals in final developmental stages), to
monitor the top trophic (and functional-feeding)
levels in the community.

The  third  group, the  abundance metrics,
includes the total numbers of individuals (27)

Karr and Kerans
Table 4. Biotic categories used in classification
of invertebrates.

   A. Herbivore
   B. Carnivore
   C. Detritivore
   D. Scavenger (Detritivore,Herbivore)
   E. Omnivore {Detritivore, Herbivore,
   A. Shredder
   B. Collector
   C. Grazer
   D. Parasite
   E. Predator
   A. Chewers
   B. Filterers
   C. Gatherers
   D. Scrapers
   E. Engulfers
   F. Piercers
   A. Skaters
   B. Planktonic
   C. Divers
   D. Swimmers
   E. Clingers
   F. Sprawlers
   G. Climbers
   H. Burrowers
   I. Attachers

 * From Merritt and Cummins  1984

and the extent to which a single taxon or a few
taxa dominate  the  community  (28).  These
metrics have been used in other explorations of
community   assessment;   however,  their
properties as individual  metrics  have been
inadequately  investigated.   We  explored  a
number of cutoff points (1-5 species)  for the
dominance metric, and at present are using the
percent of individuals in the two most abundant
Our second objective is to determine how well
each individual metric distinguishes biological
integrity. To begin this process we examined
individual  metrics  to  determine if they  vary
predictably across rivers and streams monitored
by the Tennessee Valley Authority. We  use
data from the Fixed Water Quality Monitoring
Sites, an assessment program begun in 1986.
Currently, there are sites  on  12 tributaries of
the  Tennessee   River;  however,  initially
invertebrate data were collected for only six
sites - Clinch,  Powell,  Sequatchie, Elk,  and
Duck Rivers and Bear Creek. TVA used the fish
IBI  at  four of  these  sites  (Clinch,  Powell,
Sequatchie, Bear) and we have used IBI scores
to make preliminary rankings (Clinch = Powell
> Sequatchie  >  Bear) (Saylor et al. 1988,
Saylor and Ahlstedt 1990). Our intuition is that
the Elk and Duck Rivers probably fall between
Sequatchie and Powell. We examined patterns
of metrics across the sites (rivers). If a metric
exhibits  no  discernable  pattern or extreme
variability across rivers (the four with "known"
impact) then it is thought not to be able to
distinguish sites.  Using  this  approach,  we
deleted 10 metrics from further consideration
(Table 5).

Although  we looked at patterns for  all our
metrics, we discuss only a few. Total number
of invertebrate taxa increases almost  linearly
from most degraded to least degraded  sites
(Fig.  4A). Surface-taxa  richness shows the
same pattern, probably indicating that the lost
taxa  occupy  hard surfaces  (Fig. 4B).   The
proportion of Corbicula (Fig. 5A) and proportion
of  collector-filterers (Fig.  5B) show  reverse
relationships, increasing from least degraded to
most  degraded  sites.   Corbicula  increases
dramatically  only  in the  two most degraded
systems,  while  collector-filterers   increase
almost linearly. These  two  metrics seem to
provide   differing  areas  of  sensitivity  to
degradation, much as percent omnivores and
percent anomalies do in the IBI (Karr, in press).

After we determine the relationships between
our metrics and the levels of degradation at the

                                                           Components of Biological Integrity
Table 5. Metrics proposed and being tested for
inclusion in an invertebrate IB).

I.  Taxa Richness and Community Composition
1.  Total taxa richness                decline
2.  Native snail and mussel taxa*      decline
3.  Unionid taxa*                     decline
4.  Intolerant snail & mussel taxa      decline
5.  Ephemeropteran taxa              decline
6.  Trichopteran taxa                 decline
7.  Dipteran taxa*                     increase
8.  Plecopteran taxa                  decline
9.  Water-column taxa*               decline
10. Sediment-surface taxa            decline
11. Climbing taxa*                    decline
12. Burrowing taxa*                   hcrease
13. % individuals as Corbicula         increase
14.% bivalves as Corbicula'           increase
15. % individuals as oligochaetes      hcrease
16. % individuals as chironomids      hcrease

   II. Trophic and Functional-Feeding Group
17. % individuals as omnivores
     and scavengers                   rcrease
18. % individuals as detritivores       hcrease
19. % individuals as herbivores*      decline??
20. % individuals as carnivores*       decline
21. % individuals as shredders        ???
22. % individuals as
    collector-gatherers               hcrease
23. % individuals as
    collector-filterers                 hcrease
24. % individuals as grazer-scrapers  decline??
25. % individuals as parasites*        hcrease
26. % individuals as strict predators     dedhe

               III. Abundance
27. Abundance                       decline
28. % individuals in the two most
    abundant taxa                    hcrease
 •Metrics that have been dropped from further
Fixed Station Sites, we will test our metrics in
one of two ways. First, water quality measure-
ments have been taken  on the Fixed Station
Sites during the period  of  the invertebrate
studies  (Parr 1991). We will combine specific
water quality data with information on land use
practices occurring in each of the drainages and
correlate these factors with our invertebrate
metrics. This will  strengthen  our  inferences
concerning  the  ability   of  the  metrics   to
distinguish degraded conditions. Second,  we
will independently evaluate individual metrics by
applying them  to   data  collected  on  other
streams with known impacts.

Concurrently, we  are exploring a number  of
properties associated with benthic sampling and
level of  taxonomic identification. For instance,
at the Fixed Station Sites replicate samples are
collected using different methodologies at each
site; typically, several Hess samples  (taken in
pools and runs, the data presented previously),
several  Surber samples  (taken in riffles), and
qualitative samples (taken across all  habitats)
are available. We are evaluating the extent to
which  any  one sampling   method  provides
reliable  evaluations of the biotic integrity. Our
early conclusions suggest that Hess samples
better differentiate sites than either Surber or
qualitative samples. We also are interested in
whether metric behaviors are consistent across
stream size (headwater streams to large rivers),
or if we have to score metrics differently as a
function of stream size.

Our goal with respect to the development of an
invertebrate community index is to investigate
and test  a number of  hypotheses regarding
individual metrics (as  discussed above), score
those metrics that seem  to successfully reflect
biological conditions, and develop an index from
that based on a number from as few as 8 to as
many as  14  useful  metrics. After putting
together an index that we feel is  relatively
reliable, we will seek data from areas of known
impact   to  determine  the   ability  of  our
invertebrate   index  to   evaluate  conditions
resulting from varying human impacts.

Karr and Kerans

                                               ^ *o •
                                                 20 •


              SM.   OucK   BK  PowM  OireO
                                                            S*a   Duck   Elk
                                                                            Powwtt  Clinch

 I-  20
        Bwr   S»a   Due*   ek   PowWl  Olnen

       • 1966  CZH 1967 BSB 1968  •• 1969


                                       U  4O
                                       a  30



 Figure 4. Mean and standard error bars for two
 invertebrate IBI metrics for six Tennessee River
 tributaries during four consecutive years (1986-
 1989). Note that the rivers are plotted from low
 quality (Bear Creek) to highest quality (Clinch
 River). A. Total Taxa Richness. B. Surface Taxa
 Richness. Both decrease with increasing human
 In closing, we want to emphasize three points.
 Historically,  water chemistry and  fish  tissue
 sampling and toxicity testing have been the
 principal approaches used in the evaluation of
                                       Figure 5. Mean and standard error bars for two
                                       invertebrate IBI metrics for six Tennessee River
                                       tributaries during four consecutive years (1986-
                                       1989). Note  that  Corbicula  (A)  increase,
                                       especially in the most degraded sites, while the
                                       collector-filterers (B) increase gradually  along
                                       the gradient toward Bear Creek.
                                       water resources. Unfortunately, a number of
                                       forms  of  degradation  imposed on aquatic
                                       resource systems by human  society are not
                                       fundamentally chemical. As long as we depend
                                       solely on chemical analyses we are not likely to
                                       detect and treat the degradation caused by

                                                           Components of Biological Integrity
 those factors. We view the array  of water
 quality or water resource sampling programs as
 a tripod (some have inappropriately compared it
 to  a three-legged  stool). With  a tripod, the
 relative   length   of   the  legs   (monitoring
 approaches)  can  be  altered  to  suit  the
 landscape of local water resource problems.

 Second, a number of major transitions during
 the past decade have stimulated rapid changes
 in  societal  approaches  to  water  resource
 protection.  Assessment  of  water resources
 today is broader than the assessment of the
 chemical quality  of the water, a development
 that  was  precipitated  by  the   widespread
 recognition  that the quality of water resources
 continues to decline. The broader societal goals
 for the protection of water resources requires
 the use of the broader range of disciplines to
 inform water resource  decisions. Furthermore,
 decisions based  on  a  narrow  disciplinary
 approach often  must  be questioned  when
 placed in the  larger disciplinary context.  As a
 result, the dogma of many disciplines is under
 more careful scrutiny.  The dogma  of the past
 cannot be accepted uncritically  if we are to
 properly protect water resources. The planning
 perspective  in protection of water resources is
 expanding in  both space and time. Planning
 should be done for periods of decades, not for
 next  year. Planning can and should be done
 over  entire  watersheds, not  over short  river
 reaches. The emergence of a landscape ecology
 view  of water and other natural resources is
 instrumental in making society aware of the
 need to plan in a wider geographic context.

 The role of biology will continue to expand. The
 most  important challenge is for biologists and
 ecologists to be more  effective at translating
the knowledge about biological systems into
the tools that can be used by society in pro-
tecting the biological and non-biological com-
ponents of those water resources (Karr 1991).

We express  our appreciation to the Tennessee
Valley Authority (TVA) and Office of Water,
 U.S.  Environmental  Protection  Agency  for
 supporting our efforts to develop an improved
 assessment   method   using   invertebrate
 communities. We especially thank D. Kenny for
 his  help  in  determining  invertebrate biotic
 categories. Many scientists at TVA have freely
 shared  their  historical  data  sets   and  their
 extensive  knowledge of  the biota  of  the
 Tennessee River and its tributaries. Finally, we
 thank Wayne Davis for inviting us to participate
 in the  1991 meeting  of the Midwest Pollution
 Control Biologists.

 Literature Cited
 Ahlstedt.  S. A. 1983. The molluscan fauna of
 the  Elk  River in Tennessee  and  Alabama.
 American Malacological  Bulletin 1: 43-50.

 Chutter, F. M. 1972.  An empirical biotic index
 of the quality of water in South African stream
 and rivers. Water Research 6:19-20.

 Cummins, K.  W.  1973. Trophic relations of
 aquatic insects. Annual Review of Entomology
 18: 183-206.

 Davis,  W. S. (ed.J. 1990. Proceedings of the
 1990  Midwest Pollution  Control   Biologists
 Meeting.  U.   S.   Environmental   Protection
 Agency, Chicago, Illinois. EPA 905/9-90-005.

 Davis,  W. S. and  T.  P. Simon, (eds.). 1989.
 Proceedings of the 1989  Midwest Pollution
 Control Biologists Meeting. U. S. Environmental
 Protection  Agency,  Chicago.  Illinois.  EPA

 Forbes, S. A. 1928. The biological survey of a
 river system - its objects, methods, and results.
 Bulletin Illinois Natural History Survey 17:277-

Forbes,  S. A. and R. E. Richardson. 1919.
Some recent  changes in  Illinois River biology.
Bulletin Illinois Natural History Survey 13:139-

Karr and Kerans
Frey, D. G. 1975. Biological integrity of water -
an historical approach. Pp. 127-140  in R. K.
Ballentine  and  L.  J.  Guarraia  (eds.).   The
Integrity  of  Water.  U.  S.  Environmental
Protection  Agency,  Office  of  Water   and
Hazardous Materials, Washington, DC.

Heiskary, S. A., C. B. Wilson, and D. P. Larsen.
1987.  Analysis  of  regional  patterns in  lake
water  quality:   using  ecoregions  for   lake
management  in Minnesota.  Lake  Reservoir
Management 3:337-344.

Hilsenhoff, W. L. 1977. Use of arthropods to
evaluate water quality of streams. Department
of Natural  Resources, Madison,  Wisconsin.
Technical Bulletin Number 100.

Hilsenhoff, W. L. 1982. Using a biotic index to
evaluate water quality in streams. Department
of Natural  Resources, Madison,  Wisconsin.
Technical Bulletin Number 132.

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

Hilsenhoff, W. L. 1988. Rapid field assessment
of organic pollution with a family-level biotic
index.   Journal  of  the  North   American
Benthological Society 7: 65-68.

Isom, B. G. 1969. The mussel resource of the
Tennessee River. Malacologia 7: 397-425.

Karr, J. R. 1990a. Bioassessment and non-point
source pollution: An overview. Pp. 4-1 to 4-18
in Second  National  Symposium  on  Water
Quality Assessment. Meeting Summary. Office
of  Water,  U.  S.  Environmental  Protection
Agency, Washington, DC.

Karr, J. R.  1990b. Biological integrity and the
goal of environmental legislation: lessons for
conservation  biology.  Conservation  Biology
Karr,  J.  R. 1991. Biotic  integrity:  A  long-
neglected   aspect   of   water   resource
management. Ecological Applications 1:66-84.

Karr,  J.  R.  In press.  Lessons from  practical
attempts to measure biotic integrity. Pp.  in S.
Woodley,  G.  Francis,  and  J.  Kay (eds.).
Ecological  integrity and the  management of
ecosystems. University of Waterloo, Waterloo,

Karr, J. R. and D. R. Dudley. 1981. Ecological
perspective   on   water   quality  goals.
Environmental Management 5:55-68.

Karr, J. R., L. A. Toth, and D. R. Dudley. 1985.
Fish communities of  midwestern  rivers:  A
history of degradation. BioScience 35:90-95.

Karr, J. R., K. D. Fausch, P. L. Angermeier, P.
R. Yant, and I. J. Schlosser.  1986. Assessment
of  Biological Integrity in  Running Water:  A
Method and its Rationale. Illinois Natural History
Survey Special Publication No. 5., Champaign,
IL. 28 pp.

Kolkwitz, R. and M. Marsson.  1908. Okologie
der  pflanzlichen  saprobien.   Berichte  der
Deutschen Botanischen Gesellschatt 26:505-

Lang,  G.,  G.  I'Eplattenier  and 0. Reymond.
1989. Water quality in  rivers  of western
Switzerland: application of  an adaptable index
based  on  benthic   invertebrates.  Aquatic
Sciences 51:224-234.

Lenat, D. R. 1984 Agriculture and stream water
quality:  a  biological  evaluation  of  erosion
control procedures. Environmental Management
8: 333-343.

Lenat, D. R. 1988. Water quality assessment of
streams  using a qualitative collection method
for benthic macroinvertebrates. Journal of the
North American Benthological Society 7:  222-

                                                         Components of Biological Integrity
 Merritt, R. W. and K. W. Cummins. 1984. An
 introduction   to  the   aquatic   insects.
 Kendall/Hunt • Publishing Company,  Dubuque,

 Miller, R. R., J. D. Williams, and J. E. Williams.
 1989. Extinctions of North American fishes
 during the past century. Fisheries (Bethesda)

 Moyle, P.  B.  and  J. E.  Williams.  1990.
 Biodiversity loss in the temperate zone: decline
 of  the  native  fish  fauna  of  California.
 Conservation Biology 4:275-284.

 Ohio  EPA.  1988. Biological  criteria for the
 protection  of aquatic life. Ohio Environmental
 Protection  Agency, Division of Water Quality
 Monitoring and Assessment,  Surface  Water
 Section, Columbus, Ohio.

 Ohio EPA.  1990. The use of biocriteria in the
 Ohio   EPA  surface  water monitoring  and
 assessment  program.   Ohio   Environmental
 Protection Agency, Division of  Water Planning
 and Assessment, Columbus, Ohio.

 Ohio EPA.  1991. The cost of  biological field
 monitoring.  Ohio  Environmental  Protection
 Agency, Division of Water Quality Planning and
 Assessment, Columbus, Ohio.

 Parr, K. P. 1991. Water quality of the TVA
 fixed-station monitoring network.  Tennessee
 Valley  Authority,  Chattanooga,  Tennessee.

 Plafkin, J. L., M. T. Barbour, K.  D. Porter, S. K.
 Gross,  and R. M.  Hughes.   1989.  Rapid
 bioassessment protocols for  use in streams and
 rivers:  benthic  macroinvertebrates and fish.
 United States Environmental Protection Agency,
Washington, D. C. EPA/440/4-89-001.

Prezant,  R. S. and  K. Chalermwat.  1984.
Flotation of the  bivalve Corbicula fluminea as a
means  of dispersal. Science  225: 1491-1493.

Rankin, E.  T.,  C. 0. Yoder, and  D.  Mishne.
 1990.  1990 Ohio Water Resource Inventory.
Ohio Environmental Protection Agency, Division
 of  Water Quality  Planning  and Assessment,
 Ecological  Assessment  Section,  Columbus,

 Rosenberg,  D. M., H. V.  Danks  and D. M.
 Lehmkuhl.  1986.  Importance  of  insects  in
 environmental   impact   assessment.
 Environmental Management 10: 773-783.

 Saylor, C. F., G. D. Hickman, and M. P. Taylor.
 1988. Application of  index of biotic integrity
 (IBI) to fixed station water quality monitoring
 sites.  Tennessee  Valley  Authority, Norris,

 Saylor,  C.  F. and  S.  A.  Ahlstedt.  1991.
 Application of index of biotic integrity (IBI) to
 fixed station  water quality  monitoring  sites.
 Tennessee Valley Authority, Norris, Tennessee.

 Starnes.  L.  B. and A. E. Bogan.  1988. The
 mussels  (Mollusca:  Bivalvia:  Unionidae)  of
 Tennessee. American Malacological Bulletin 6:

 Simon, T. P., L. L. Hoist, and L. J. Shepard
 (eds.). 1988. Proceedings of the first national
 workshop  on  biological   criteria.   U.  S.
 Environmental  Protection Agency,  Chicago,
 Illinois. EPA 905/9-89/003.

 Stansbery, D. H. 1970. 2. Eastern  freshwater
 molluscs  (I). The Mississippi and St. Lawrence
 River Systems. Malacologia  10(1):9-22.

 Steedman,  R. J.  1988.  Modification  and
 assessment  of an index  of biotic integrity to
 quantify  stream quality in  Southern Ontario.
 Canadian  Journal  of  Fisheries and  Aquatic
 Sciences 45:492-501.

Thoma, R. F. 1991. A preliminary assessment
of Ohio's Lake Erie estuarine fish communities.
 Ohio Environmental Protection Agency, Division
of Water Quality Planning  and Assessment,
Columbus, Ohio.

Williams,   J.  E.,  J.  E.  Johnson,  D.  A.
Hendrickson,   S.  Contreras-Balderas, J.  D.
Williams,   M.  Navarro-Mendoza,   D.  E.

Karr and Kerans
McAllister, and J. E. Deacon.  1989. Fishes of
North America: endangered, threatened, or of
special concern:  1989. Fisheries (Bethesda)

Winner, R. W.f M. W. Boesel and M. P. Farrell.
1980. Insect community structure as an index
of heavy-metal pollution in lotic ecosystem.
Canadian Journal  of  Fisheries and  Aquatic
Sciences 37: 647-655.

        Effects of Extremely Low Frequency (ELF) Electromagnetic Fields
              on the Diatom Community of the Ford River, Michigan

Thomas M. Burton
Departments of Zoology and Fisheries and Wildlife
Michigan State University
East Lansing, Ml  48824

Dennis M. Mullen
Department of Zoology
Michigan State University
East Lansing, Ml  48824

Susan L. Eggert
Department of Fisheries and Wildlife
Michigan State University
East Lansing, Ml  48824

The effects of 76 Hz extremely low frequency (ELF) electromagnetic radiation produced by the U.S.
Navy's ELF andtenna on riverine diatom communities have been intensively studied since 1983 at
two sites in the fourth order Ford River in northern Michigan. Data from a control site were compared
to data from an experimental site under the antenna. Background  data on the diatom community
were collected from June 1983 through April 1986; transitional data were collected on ELF effects
during  a variable power testing period from May 1986 until October, 1989; and data from a fully
operational system were  collected from October 1989 through September 1990. Diatoms were
monitored at  a site near the antenna  and  at a  control site where they received  5.5 times less
exposure to longitudinal electric fields and 330 times less exposure to magnetic  flux  than they
received  at the antenna site. Paired t-test analyses between sites showed  that chlorophyll  a and
organic matter accrual rates were different between the sites in 1990,  the first year of fully
operational ELF exposure, whereas they were not different when data from 1983 through 1990 were
compared. Before and after, control and impact analyses (BACI) and randomized intervention analyses
(RIA) suggested that the relationship between sites had changed for chlorophyll a biomass, diatom
cell density, total diatom biovolume, and species diversity and evenness after ELF testing began as
compared to data collected prior to such testing indicating a possible ELF effect. Chlorophyll a gave
the strongest evidence for  an ELF  effect  of any of the 9 community based algal parameters
examinded. Analysis of covariance (ANCOVA) with longitudinal electric field included as a covariant
suggested that this type of electric field was not responsible for the changes observed through 1989
(1990  exposure data were not available at  the time of this analysis). Ongoing analyses are being
conducted to determine if observed changes are related to the greater exposures to longitudinal fields
experienced by the algae in 1990 or to changes in magnetic flux or to weather related variables such
as  water temperature. Stepwise  regressions  indicated  that  water temperature was the most
important of the weather related variables in explaining variance in diatom community data. We also
examined ELF effects on 3 of the most common  species of diatoms present in the summer and 4 of
the most common species present  in the winter. The abundance of none of  these species changed
between the before and  after data sets providing no evidence for ELF  effects at the individual
population level.

Key Words: Diatoms, ELF electromagnetic radiation, rivers, diversity, density, species composition,
chlorophyll a.	

Burton, Mullen, and Eggert
In 1982, a study to determine the effects of
extremely low frequency (ELF) electromagnetic
radiation on diatom  (Division Bacillariophyta:
Class Bacillariophyceae)  (classification  after
Patrick and Reimer 1966) communities in the
Ford River, Michigan was initiated. The ELF
electromagnetic radiation was to be produced
by a 56 mile long antenna (one 28 mile NS leg
and two 14 mile long EW legs that crossed the
NS leg)  to  be  built  by the U.S.  Navy for
communication   with  deeply   submerged
submarines. The initial  subcontract with  (IT
Research  Institute  (funded  as  part  of their
contract with the U.S. Navy) called for three
years of before and three years of after data for
determining ELF effects at paired sites with one
located  under the site where the antenna was
to be built (experimental site) and the other far
enough  away so  that  ELF  electromagnetic
radiation  would  be  at  least  an  order  of
magnitude less than the radiation received by
organisms  under  the  antenna  (control  or
reference site). After 10 months of selecting
preliminary sites and collecting  data  on the
biota, chemistry, and physical characteristics of
the  sites,  and  on  the  ambient  ELF  fields
(collected by IITRI) at the sites to make sure
that the final sites selected were as comparable
as possible, background data collection began
at the two selected sites. Background or before
data collection continued from June, 1983 to
May, 1986; transitional data were collected
from May, 1986 to October, 1989; and fully
operational  data  have  been collected  since
October,  1989. The transitional  data were
collected during a period of intermittent antenna
testing  that  consisted  of  only  35  days of
testing for the NS leg that crossed the aquatic
sites in 1986 (75 days for all legs) at 4-6 amps;
limited exposure at 15 amps in 1987;  limited
exposure at 75 amps in 1988; intermittent
exposure at 150 amps in most of 1989 before
going operational  in  October,  1989  at 150
amps and 76 Hz. This paper includes data on
the effects of ELF fields on the benthic algal
community  through early September,  1990.
The objective of this paper is to summarize our
results on ELF effects on the diatom community
in the Ford River. To our knowledge, no other
studies of ELF effects on natural, in situ algal
communities have been conducted.

Methods and Materials
The  two sites  chosen for the study of ELF
effects were in  riffle zones of the fourth order
Ford River in  northern Dickinson County  in
Michigan's upper peninsula. The Ford River is
part  of the Lake Michigan drainage. The Ford
River catchment lies between the Escanaba and
Menominee River catchments, and the river
empties into upper Green Bay just south  of
Escanaba,  Michigan. At the two study sites,
the stream varies from 9-12 m wide during low
flow, and depth varies from less than 0.3  m
during  low flow to more than  2  m during
floods. Chemical and physical data for the river
have  been   extensively   monitored,   and
preliminary data are available from Burton et al.
(1991 a. b) and Oemke and Burton (1986). The
river can be characterized as a hardwater, high
alkalinity (about 120 - 180 mg CaC03/L), low
nutrient (less than 10 pg  soluble reactive P/L
and  less than  60 fjg  inorganic  N/L), brown
water trout stream  with a bottom that varies
from sand and silt in  the pools  to gravel  to
cobble in  the riffles. Most of  the land in the
catchment  is  in   successional  forest with
Populus  tremuloides  dominating the  upland
areas and P. balsamifera and/or Alnus ruoosa
dominating  riparian zones.  Extensive  Alnus
ruaosa and some Thuja occidentals swamps in
the drainage basin along with forest humus and
litter are likely  sources of the organic matter
that imparts the brown color to the river. The
pH in  the river varies between 7.5  and 8.2
most of the time, and dissolved oxygen remains
near saturation throughout the year.

The experimental site was located within 50 m
of the point where the ELF antenna crossed the
Ford River (T43N:R29W:Sec. 14), while  the
control or reference site was about  8  km
downstream (T43N:R28W:Sec. 21). These two
sites were closely matched  in  exposure  to
ambient ELF electromagnetic levels  prior  to

                                                          Effects of ELF on Diatom Community
operation  of  the antenna. At  full  power  in
1989, 5.5-> 10 times more ELF radiation was
received by the biota  at the experimental site
than at the control site (e.g. the experimental
site  received  5.5 times  more  exposure  to
longitudinal electric fields;  540 times  more
exposure to transverse electric fields, and 330
times more exposure to magnetic flux than did
the  control   site  at  76  Hz  according  to
measurements made by IITRI in 1989). At full
power of 150 amps, biota at the experimental
site  received ELF  exposure at  a rate of 61
mv/m2.  Placement of  samplers in  1990 was
adjusted to insure that  biota at the experimental
site received  at least 10 times more exposure
for all components  of  ELF electromagnetic
radiation measured at  76 Hz during full power
operation than did the biota at the control sites.

Standard (25X76 mm), glass microscope slides
were held  vertically  into the  current  in
plexiglass diatom racks using a modification of
the diatometer illustrated in Patrick and Reimer
(1966) except  that no  deflector shield was
placed in front of the slide holder, and the racks
were fastened directly  to standard construction
style clay bricks that  were placed directly on
the stream bottom. The slide racks were placed
in riffle areas, and current velocity through the
slide holders  was carefully matched  using a
current  meter.   Positions of the racks were
checked  and  adjusted weekly to insure that
current velocity  differed  little  between sites.
Care was also taken to place the racks in areas
of the stream that were open to sunlight during
most of the day. Some early morning  and late
afternoon shading by riparian vegetation and/or
banks of  the streams was unavoidable but
shading differences between sites was minimal.

Colonization times of  28 days for the mature
community and  14  days for  chlorophyll a
accrual rates  during the ice-free seasons were
selected based on the findings of Oemke and
Burton (1986) for this  river during start up
studies in 1982. Winter sampling occurred at
28  day   intervals  from   1983-84  through
1986-87, at 56-60 day  intervals in 1987-88,
and has remained at 42 day intervals since that
time.  Processing  of  diatoms  for  counting,
calculations, and other procedures followed the
methods discussed  by  Oemke  and  Burton
(1986)  as  did  procedures  for  determining
chlorophyll  a and organic matter accrual rates
and biomass standing  crops. We use accrual
rates as crude estimates of primary productivity
as  recommended  in  Standard   Methods
(A.P.H.A. 1985). We prefer to refer to these
measures as accrual rates, since productivity is
underestimated  or  affected by  an unknown
amount due  to  sloughing  losses,  leakage of
cellular  contents,  death and  senescence of
cells, etc.

All statistical procedures except for before and
after, control and impact (BACI) analyses and
random  intervention  analyses   (RIA)  were
performed  using StatView 512+  (copyright
1986 by Abacus Concepts, Inc.),  a software
program for  the Apple Macintosh  plus  from
Brainpower, Inc. of Calabasas, CA. Procedures
for  BACI and RIA  analyses are summarized in
Eggert et al. (1991), a companion paper in this
volume. Differences between treatments were
accepted as significant at the p>0.05  level
unless otherwise specified.

There were no significant differences for any of
the  nine  community  level  benthic  algal
parameters  monitored  at the  antenna  and
control   sites  from   1983  through  1990
according to paired t-tests analyses (Table 1).
Values at the control site and the antenna site
were significantly correlated for each of these
parameters  (Table 1).  Minimum  detectable
differences were in the 25-30 % range for most
parameters with Shannon-Wiener diversity (H')
and evenness  (J') being the most sensitive
indicators for detection of differences between
the sites (5-7 % - Table 1). Diatom cell density
and total diatom  biovolume (calculated  from
individual cell volumes and density) were the
least  sensitive  parameters for  detection of
differences  between  the  sites  (48   -53 %
minimum detectable differences - Table 1).

Burton, Mullen, and Eggert
Table 1. Paired t-test results (DF  = 83 -  85), correlation coefficients*,  and percent minimum
detectable  differences  (p  < 0.05) between  the  antenna and  control sites for benthic algal
parameters for 1983 - 1990.
Parameter                   t-test         correlation

Chlorophyll a Biomass        NS           0.85

Chlorophyll a Accrual         NS           0.83

Organic Matter Biomass      NS           0.70

Organic Matter Accrual       NS           0.61

Diatom Cell Density          NS           0.90

Diatom Cell Volume          NS           0.96

Total Diatom Biovolume      NS           0.70

Species Diversity            NS           0.70

Species Evenness            NS           0.79

* all are significant at the p < 0.01 level.
        minimum detectable

        29 %

        32 %

        23 %

        27 %

        48 %

        25 %

        53 %

        7 %

        5 %
We also compared the antennna to the control
site for each year of the study using the paired
t-tests. Few  differences were detected on a
year by year basis. The first fully operational
year for ELF exposure was 1990. Comparisons
between the antenna and control sites for 1990
using paired t-test analyses suggested that two
of the  parameters listed in Table 1, chlorophyll
a biomass and daily accrual rates of organic
matter, had been  affected by ELF exposure.
Chlorophyll   a  biomass  was  slightly   but
significantly higher under the antenna than it
was at the control site. However, chlorophyll a
daily  accrual  rates  were  not  significantly
different   between   the  sites   in  1990.
Conversely, organic  matter biomass was  not
signifcantly diffferent between the two sites in
1990 even though organic matter daily accrual
rates were. Since algal and bacterial biomass
should dominate production of organic matter
on slides oriented vertically to the current (little
settling of suspended  organic matter should
occur in this orientation), it is surprising that
results from the two parameters are not more
closely correlated. No significant differences for
any  of the  other  seven  community  based
parameters listed in Table 1 were detected in
1990. No significant differences had occurred
between sites  overall  (Table  1) or even for
1989, the year of intermittent exposure to 150
amp  operation  of  the  antenna.  Significant
results should occur by chance  alone 5 % of
the time.  With  two of  the nine community
based parameters (22 %) being significantly
different between the sites in 1990, differences
appear to  be real. We feel that it is too early to
put much emphasis on these analyses after only
a single year of fully operational data and will

                                                         Effects of ELF on Diatom Community
Table 2. Summary BACI and RIA statistics. Before statistics are from June 1983 to April 1986. After
statistics are from May 1986 to September 1990. N in parentheses for BACI and RIA respectively.
NS = p > 0.05. N  = 83 - 85 overall, 48 - 53 for summer, 33 - 35 for winter samples.
Chlorophyll a. Biomass
Summer Data Only
Winter Data Only
Organic Matter Biomass
Summer Data Only
Winter Data Only
Diatom Cell Density
Summer Data Only
Winter Data Only
Diatom Cell Volume
Summer Data Only
Winter Data Only
Total Diatom Biovolume
Summer Data Only
Winter Data Only
Species Diversity
Summer Data Only
Winter Data Only
Species Evenness
Summer Data Only
Winter Data Only
p < 0.01
p = 0.06
p < 0.01
p < 0.01
p < 0.01
p < 0.05
p < 0.05
p < 0.01
p < 0.05
p < 0.05
p < 0.05
p < 0.01
p < 0.01
p = 0.06
p = 0..18
p = 0.09
p < 0.05
p < 0.01
p < 0.05
p < 0.05
await  the  1991 data  before  drawing firm

Studies that use comparisons between a single
control  and reference site have a  potential
problem with pseudoreplication (Stewart-Oaten
et al. 1986, Carpenter et al. 1989), and there
is some question as to whether or not paired
t-tests are appropriate ways to analyze the
data. The two statistical procedures suggested
as alternatives by Stewart-Oaten et al. (1986)
and  Carpenter  et  al. (1989)  to  the more
                   traditional  paired t-test approach  have been
                   used on our data as an additional way to detect
                   potential differences between the two sites that
                   may be related to  ELF  electromagnetic fields
                   (Eggert et al. 1991). Since RIA analyses require
                   a   minimum  number  of  40 observations,
                   analyses are only possible at this time using all
                   the transitional and operational data (May 1986
                   -  September  1990). We  ultimately plan  to
                   stratify results into before data (June 1983 -
                   April 1986),  transitional  data with the ELF
                   antenna   undergoing  testing   (May  1986  -

Burton, Mullen, and Eggert
September  1989),   and  operational  data
(October 1989 -  September 1992). Both  the
RIA and BACI analyses test for differences in
the relationship of the data between the two
sites before and after a potential impact occurs.
The means do not have to be the same to show
no significant difference. Suppose, for example,
that density had consistently been higher at the
control site than at the impact site prior to the
onset of the potential environmental stress (e.g
ELF exposure in this study). No impact would
be indicated by either analysis if this difference
remained consistent after onset of the potential
stress (e.g. exposure to ELF fields). If, however,
the  relationship  changed such that  mean
density decreased at the impact site but not at
the control site, BACI and RIA analyses would
indicate a significant change even though  the
means might now be comparable. This signifi-
cant change could signal a potential significant
change in density at the impacted site related
to ELF exposure (Eggert et al. 1991).

Results of the preliminary  BACI analyses for 7
of the 9 community based parameters reported
in Table 1 indicated that significant differences
have  occurred   between  the  before   and
transitional/operational data for 5  of the 7
parameters examined using the overall  data set
(line one for each parameter in Table 2).  RIA
analyses confirmed these  differences as being
significant for 3 of the 5 parameters. Both  RIA
and  BACI analyses showed that before data
were different from transitional/operational data
for chlorophyll a biomass and species diversity
and evenness indicating a potential ELF effect
on these parameters (Table 2). BACI analyses
showed that  before  diatom  cell density  and
total  diatom biovolume data  differed signifi-
cantly from after data, but RIA analyses failed
to confirm these  results at the p > 0.05 level
(differences were significant at the p  > 0.10
level - Table 2). No differences were detected
for any of these 5 parameters using the paired
t-tests on the  combined data (Table 1).

Winter data  are  more variable  than  summer
data, are collected at less frequent intervals,
and could mask differences in the  data sets.
Therefore,  we stratified the  data  into data
collected  from  mid-April   through  October
(summer  data)   and  data   collected  from
November to April (winter) to  see if we could
improve results  (Table 2).  Of  course, smaller
sample  size  results  in less power to detect
differences. Even so, differences occurred in
the organic matter biomass summer  before and
after data, that were not evident in  the overall
data set even though significant differences in
some of other parameters  disappeared  (Table
2). Species evenness was  the only  parameter
that  showed  a  consistent difference in the
winter as well as in the summer and overall
(Table 2).

BACI analyses were used to compare year to
year differences, since  it is less sensitive to
sample size than is RIA analyses. While some of
the other parameters showed  differences in at
least one of  the before years  compared to at
least one of the transistional/operational years,
only chlorophyll a biomass showed a significant
difference between data collected in 1990, the
only fully operational year  so  far, and each of
the before ELF years (83,  84, and  85). Since
paired t-test  analyses had  also indicated  that
differences between the control  and antenna
site  were siginifcant in 1990, chlorophyll a
offers   the   best  evidence   that   ELF
electromagnetic radiation may have caused a
difference in  benthic algae between  the two
sites.   Chlorophyll   a  was  slightly  but
significantly greater at the antenna site than at
the reference site in 1990. If ELF has caused an
effect,  it has resulted  in slightly  increased
biomass of chlorophyll a at the antenna site.

Since a  variety of  differences were  detected
between the  before and transitional/operational
data, we analyzed  the data using  analysis of
covariance   (ANCOVA)  with   cumulative
exposure to ELF longitudinal fields over the 28
day colonization period as the  covariant for the
1986 through 1989 after  data (Table 3). The
1990 ELF exposure data are not  yet available.
The only difference between sites detected in

                                                          Effects of ELF on Diatom Community
Table 3. Analysis of  covariance (ANCOVA) statistics  for 1986 - 1990. Before and after t-test
comparisons are for before period from June 1983 to April 1986 and May 1986 to September 1990
respectively. NS = p > 0.05.
Chlorophyll a. Biomass
Organic Matter Biomass
Diatom Cell Density
Diatom Cell Volume
Total Diatom Biovolume
Species Diversity
Species Evenness
Between Site t-Tests
Before After ANCOVA
p < 0.05
p < 0.05
p < 0.01
p < 0.01
P <
P <

the after data using between site t-tests were
for species diversity and evenness  (Table 3).
ANCOVA  did  not  change  any  of these
relationships (Table 3) indicating that diffences
between sites was not related to exposure to
longitudinal electric fields from 1986  through
1989. Therefore, differences detected between
sites  as  summarized  in Tables  1 and 2 and
above must be related  to  some factor other
than ELF longitudinal  electric fields or to the
greater differences in longitudinal electric fields
in  1990, the first year of fully operational data.
Exposure to increased magnetic  flux could  be
the  explanation  as   could  between  year
differential responses to weather related factors
between  the  sites.   Stepwise   regression
analyses strongly implicated water temperature
as the most important weather related factor in
explaining  variance  in  the benthic   algal
community. Future analyses will explore water
temperature and ELF generated magnetic flux
exposure  as  covariants  that  may  explain
differences  in  the  before  and after  data
between sites. As soon  as  1990 data  on
longitudinal electric fields are available to us,
we will incorporate these data in our ANCOVA

Since BACI and  RIA anayses  had  indicated
differences  in  the  before  and  transitional/
operational  data  in  species  diversity  and
evenness,  we  selected three  of  the  more
common species of the diatom flora from the
summer period and four from the winter period
for further analyses (Table 4). There were no
significant differences for any of these species
between the before and transitional/operational
data (Table 4). Differences  in diatom species
diversity and  evenness must,  therefore,  be
related to differences  in rarer  species  rather
than these more common forms.

In summary, data collected to date on response
of  the   benthic  algal   community  to  ELF
electromagnetic fields suggest that changes
may be  occurring in the algal community that
may  be   related   to  exposure   to   ELF
electromagnetic fields. These changes include

Burton, Mullen, and Eggert
Table 4. Summary BACI and RIA statistics for selected diatom species. Before statistics are from
June 1983 to April 1986; N = 44-45 for summer and 31 for winter. After statistics are from May
1986 to September 1990; N = 45-46 for summer and 33 for winter. NS = p > 0.05.

Achnanthes minutissima

Cocconeis placentula

Cvmbella minuta


Achnanthes minutissima

Fragjjaria vaucheriae

Gomphonema olivaceum

Svnedra ulna














changes  in  chlorophyll  a  biomass,  organic
matter accrual rates,  diatom  density,  and
diatom species diversity  and evenness.  The
strongest  evidence for such changes  comes
from the chlorophyll a biomass data. There is
no  evidence to suggest  that any significant
changes have occurred in numbers or volume of
the  more common  algal  species  in   the
community. Potential effects on the community
cannot be completely confirmed at present due
to only a  single year of exposure to the  fully
operational ELF antenna. Studies on the benthic
algal community over the next one or two years
should clarify the preliminary results reported in
this paper.

We thank Jennifer Molloy and Mark P.  Oemke
for diatom enumeration and identification. We
also thank M. O'Malley,  D. Repert, and  R.
Stelzer for technical assistance. Support for this
             research was  provided  by  the  U.S. Navy
             through  a  subcontract  from  I IT  Research
             Institute under  contract numbers N00039-81,
             N00039-84-C-0070, and N00039-88-C0065.

             Literature Cited
             A.P.H.A. 1985.  Standard methods for the
             examination of water  and wastewater. 16th
             edition.  American Public  Health Association,
             Wash., D.C., 1268pp.

             Burton, T. M., M. P. Oemke, and J. M. Molloy.
             199la.  Contrasting  effects  of nitrogen  and
             phosphorus additions  on  epilithic  algae  in a
             hardwater and  a softwater stream in Northern
             Michigan.  Verh.  Internet.  Verein.  Limnol.

             Burton, T. M., M. P. Oemke, and J. M. Molloy.
             1991b.  The  effects  of  stream  order  and
             alkalinity  on  the  composition   of  diatom

                                                        Effects of ELF on Diatom Community
communities in two Northern Michigan river
systems. In:  P. Kociolek (ed.)  Proc.  11th
International Diatom Symposium.  Mem. Calif.
Acad. Sci. (In press).

Carpenter, S. R., T. M. Frost, D. Heisey, and T.
K.  Kratz.  1989.  Randomized   intervention
analysis   and   the  interpretation   of
whole-ecosystem   experiments.   Ecology

Eggert, S. L, T. M. Burton, and D. M. Mullen.
1991. A comparison of RIA and BACI analysis
for  detecting  pollution   effects  on  stream
benthic algal communities, pp. In:  W. S. Davis
and T.P. Simon (eds.)  Proc.  1991  Midwest
Pollution Control Biologists Meeting. U.S. EPA
Region  V,  Environmental Sciences  Division,
Chicago, IL. (this volume).

Patrick,  R. and C. W.  Reimer.  1966. The
diatoms of  the United  States  exclusive  of
Alaska and Hawaii, Volume I. Acad. Natur. Sci.
Philadelphia, Philadelphia,  688 pp.

Oemke,  M. P. and T. M. Burton. 1986. Diatom
colonization dynamics  in  a lotic  system.
Hydrobiologia 139:153-166.

Stewart-Oaten, A., W. W. Murdoch, and K. R.
Parker.   1986.  Environmental   impact
assessment:   "Pseudoreplication"   in  time?
Ecology  67:929-940.

         A Comparison of RIA and BACI Analysis for Detecting Pollution
                   Effects on Stream Benthic Algal  Communities

Susan L. Eggert
Department of Fisheries and Wildlife
Michigan State University
East Lansing, Ml 48824

Thomas M. Burton
Departments of Fisheries and Wildlife and Zoology
Michigan State University
East Lansing, Ml 48824

Dennis M. Mullen
Department of Zoology
Michigan State University
East Lansing, Ml 48824

Before and After, Control and impact analysis (BACI) and Randomized intervention Analysis (RIA) are
used to overcome pseudoreplication in sampling designs that emphasize data collection  at paired
control and impacted sites. Both methods were applied to data collected on the potential effects of
76 Hz extremely low frequency (ELF) electromagnetic radiation on benthic diatom communities at
two sites in the Ford River, Michigan. Data on the diatoms were collected as part of the U.S. Navy
funded studies (through I IT Research Institute) on ELF effects from an ELF communication antenna.
Data on chlorophyll a., diatom species diversity, and diatom abundance were collected at an impact
site under the antenna and at a control site receiving 7-10 times less ELF radiation than the impact
site. Data were divided  into a 3 year before period and a 6 year after period. Both procedures
detected differences (p < 0.05) in the relationship between the two sites in chlorophyll a and species
diversity before  antenna operation  as compared to the period after antenna operation began, but
detected no difference in diatom abundance. The RIA procedure was limited by sample sizes less
than 40, limiting our ability to detect monthly differences with it. The small sample size problem
emphasizes the need for  long term monitoring. The parametric (BACI) and  randomization (RIA)
procedures offer powerful, complimentary tools for the detection of pollution effects using control
and impact sites.

Keywords: Statistical applications,  BACI analysts, RIA analysis, diatoms, rivers, ELF electromagnetic
The  experimental  design   of   large-scale
monitoring   efforts   used   to   determine
environmental  effects  of  a  pollutant  has
commonly been  plagued  by the problem  of
psuedoreplication as defined by Hulbert (1984).
A typical sampling design, set up to determine
whether a known pollutant will adversely affect
a biological parameter, consists of replicated
sampling over time at a control and impact site.
However, funding  and  logistic  constraints
associated with environmental monitoring often
makes the  replication of  treatments  (sites)
impossible. This  lack of replication,  therefore,
invalidates the use of inferential statistics such
as analysis of variance in interpreting the data
(Stewart-Oaten 1986).

The  problem   of   statistically   detecting
non-random changes between  a single control

                                                Comparison of RIA and BACI Analysis
and impacted site has received some attention
in the literature (Stewart-Oaten et al.  1986,
Carpenter et al. 1989, Carpenter 1990, Jassby
and Powell 1990, Reckhow 1990).

Stewart-Oaten et al. (1986) have  introduced
the  Before  and  After, Control and  impact
analysis (BACI), which requires paired sampling
at control and impact  sites, both before and
after perturbation. The mean of the  "before"
differences between sites is compared  to the
mean of the "after" differences between sites
by a t-test. If the magnitude of the difference
between sites changes significantly following
addition  of  the  pollutant, there may be  a
perturbation effect.  The procedure assumes
that the following  criteria are met: (1) the
measures of the  parameters at any time are
independent of the measures at any other time,
and  (2) the differences between control and
impact sites of the "before" period are additive.

Randomized   intervention  Analysis  (RIA)
represents a non-parametric alternative  to the
BACI analysis  (Carpenter et al.  1989).  RIA  is
based on replicated sampling over time,  before
and after an impact, at control and experimental
sites. A  mean difference  between  sites  is
calculated from both "before" and "after" data
sets.  The  absolute  value  of  the  difference
between   these  means represents the  test
statistic.  Random permutations of  the time
series  of inter-site  differences provide  an
estimate of the distribution of the test statistic.
The proportion of  randomly created differences
between  means  that  are  greater than the
observed  difference  between   means,
determines whether  a  significant change has
occurred  between sites. By using a randomly
created error distribution, the RIA design does
not require transformations for non-additive
data.  RIA does require the assumption that
errors  are independent  from  one  another
through time. However, Carpenter et al. (1989)
indicated that serial  correlation did not lead to
ambiguous results   in   97  percent  of  the
autocorrelated cases that they examined.
The objective of this paper is to demonstrate
the potential use of RIA and BACI as statistical
tools for the detection of pollution  effects in
environmental monitoring. Here we compare
and  contrast  both  methods  using  diatom
abundance, species diversity and chlorophyll a
standing stock data sets collected as part of an
on-going study to monitor potential  effects of
76   Hz   extremely   low  frequency   (ELF)
electromagnetic  radiation  on stream  benthic
diatom communities  at a  control and impact
site in the Ford River, Michigan.

Methods and Materials
The data were obtained from an ongoing study
of the effects of ELF electromagnetic radiation
(generated by the U.S. Navy's ELF antenna) on
the Ford River ecosystem  in Michigan's upper
peninsula (see Burton et. al. in this volume for
more details on this study). Glass microscope
slides held in plexiglass carriers and  incubated
in the river for 28 days (42  days during  the
winters  of the last three years of the study)
were used to sample the periphyton community
at two sites in  the river. The impact site  lies
directly beneath the antenna and receives about
7 times the exposure of the control site, which
is about 8 km downstream from the antenna.
Pre-operational data were collected every 28
days  between  June  1983  and  May  1986.
Operational data were collected every 28 days
during the spring, summer and fall and every 48
days during the winter between June 1986 and
September 1990. Testing on the antenna began
in May 1986 at 4 amps (impact site exposure
rate  =  1.6 mv/min).  On April  28,  1987  the
power was increased to 15 amps (6.1  mv/min
at the impact site) and on November 15, 1987
the power was increased to 75  amps (31.0
mv/min  at the impact site). Testing at  full
power (150 amps, 61.0 mv/min at the impact
site) began on May 1, 1989 and full operations
at 150 amps began on October 7, 1989. The
antenna  was  operated  in a  fairly irregular
pattern  during the testing period from May
1986 to October 1989.

Eggert, Burton and Mullen
BACI  analysis and  RIA  were conducted on
chlorophyll a standing crop (10 replicates per
site per sampling date), species  diversity (3
replicates  per site  per sampling date),  and
individual   diatom  species   abundances   (3
replicates per site per sampling date). Each data
set was split up  into  "before"  and "after"
periods with all sampling dates from June 1983
to April 1986 as the "before"  period and all
dates from May 1986 to September 1990 as
the "after" period. The data for the biological
parameters and diatom abundances  were also
divided  into  summer  and winter seasons to
statistically  examine seasonal  variations for
these parameters.  Seasons  consisted  of a
Summer  (May  to  October)  and  a Winter
(November to April) period. Those seasons prior
to Summer  1986  represented the "before"
period, while the "after" period consisted of all
seasons  after  May 1986.  Using  the  BACI
technique, we also compared individual seasons
of the  "before"  period  to other  "before"
seasons to determine whether any differences
occurred prior to impact. Each of the "before"
seasons was then  compared to each  of the
"after"  seasons  to see  whether  significant
differences between sites had occurred. Finally,
peak  diatom  abundances  were  compared
between sites using BACI for those  months of
the year  when species such  as  Achnanthes
minutissima become dominant.

The  BACI analyses were  run as specified by
Stewart-Oaten (1986). Data points  from each
sampling  date  were entered into files using
Statview   512+   on   a  Macintosh    Plus
microcomputer. The set of "before"  data were
tested for additivity using Tukey's one  degree
of freedom test for non-additivity. If the slope
of the regression of differences between sites
against   the means  of  both  sites   varied
significantly from zero (9 < 0.05), then data
were  transformed.  The  log  (x   +   1)
transformation   was  used  for  non-additive
biological data, while the arcsin square root of
the mean'transformation suggested by Steel
and  Torrie (1960)  was used for proportional
data. According to Stewart-Oaten et al. (1986)
the independence of error assumption required
by the BACI analysis may be considered to be
"plausible if large,  local, long-lasting random
effects are unlikely". While our initial analysis of
the data indicated  that this assumption was
indeed plausible, any significant or questionable
results were closely examined for possible serial
correlation  problems. The Ourbin-Watson test
(1951) was used to test for independence of
errors. If the  "before"  data  sets  met  the
additivity and  independence of error criteria,
differences between sites for each time period
were compared with an unpaired two-tailed
t-test. Data sets  which failed  to meet  the
stringent requirements of BACI analysis were
analyzed using the non-parametric  RIA test.

Data for RIA were entered into data files using
the Supercalc software program and transferred
to separate ASCII files for each parameter for
each site.  By  using a randomly created error
distribution, RIA  eliminates  the  criterion  for
additive data. RIA calculations were performed
on  an IBM microcomputer using the RIAPUB
program   obtained  from   Dr.  Stephen  R.
Carpenter.  The program is interactive in nature
and is applicable for most studies of this type.
A probability distribution created  by RIAPUB
through random permutations of the time-series
of inter-site differences determined whether a
significant change between sites occurred after
antenna operation.

Results and Discussion
Pooled and seasonal chlorophyll § standing crop
data for the "before" period  were  found to be
non-additive using Tukey's test. A plot of mean
standing crop against differences between sites
using  raw  data  showed   the  slope  to  be
significantly greater than zero  (Figure 1a and
1c). The log (x +  1) transformed data set was
retested for  additivity  and  the slope  of  the
regression line was found to be not significantly
different from zero (Figure  1b and  1d). Results
of a BACI unpaired t-test on pooled chlorophyll
a data indicated  a significant difference (p <
0.01) between  "before"  and "after"  period
mean differences (Table 1). When the data

                                                  Comparison of RIA and BACI Analysis
             Q O
              ui 5
              ui Si





                                     p < 0.01
                                               » p > 0.05
                    '"T	•	1     '     |        -u.* T~—»—^—^^—j—•>——|
                      0        10        20          0.0    0.4    0.8     1.2
Figure 1. Non-additive  and additive chlorophyll  a. data. (A) Difference between sites  of raw
chlorophyll a data for "before" period.  (B)  Difference between sites of log(x + 1) transformed
chlorophyll a data for "before" period. (C) Tukey's test for non-additivity on raw chlorophyll a. data
(significant at p < 0.05). (D) Tukey's test for non-additivity on log transformed chlorophyll a. data
(non significant).

Eggert, Burton and Mullen
Table 1.  Summary of BACI and RIA comparisons for chlorophyll a, species diversity and diatom
abundance between control (FCD) and experimental (FEX) sites for 1983-1990. N in parentheses for
BACI and RAI, respectively.
(p < 0.05)
 (p < 0.05)
     Chlorophyll a.    6/83-4/86 vs. 5/86-9/90  (84) (84)
                                       p  <  0.01
                                         86-90  (49) (51)
  Summer 83-85 vs.
       S 83/88 (9)
       S 83/90 (9)
       S 84/87 (11)
       S 84/88 (10)
       S 84/90 (10)
       S 85/87 (11)
       S 85/88 (10)
       S 85/90 (10)
  Winter 83-85 vs.  86-89  (33) (33)
                                                             P «
                                                             P <
                                                             P <
                                                             P <
                                                             P <
                                                             P <
                                                             P <
                                                             P <
                                                             P <
 p < 0.05

 p < 0.01
6/83-4/86 vs. 5/86-9/90  (83)(85)

  Summer 83-85 vs. 86-90  (48) (51)
  Winter 83-85 vs. 86-89  (33)(34)
     Diatom Abundance
        A., minutissima
   p < 0.05

                        Summer 83-85 vs. 86-90  (45) (46)         NS
                        Winter 83-85 vs. 86-89  (31)(33)         NS

                  Hay  &  June 83-85 vs. Hay £ June 86-90  (12)   NS
  p < 0.05

were broken  down  on  a  seasonal  basis,
significant differences between the summers of
83, 84 and 85, and the summers of 87, 88 and
90 were found.  A closer inspection of the
chlorophyll a "before" data set revealed that the
independence  assumption was not completely
satisfied. Results of a Durbin-Watson test on
"before" period data indicated that a significant
(d = 1.14, p < 0.05) serial correlation problem
                         Chlorophyll a.  standing crop  data  were then
                         analyzed using the non-parametric randomized
                         intervention analysis. The inter-site relationship
                         changed over time for both the entire data set
                         and the summer seasonal data (Table 1). Due to
                         the lack of sensitivity of RIA at observation
                         numbers less than 40, year-to-year comparisons
                         could not be run (Carpenter et al. 1989).

                         Diatom species diversity data were found to be
                         additive using Tukey's test for pooled (p <

                                           Comparison of RIA and BACI Analysis

                                                         Impact ste
                                                         Reference site
                    a CD co CD 00 oo 
Eggert, Burton and Mullen
additional  statistical analysis of diversity data
using   the   non-parametric   randomized
intervention analysis was redundant. A check of
all data sets indicated that all assumptions of
BACI had  been satisfied,  thus making  RIA

The  abundance of a dominant diatom species
Achnanthes  minutissima  has  followed  a
predictable pattern of high dominance during
summer periods and low dominance during  the
winter (Figure 2).  All diatom abundance data
were transformed using the arcsin square root
of the  mean  transformation and  tested  for
additivity. Seasonal transformed data for  the
before period were marginally additive (summer:
p   <   0.06   and  winter:  p   <   0.08).
Ourbin-Watson independence tests of summer
data  found   no  significant  autocorrelation
problem (d  = 1.98), while  winter data were
significantly autocorrelated  (d  - 0.90, p  <
0.05).  Results  of  unpaired  t-tests  for  A.
minutissima  indicated that  there  were  no
significant  inter-site  changes  in  mean
differences for either seasonal  "before" and
"after" periods, or year-to-year comparisons.
Since the winter abundance data were found to
be  significantly  autocorrelated, BACI  results
were verified using  RIA.  RIA  reflected  the
results obtained in the  BACI comparisons of
both the summer  and winter abundance data
(Table 1).

In an  attempt  to detect  even more subtle
changes in diatom abundances, we ran BACI
analyses  on   monthly  data  at  peak   A.
minutissima abundances (Figure 2). All May and
June data  for  the years  1983-1985 were
pooled to represent the "before" period and all
May and June  data for 1986-1990  as  the
"after" period so that mean differences beween
sites could be examined. The data appeared to
be significantly negatively serial correlated (d =
3.33,   p  <   0.05).   Since   negative
autocorrelations are conservative with regard to
probability levels, an unpaired t-test was run on
the  monthly data. There was no significant
change  in  the  inter-site  relationship after
antenna  operation  according  to the  BACI
analysis (Table  1). This comparison could not
be verified with RIA due to the limited number
of observations available.

For the chlorophyll a. standing crop and species
diversity   parameters   where   significant
differences were  found using either BACI or
RIA,  the  data were  scrutinized further  to
determine   whether   ELF   electromagnetic
radiation or another factor had  caused  the
observed differences.  Significant differences
found by  BACI or  RIA do  not imply that a
suspected perturbation has caused a change,
nor do these tests reveal at what  point in time
the change occurred. Ecological and procedural
considerations should be examined in all cases.
Analysis   of   covariance   (ANCOVA)   of
chlorophyll a standing crop with ELF  exposure
included as a covariant  indicated that a variable
other than ELF electromagnetic radiation caused
the  change  in relationship  between  sites.
Significant positive correlations between water
temperature and chlorophyll a.  during drought
periods  from 1986 to  1990 suggest that the
observed differences were related to weather
variables.  ANCOVA of species diversity data
using  ELF  exposure  as a  covariate also
indicated  that  a  factor  other than  antenna
exposure was  responsible for  the significant
BACI and RIA results.

Along with the need for careful interpretations
regarding the possible sources of variation, data
set sample sizes  should  be considered when
deciding on the appropriate statistical analysis.
Generally,  the  statistical   power  of
non-parametric tests  such  as  RIA is smaller
than  that  of  a   similar  parametric  test
(Welkowitz et at.  1976). When populations are
normally   distributed,   the    number  of
observations required by  RIA should be larger
than  the  sample  size required  by  the BACI
analysis in order to obtain the same amount of
power.  Carpenter et al.  (1989)  reported that
RIA  could  consistently   detect  manipulation
effects with  sample  sizes  of 40  or more.
Stewert-Oaten et al. (1986) did not  suggest a

                                                Comparison of RIA and BACI Analysis
minimum sample  size for the  BACI analysis.
Although the authors did present an example in
which only 23 data points were used to detect
a manipulation effect, Monte Carlo simulations
are required  to  definitively determine  the
effective sample  size required  by the BACI
method. Thus, year-to-year BACI comparisons
presented in this  study  should be interpreted
with some caution, since  BACI's ability to
detect perturbation effects with sample sizes of
nine to  eleven (Table 1) data points remains
unknown.  The  sample  size  problem   also
emphasizes the  need for long term monitoring
of potential pollutant effects.

In summary,  an environmental  impact  study
should provide quantitative evidence to support
regulatory   decisions   regarding   potential
environmental  pollutants.   Both  the  BACI
analysis and RIA offer a means of quantitatively
detecting whether perturbations such as toxic
effluents, pipeline construction, or power plant
discharges  may be  impacting an ecosystem.
The BACI and RIA results should be interpreted
with some  care  and  caution  however,  as
significant  findings  do  not imply  that  the
suspected pollutant  has caused the observed
differences. Our analysis of  ELF effects on a
riverine  algal  community using BACI  and  RIA
suggests that the  following statistical protocol
will accurately  and  quantitatively allow  the
detection of environmental perturbations. First,
the parametric BACI  analysis should be used for
data sets satisfying plausible assumptions of
independence and additivity.  If the relationship
between control and  impacted  sites  has
changed significantly over  time, or if  the
independence,    normality   or   additivity
assumptions appear to be questionable, then
the non-parametric  randomized  intervention
analysis may be  used to  examine the data.
Finally, if the inter-site relationship is found to
change over time using  RIA, final conclusions
of the perturbations effects should be based on
ANCOVA results  (using  the  magnitude of the
perturbation  as  the covariate) and/or other
ecological considerations. When used in  this
manner, BACI and RIA represent complimentary
and practical tools with which to make sound
ecological  decisions   regarding   potential
environmental impacts.

We thank Stephen R. Carpenter for copies of
the programs  used  to  run the  randomized
intervention analyses. Jennifer Molloy and Mark
P.  Oemke  assisted  with   the   diatom
identification and enumeration. Support for this
research was provided by the Naval Electronic
Systems Command through a subcontract to IIT
Research  Institute  under   contract  numbers
N00039-81,   N00039-84-C-0070,   and

Literature Cited
Burton, T. M.,  D.  M. Mullen, and S. L. Eggert.
1991. Effects of extremely low frequency (ELF)
electromagnetic fields on the diatom community
of the Ford River,  Michigan, pp. 17-25 In: T.P.
Simon and W.S. Davis (editors). Proceedings of
the 1991 Midwest Pollution Control Biologists
Meeting. U.S. EPA Region  V, Environmental
Sciences  Division,  Chicago, IL.  EPA-905/R-

Carpenter,  S.   R.  1990.   Large-scale
perturbations:  opportunities  for   innovation.
Ecology 71:2038-2043.

Carpenter, S. R., T. M. Frost, D. Heisey, and T.
K.  Kratz.  1989.  Randomized   intervention
analysis   and  the  interpretation  of
whole-ecosystem   experiments.  Ecology

Durbin, J. and G. S. Watson. 1951. Testing for
serial correlation in least squares regression II.
Biometrika 38:159-178.

Hulbert, S. H. 1984. Psuedoreplication and the
design  of  ecological   field   experiments.
Ecological Monographs 54:187-211.

Jassby,  A. D.   and T. M. Powell.  1990.
Detecting changes  in ecological  time  series.
Ecology 71:2044-2052.

Eggert, Burton and Mullen
Reckhow,  K. H. 1990. Bayesian inference in
non-replicated  ecological  studies.  Ecology

Stewart-Oaten, A., W. W. Murdoch, and K. R.
Parker.   1986.   Environmental  impact
assessment:  "Pseudoreplication"  in  time?
Ecology 67:929-940.

Steel, R. G. and J. H. Torrie. 1960. Principles
and procedures of statistics. McGraw-Hill, New
York, 481  pp.

Tukey, J. W. 1949. One degree of freedom for
non-additivity.Biometrics 5:232-242.

Welkowitz, J., R. B. Ewen and J. Cohen. 1976.
Introductory  statistics  of  the  behavioral
sciences. 2nd  edition. Academic Press, New
York, 316 pp.

    The Freshwater Annelida (Polychaeta, Naidid and Tubificid  Oligochaeta,
             and Hirudinea) of the  Great Lakes Region-an Overview

Donald J. Klemm
Bioassessment and Ecotoxicology Branch
Environmental Monitoring  Systems Laboratory
U.S. Environmental Protection Agency
3411  Church Street, Cincinnati, Ohio 45244

Jarl K. Hiltunen
Sugar Island
Sault Ste. Marie, Michigan 49783

The segmented worms are important components of benthic communities in nearly every freshwater
biotope. They are widely distributed, and some groups are found in great abundance. Several of the
annelid groups have been  used for monitoring and detecting changes in water quality and physical
habitats. The habitat and water quality requirements as well as the  pollution tolerance of many
species of freshwater annelids have been documented in the literature by  a few investigators.
Practical taxonomic keys are now available to species, but many benthic water quality assessment
studies still do not treat the annelid groups adequately because the investigators lack the knowledge
and experience in using these keys. Furthermore, most bioassessment monitoring studies do not use
adequate sampling and processing (preservation) techniques for aquatic annelids. The inadequate
treatment  by some investigators represents a loss of valuable  ecological  information for use in
biological assessment of the quality of water resources, water pollution, or other changes in aquatic
ecosystems  resulting  from  natural causes or anthropogenic activities. The current aspects of
morphology, taxonomy, distribution, and organic  pollution  to polychaetes,  naidid  and  tubificid
oligochaetes, and leeches of the Great Lakes Region species are presented and discussed.

Kev Words: macroinvertebrates, polychaetes, oligochaetes, leeches, pollution, water quality, organic
Benthic animals,  including  the  segmented
worms, are commonly used to demonstrate the
effects of pollution on the biological integrity of
surface waters  and changes  in  the  biotic
community (species composition, presence or
absence, and relative abundance of tolerant and
intolerant species) resulting from natural causes
and destructive activities by man (Aston 1984;
Brinkhurst 1974a,b; Carr and Hiltunen  1965;
Goodnight and Whitley  1960; Hiltunen  1967,
1969a-c,  1971; Hiltunen and Manny  1982;
Howmiller and  Scott 1977; Milbrink  1983;
Sawyer 1974; Klemm 1991  and papers cited
therein). This paper is a taxonomic overview of
the freshwater polychaetes, naidid and tubificid
oligochaetes, and leeches of the Great Lakes
region   with   emphasis   on  their  use  to
demonstrate pollution effects and changes in
biotic community. A checklist of the species is
found in Table 1.

Annelida is an important  and major  phylum in
the animal kingdom. The body of annelids is
divided into rings (somites) or segments with
serially arranged organs.  The phylum includes
three major classes,  Polychaeta, Oligochaeta,
and  Hirudinea. The  distribution of  aquatic
annelids is usually determined by the physical,
chemical, and  biological characteristics of  the
environment. Published accounts are relatively
sketchy for understanding the roles that some
of these characteristics play in the distribution
of annelids (gross chemical pollution not

Klemm and Hiltunen
Table 1.  Checklist of Polychaetes, Naidid and Tubificid Oligochaetes, and Leeches in the Great Lakes
       Family Sabellidae
       Manavunkia soeciosa
       Family Naididae
       Allonais pectinata

       Amphicaeta america

       Amphicaeta levdioii

       Arcteonais lomondi

       Bratislavia unidentata
        Chaetogaster diaphanus
        Chaetoqaster diastroohus
        Chaetooaster limnaei
        Chaetogaster setosus

        Dero digitata
        Dero f urcata
        Dero nivea
        Dero obtusa
        Dero vaga

        Haemonais waldvoqeli

        Nais alpina
        Nais bretspheri
        Nais barbata
        Nais behninoi
        Nais communis
        Nais elinouis
        Nais pardalis
        Nais pseudobtusa
        Nais simplex
        Nais variabilis
                               Class Polychaeta: Order Sabellida
Class Oligochaeta: Order Tubificida

             Qphidonais seroentina
             Paranais frici

             Piguetiella michiaanensis
             Piouetiella blanci
             Pristina aequiseta
             Pristina breviseta
             Pristina leidyi
             Pristina lonqiseta bidentata
             Pristina lonaiseta lonoiseta
             Pristina plumaseta
             Pristina svnclites
             Pristinella acuminata
             Pristinella ienkinae
             Pristinella osborni

             Ripistes oarasita

             Slavina appendiculata

             Soecaria iosinae

             Steohensoniana trivandrana

             Uncinais uncinata

             Veidovskvella comata
             Veidovskvella intermedia

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
Table 1. Checklist of Polychaetes, Naidid and Tubificid Oligochaetes, and Leeches in the Great Lakes
        Region (continued).

                             Class Oligochaeta: Order Tubificida
       Family Tubificidae
       Aulodrilus americanus
       Aulodrilus limnobius
       Aulodrilus piaueti
       Autodrilus pluriseta

       Bothrioneurum veidovskvanum

       Branchiura sowerbvi

       Haber cf. soeciosus

       llvodrilus temoletoni

       Isochaetides f revi
       Isochaetides curvisetosus
       Limnodrilus cervix
       Limnodrilus cervix (variant form)
       Limnodrilus claoaredianus
       Limnodrilus hoffmeisteri
       Limnodrilus hoffmeisteri (soiralis form)
       Limnodrilus hoffmeisteri (variant form)
       Limnodrilus maumeensis
       Limnodrilus profundicola
       Limnodrilus udekemianus

       Phallodrilus hallae
Potamothrix bavaricus
Potamothrix bedoti
Potamothrix hammoniensis
Potamothrix moldaviensis
Potamothrix veidovskvi

Psammorvctides californianus

Quistadrilus multisetosus

Rhyacodrilus coccineus
Rhyacodrilus montana
Rhvacodrilus punctatus
Rhvacodrilus sodalis

Soirosperma ferox
Spirosperma nikolskvi

Tasserkidrilus harmani
Tasserkidrilus kessleri
Tasserkidrilus superiorensis

Teneridrilus flexus

Tubifex ianotus
Tubifex tubifex

Varichaetadrilus auoustipenis
                           Class Hirudinea: Order Arhynchobdellida
       Family Haemopidae
       Haemopis arandis
       Haemopis lateromaculata
       Haemopis marmorata
 Haemopis plumbea
 Haemopis terrestris
       Family Hirudinidae
       Macrobdella decora
 Philobdella gracilis

Klemm and Hiltunen
Table 1. Checklist of Polychaetes, Naidid and Tubificid Oligochaetes, and Leeches in the Great Lakes
         Region (continued).
       Family Erpobdellidae
       Eroobdella dubia
       Erpobdella parva
       Erpobdella punctata
        Mooreobdella bucera
        Mooreobdella fervida
        Mooreobdella microstoma
       Nepheloosis obscura

                           Class Hirudinea: Order Rhynchobdellida
       Family Glossiphoniidae
       Actinobdella annectens                            Helobdella fusca
       Actinobdella ineouiannulata                        Helobdella paoillata
       Actinobdella pediculata                            Helobdella staanalis
                                                        Helobdella transverse
       Alboqlossiphonia heteroclita                       Helobdella triserialis
       Desserobdella michiaanensis
       Desserobdella phalera
       Desserobdella Dicta

       Gloiobdella elonqata

       Glossiphonia comolanata
        Marvinmeveria lucida

        Placobdella hollensis
        Placobdella montifera
        Placobdella ornata
        Placobdella papillifera
        Placobdella parasitica
       Family Piscicolidae
       Cvstobranchus meveri
       Cvstobranchus verrilli

       Mvzobdella luoubris
        Theromvzon biannulatum
        Theromvzon rude

        Piscicola geometra
        Piscicola milneri
        Piscicola punctata

        Piscicolaria reducta
withstanding).  Despite the  fact that annelids
may occur in all aquatic habitats and in great
numbers, especially certain oligochaete groups,
it must be stressed that much work remains to
be done on the ecology and pollution biology of
the  annelids.  The  improper  or inadequate
treatment of the segmented worms is attribut-
able,  in  part,  to  investigators  that  lack
appropriate experience or do not understand the
morphological  terms and characters used in
practical keys. To interpret the quality of water
resources, the water quality requirements and
pollution tolerances,  the animals  should  be
identified  to  the  species  level  (Resh  and
Unzicker 1975).

Class  Polychaeta:  Order  Sabellida:   Family

General Morphology and Taxonomy
Polychaetes are segmented annelids typically
with  parapodia associated with  each body

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
segment. A high degree of modification in the
basic plan of many polychaetes has resulted in
different modes of  existence,  ranging  from
sedentary  (tubicolous)  forms to  highly  free-
moving (errant) forms.  The group is   very
diverse, most  forms are mainly found  free
living; some are commensal with other inverte-
brates, and only a few are parasitic. The class
contains about 85  families and many species,
most of which are marine forms.  World-wide,
only 10 families are represented in freshwater.

In North America  only  four families and 11
species are  represented  (Klemm  1985a,c):
Nereididae  with six species, Ampharetidae with
one species, Sabellidae with one species, and
Serpulidae with two species. In the Great Lakes
Region   only   the  family,   Sabellidae  is
represented  by  one  species,   Manavunkia
soeciosa; it is sedentary and inhabits a  tube
built of mud or sand and mucus. The size  of
this species is  usully 2 to 5 mm long. The body
is divided into distinct regions, the thorax and
abdomen, with reduced  or vestigal parapodia,
with  simple capillary chaetae and  hooks  or
uncini. The prostomium is small or  indistinct,
without appendages. In Sabellidae, the anterior
end is modified to form a branchial (tentacular)
plume or crown surrounding the mouth which
is used for food  getting  (filter feeding) and
respiration.  For more  information  on  the
taxonomy  of  the  North America freshwater
forms, see  Klemm (1985a,c).

General Distribution and Ecology
This species is widely distributed in the Neartic
region,  from  Duluth Harbor in  western  lake
Superior to St. Marys River, Lake St. Clair, the
Ottawa River in Ontario  to Western  Lake  Erie,
Lake Ontario, and the upper St. Lawrence River;
eastward to the Finger Lakes and Hudson River,
New York, Schuylkill River and Delaware River,
Pennsylvania,  Egg  Harbor River, New Jersey,
and Lake Champlain, Vermont, south to North
Carolina, South Carolina, and Georgia.  It has
also been reported  from the Pacific northwest,
California, and Oregon to Alaska. One reason
for not detecting this species more often is its
small size; the sieve used to process the sample
may have mesh openings too large to retain the
specimens. Most specimens may pass through
a Standard No. 30 sieve,  and  a  Standard No.
60 should be used.

Mackie  and Qadri (1971) reported,  during a
limnological survey of the Ottawa River, that
specimens of M-  soeciosa  occurred only  in
substrates composed  of silt and sand and in
moderately  moving waters. Hiltunen (1965)
also found a large number of M- speciosa at the
mouth of  the Detroit  River, suggesting  some
relationship between water movement and the
frequency  of  occurrence. Specimens in the
Mackie  and  Qadri study did not  occur  in
polluted water  where BOD value exceeded 4
ppm nor where the DO content was less than 5
ppm. M- speciosa has also been found in lentic
habitats, many  locations in Lake  Erie  (Hiltunen
1965, Krieger 1990), and in several lakes in
Alaska  (Holmquist 1973).  Spencer (1976)
found it in Cayuga Lake, New York at depths of
20  m or  less where  densities of more than
I000/sq. m were found occasionally. Poe and
Stefan (1974) reported this polychaete from the
Schuylkill  River, Pennsylvania, near the type-
locality, and  they  reported that  it appears  to
have a  wide range of tolerances for  environ-
mental  parameters such  as DO (1.8  to 14.0
ppm), depth (0.3 to 16.0 m), pH (6.8 to 8.8),
and   water  temperature   (2.8   to   28.3°C)
(obviously as low  as 0°C  because it survives
ice-cover seasons), and concluded that the only
environmental  factor  which  may  limit  its
distribution   is  the  requirement   for  fine
paniculate material in the substrate for the
construction of the tube in  which it lives (gross
chemical pollution notwithstanding).

Class Oligochaeta

Naidids  and tubificids are predominantly found
in freshwater but some  are  strictly marine
forms. Most oligochaetes have chaetae, with a
few exceptions, and have no  parapodia as in
the polychaetes. The body is  segmented into

Klemm and Hiltunen
somites or compartments separated by septum,
and each segment by convention is indicated by
a Roman numeral, progressing from anterior to
posterior.  Segment  I  (including  mouth and
prostomium)  is devoid  of chaetae,  hence
numerical  orientation of segments is achieved
by counting  posteriad of the chaetophorous
segments,  beginning  with II.  Normally each
segment bears  four  fascicles "bundles" of
chaetae, two dorso-lateral  and two  ventro-
lateral. There are two basic types of chaetae
(crotchets and capillif orms) whose numbers and
morphology in  the various  body regions  are
taxonomically important.  A crotchet  can be
straight  or  curved  (sigmoid),  and  usually
possess a more or less median thickening (the
node or nodulus), and may be simple-pointed or
have a bifid (cleft) distal end.  Crochets are
found  in all oligochaetes. Capilliform chaetae
which are elongate and simple-pointed, may be
smooth or finely serrated. Capilliform chaetae
when present, are found only in the dorsum of
the Naididae  and Tubif icidae. For more detailed
information on aquatic oligochaete biology, see
Brinkhurst and Jamieson (1971), Brinkhurst and
Cook (1980), Bonomi and Erseus (1984), and
Brinkhurst and Diaz (1987).

Order Tubificida: Family Naididae

General Morphology and Taxonomy
Naidids are relatively small, commonly I mm to
10 mm and more or less transparent when alive.
All or nearly  all can be identified to species by
the external morphology, particularly the shape
and arrangement of the chaetae. There are 20
genera and more or less 48 species known or
likely to occur in the Great Lakes region. In
North America 21 genera and 75 species are
reported. Keys work well for most species, but
some  species descriptions are incomplete for
North American material. The kinds of chaetae
are much like those in the Tubificidae, except
the naidids have dorsal acicular (short  needle-
like) chaetae that accompany the long  capilli-
form (hair) chaetae.  Some  species of naidids
also bear pectinate chaetae like the tubificids.
Dorsal chaetae can  begin  in segment II or
posteriad to it; dorsal chaetae that accompany
capilliform  (hair)  chaetae  are  often  very
different from ventral chaetae. Dorsal fascicles
often contain 1-2  capilliform chaetae and I-2
acicular chaetae. In summary, some naidids
have ventral chaetae only (Chaetooaster spp.);
other species have dorsals and ventrals with
bifids only, while  still  other species  have
ventrals bifid and dorsals with capilliform plus
simple,  bifid, pectinate, or palmate  acicular
chaetae. The dorsal chaetae may begin in II, III,
or further back, usually V or VI, rarely beyond.
Some species may have eyes; may be found
budding (a form of asexual reproduction), and
when  sexually  mature,  may  bear  genital
chaetae in  segments V or VI; spermathecae in
segments IV, V, or VII; male pores on segments
V, VI, or VIII. However, these features are not
used  in  species  identification.   For   more
information on  taxonomy of the  naidids, see
Hiltunen and Klemm (1980, 1985), Brinkhurst
(1986), and Brinkhurst and Kathman (1983).

General Distribution and Ecology
The naidids are an ecologically diverse group
(Leaner 1979) and  are found in both lotic and
lentic waters. The naidids are widely distributed
and commonly inhabit the littoral zones of lakes
or other shallow waters in streams, ditches,
and ponds. Some species are sediment dwellers
(like  tubificids)   while  other  species  are
characteristically found among  the aquatic
plants. Naidid populations are usually reduced
where siltation and mud occur. Plants  with  a
thick growth  habit and well-developed peri-
phyton community can support sizeable naidid
populations. Riffles and similar areas where the
substrate is  primarily  sand  and  gravel often
contain substantial naidid  populations.  Longi-
tudinal zonation of naidids along rivers has been
demonstrated.  Learner et al.  (1978) concluded
that factors associated with changes in altitude
and slope  of a river (water velocity, substrate
type, presence and type of vegetation, and the
influence of municipal  and industrial wastes)
can be important in influencing the distribution
of naidids. They are generally less significant in
lakes where they are confined primarily to the

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
littoral  zone. The behavior of most naidids is
unlike that of most other oligochaetes because
some naidids can swim as well as crawl, and
others are small enough to be passively carried
by strong water movements.

Feeding habits of most species are unknown,
but some  have been observed  to  feed on
detritus, grazing upon bacteria, protozoans, and
algae.   Probably  most  oligochaetes   are
herbivorous, but some Chaetooaster species are
primarily or perhaps  entirely   predaceous.
Reproduction occurs by paratomy (architomy,
asexually   budding),  where  the   posterior
segments of the naidid develop into  daughter
zooids  that break free  after development is
complete, or by fragmentation. Sometimes the
worms are found  consisting of two or more
individuals that have not yet separated from the
parent (anterior) section. Sexual reproduction is
considered  uncommon in many species.

Order Tubificida: Family Tubificidae

General Morphology and Taxonomy
Tubificids are  medium-sized to large worms,
commonly more than 20 mm long, that never
have eyes, never reproduce by asexual budding,
but occasionally regeneration of the posterior
section in  some species  suggests fragmen-
tation.  A variety of chaetae  are found among
the species. Crotchets are always present but
capilliform  (hair) chaetae may or may not be
present depending on the species. Most species
are red when  alive and  coil or loop  when
disturbed. In the Great Lakes region there are
presently 17 genera and 37 species. There are
21 genera and 64 species reported to occur in
North America. Tubificids are identified  by the
characteristic shape of the somatic chaetae and
their genital chaetae (spermathecal or penial
chaetae) if present, or by mature male genitalia.
In some species penis sheaths in segment XI
are especially helpful in species identification.
Spermathecae are located in segment X, and
males pores are in Segment XI. Dorsal chaetae
always begin on segment II, dorsal chaetae are
often broadly similar in form to ventral chaetae;
dorsal  fascicles  often bear a  complement  of
more than 2 capilliform (hair) chaetae and 2 or
more crotchets. Therefore,  some species  of
tubificids have dorsal and ventral chaetae bifid;
other  species  have  dorsal  capilliform  and
pectinate  chaetae and ventrals mostly bifid.
Pectinate chaetae may be narrow and hairlike
distally in  appearance.  Some species have
dorsal capilliform and bifid chaetae and ventral
bifid chaetae. For  more  information on  the
taxonomy  of  the  tubificids,  see  Stimpson,
Klemm,  and  Hiltunen  (1982,  1985)  and
Brinkhurst (1986, 1989).

General Distribution and Ecology
Tubificids are most commonly found in soft
sediments  rich  in  organic  matter; several
tubificid species characteristically live in large
numbers  in  habitats  that  receive  organic
pollution (Aston 1984,  Brinkhurst  1974a,b,
Carr and Hiltunen 1965, Goodnight and Whitley
1960. Hiltunen 1967,1969a-c, 1971, Hiltunen
and  Manny 1982, Howmiller and  Scott 1977,
Krieger 1990, Milbrink 1983). Tubificids respire
cutaneously, but some  species  can tolerate
anoxic conditions and  environmental stresses
(e.g.,  Limnodrilus  hoffmeisteri and  Tubifex
tubifex). A number of species in the family are
very stress-sensitive (Hiltunen 1967, Howmiller
and  Scott 1977, Milbrink  1983). Tubificids
burrow in soft sediments, often in tubes of mud
and  mucus secretions, as the classic name
implies. A few species occur in fine gravel or
sand.  The  quantity and quality of organic
matter reaching the sediment may be more
important in determining which tubificid species
will occur in a locality (gross chemical pollution
notwithstanding) than  the  physical-chemical
variables of water or sediment (Brinkhurst and
Cook 1974).

Tubificids are mostly deposit feeders living on
organic detritus and its associated bacteria,
microflora, and fauna. Tubificids typically feed
with their  heads buried below the sediment
surface with their tails protruding above it. The
feeding activities of tubificids play an important
role  in mixing  the  physical  and  chemical

Klemm and Hiltunen
characteristics of sediments (Brinkhurst 1974).
Most  tubificids  reproduce by  sexual  repro-
duction even though they are hermaphrodites.
Tubificids  enclose their eggs in cocoons and
deposit them on sediments.

Class  Hirudinea (Not Hirudinoidea)

Leeches are serially  segmented worms and are
considered closely related to the oligochaetes.
They are also hermaphroditic, i.e., they contain
both male and female organs in each individual.
Muscles and a hydrostatic skeleton are used in
locomotion.  The   nervous,  excretory  and
vascular  systems are  segmentally arranged.
Leeches   have well-developed  anterior  and
posterior  sucker, 34 segments (indicated by
Roman  numerals I-XXXIV))  which are  sub-
divided  into annul!, a  reduced coelom and
intestinal caeca, and usually two separate male
and  female gonopores with male gonopore
anterior to the female  gonopore. Leeches are
devoid  of  chaetae,   except  Acanthobdella
oeledina.  a leech which has chaetae  in the
anterior  segments  of  the  cephalic  region
(Klemm 1985b,c). Although some  leeches are
well adapted to a sanguivorous existence, the
group is also well-represented by species which
are both predatory and can engulf small animals
whole or  parasitic  fluid-feeders. Leeches are
found on all  the  continents,  in terrestrial,
freshwater,   estuarine,  and  marine

General Taxonomy and Morphology
Five families in the orders Arhynchobdellida and
Rhynchobdellida are represented in the Great
Lakes  region:   Haemopidae,    Hirudinidae,
Erpobdellidae, Glossiphoniidae, andPiscicolidae.
Nineteen  genera and  43 species have  been
recorded from that  region. Twenty-one genera
and 66 nominal  species presently are reported
to occur  in North  America. For  more infor-
mation on the taxonomy of leeches, see Klemm
(1985b,c, 1991) and Sawyer (1972,1986).

Leeches are found in most freshwater habitats,
but are often ignored by biologists because they
are thought to be difficult to identify to genus
and species. Also, investigators  neglect them
because  they  lack an understanding of  the
diagnostic (morphological) terms and characters
used  in  keys  to  identify  specimens to  the
lowest taxonomic level.

The  external  morphological   characters  are
usually sufficient for the identification  of most
leeches.  Internal characters used  to  identify
certain species of Haemopis are discussed in
Klemm   (1985b,c).   The  general   external
diagnostic  features  that  are  important  for
identifying the leeches to species are: size of
mouth, general shape of body, form of suckers,
form   of  cephalic  region,   number   and
arrangement of eyes, jaws and teeth, eyespots
(ocelli),  papillae,  pulsatile vesicles,  digitate
processes on  rim of caudal  sucker,  caudal-
sucker separation from body on narrow pedicle,
copulatory gland pores, the number of annul!
between gonopores, and pigmentation patterns.
Typically, the mouth opening of the haemopids
and hirudinids is medium to large, occupying
the entire sucker cavity, and the body is large,
linear, elongate and well-muscled, length 75-
300 mm. They are good swimmers. Haemopids
and  hirudinids  always have 5 pairs of eyes.
The mouth opening of erpobdellids is medium,
occupying the entire sucker cavity; the body of
erpobdellids is moderate  size,  linear, elongate,
length to 100 mm, and  they are also good
swimmers. They usually  have 3, or 4 pairs of
eyes   (or  eyes  absent). The  mouth  of
glossiphoniids  is a small pore on the rim or
within the oral sucker cavity; the body of this
group is dorso-ventrally  flattened with the
posterior half  usually much  wider than the
tapering cephalic end, length to 40 mm. They
have I, 2, 3, or 4 pairs of  eyes. The mouth of
piscicolids is a small pore within the oral sucker
cavity,  and  the  body of the  piscicolids  is
cylindrical, narrow, posterior half can be slightly
flattened, length to 30 mm. The body may be
divided into a narrow neck (trachelosome) and
wider body (urosome) regions;  caudal sucker
with or without eyespots, and body with or
without pulsatile vesicles.  Piscicolids can have

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
one or two pairs of eyes (or eyes absent).

General Distribution and Ecology
In North America,  freshwater leeches reach
their  greatest   species  diversity  in  lakes,
permanent  and  temporary ponds,  woodland
pools,  bogs, wetlands,  rivers,  and streams
(Klemm 1972, Sawyer 1972). Klemm (1977,
1991) summarized the distribution of leeches in
the  Great  Lakes  states (Illinois,  Indiana,
Michigan, Minnesota,  New York, Ohio,  and
Wisconsin) bordering the Great Lakes, including
Ontario, Canada. Additional  collecting in all
habitats, substrate types, and host organisms
will undoubtedly extend the regional distribution
of some taxa.

The leeches  of the Great Lakes  region are  a
significant  part of the continental freshwater
fauna.  Leeches are biologically  important in
food webs, and  at trophic levels  they function
mainly as ectoparasites, predators, or both. An
important ecological factor in the distribution of
leeches  is  the  availability  of   prey. Other
environmental factors, such as composition of
substratum, lentic or lotic waters, depth, size
and type of water body,  hardness and  pH,
dissolved solids, water temperature,  dissolved
oxygen, siltation and turbidity, and salinity are
characteristics  of aquatic  habitats that  also
influence  leech  distribution  and abundance
(Klemm  1972,   1991;  Sawyer  1974),  not
withstanding toxic  substances in the aquatic
environment.  We have  a relatively sketchy
understanding of the role that these and other
environmental factors play in the distribution of
leeches. It must  be stressed  that much work
remains to be done before we can have a clear
picture  of the problem.

The little we know of the feeding  habits of
leeches indicates that they are far more diverse
than  most  people  realize;  many  are  not
sanguivorous (blood feeders). The Haemopidae
have no  teeth  or have  varied and poorly
developed  jaws  that  are  armed with small
numbers of blunt teeth for masticating food,
and are able  to swallow  prey whole. They are
mainly predators of macroinvertebrates, but the
ones with teeth are perhaps capable of sucking
blood. In the  Hirudinidae,  the jaws are well
developed,  armed with numerous small, sharp,
saw-like teeth suitable for making cuts in the
epidermis of prey, such as reptiles, amphibians,
and  mammals.    These  leeches are  blood
sucking ectoparasites.  Surprisingly, studies on
the North American haemopids and hirudinids
indicate that these leeches are predominantly
predatory and extremely  opportunistic,  and
consume larvae and eggs of  amphibians  and
small invertebrates.  They  dwell mostly in
freshwater,  but   some  species  can  travel
overland, and  a few species are terrestrial. In
Erpobdellidae,  the mouth is large and adapted
to predation. It contains a muscular pharynx for
crushing  and   swallowing  macroinvertebrate
prey  whole. They are highly  mobile and are
good swimmers. They live   exclusively in
freshwater. The  Glossiphoniidae are without
teeth or jaws and  have  a   very small  oral
opening  (pore).  This  name   refers  to  the
mechanism by which these leeches feed. They
insert a tube-like  proboscis into their prey and
suck out the  body fluids.  The glossiphoniids
parasitize turtles, mollusks, waterfowl, fishes,
amphibians, mammals, including man, and even
other  leeches.  They  are  ectoparasites  or
predators.  Most travel slowly with  a  looping
movement,  but  a few  species are  active
swimmers. Brooding behavior is well developed,
and  cocoons  are brooded over substrate or
directly on the venter  of the parent. They are
found  exclusively   in   freshwater.   The
Piscicolidae are primarily ectoparasites on fish.
Some are   permanent  parasites  on specific
hosts, but most are opportunistic and feed on
a  variety  of  host fishes. A  few  feed  on
invertebrate groups,  such a   Decapoda  and
Cephalopoda.  Piscicolids generally have a large
and well-developed anterior sucker surrounding
the mouth. As in the glossiphoniids, the oral
opening  is  small. To feed, piscicolids insert a
protruding  sucking proboscis  into their host.
Parent leeches lay hard-shelled cocoons on a
substrate, but they do not brood their cocoons
or young.  Only a few piscicolids are active

Klemm and Hiltunen
swimmers.  Muscles  are  generally  poorly
developed and locomotion is usually by looping
movements  over a substrate  or  host. Most
species are  marine,  but some are found in
brackish or fresh waters.

Tolerance to Organic Pollution
The  species  of  annelids in  the Great Lakes
region  that   are  commonly  or occasionally
associated with organically enriched waters are
indicated in Table 2A-C. The tolerance values in
Table  2A-C  can be used  with the  Trophic
Condition Index (Howmiller and Scott 1977;
Milbrink 1983) and modified Hilsenhoff Biotic
Index (Klemm et al. 1990; Plafkin et al. 1989).
However, more pollutional studies are needed
for these annelid groups because little is known
about their tolerances and the biological effects
of various contaminants. For more information
on the water quality requirements and pollution
tolerance of freshwater naidids, tubificids, and
leeches,  see Brinkhurst (1974a,b), Carr and
Hiltunen  (1965),  Hiltunen (1967,  1969a-c),
Howmiller and Scott  (1977),  Klemm (1972,
1991), Klemm et  al. (1990), Krieger  (1990),
Milbrink (1983). and Sawyer (1974).

General Collection  and Preservation
Aquatic  worms are usually  collected using
dredges,  grabs,  cores and  other sampling
devices that provide bulk collections of bottom
subtrate. This material  is then  sieved or hand-
picked so that  the organisms are separated
from the accompanying silt and  debris.  This
must be  done carefully, especially if a sieve is
used. The abrasion of  the soft-bodied worms
against a sieve surface may break specimens or
damage   the  specimens  by  breaking   or
displacing chaetae, particularly capillif orm (hair)
chaetae, for example. Although a US Standard
No,  30 mesh sieve (28 meshes per inch, 0.595
mm openings) is  usually  used, it  should be
noted  that many small individuals may be lost
during the sieving  process and that the use of
a finer sieve (for example, No. 60 mesh, 0.25
mm opening)  or  no  sieving  at  all  may be
required  to ensure collection of all individuals.
Even when sieving has been accomplished care-
Table 2A. Pollution Tolerance of Selected
         Freshwater Annelids
Tolerance to Organic Wastes*
             T      F      I
Manavunkia soeciosa

Amohichaeta americana
Chaetogaster diaohanus
C. diastroohus
Dero dioitata
D. nivea
D. obtusa
D. pectinata
Naj§ barbata
N. behninqi
N. bretscheri
N. communis
N. elinauis
N. oardalis
N. simplex
N. variabilis
Qphidonais serpentina
Pristina aeauiseta
Slaving aopendiculata
Specaria josinae
Stvlaria fossularis
S. lacustris
Veidovskvella comata
V. intermedia


 * Ranking from 0 to 5 with 0 being the least
 tolerant. T  = tolerant; F  = facultative; I =
 fully,  some   individuals   will   nevertheless
 fragment.  Only head-end sections and whole
 worms should be enumerated. The initial sorting
 of specimens  from sediment  residue in the
 laboratory  should  be  done  at  a  5-10  X
 magnification using a dissection microscope or

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
Table 2B. Pollution Tolerance of Selected
         Freshwater Annelids.

            Tolerance to Organic Wastes*
Taxa                       T      F      I
Aulodrilus americanus
A. limnobius
A. DJoueti
4, pluriseta
Bothrioneurum veidovskvanum
Branchiura sowerbvi         4
llvodrilus tenrmletoni
Isochaetides curvisetosus
Limnodrilus cervix
L. claoaredianus
L. hoffmeisteri
L. maumeensis
L. udekemianus
Potamothrix moldaviensis
P. vejdovskvi
Quistadrilus multisetosus
Spirosperma carolinensis
S. ferox
S. nikolskvi
Tubifex tubifex


 'Ranking from 0 to 5 with 0 being the least
tolerant. T  = tolerant; F = facultative; I  =
lens. They can also be selectively hand-picked,
fixed, and preserved in the field.

Leeches are also  found attached to various
substrates  such  as rocks,  boards,  logs,  or
almost any inanimate object littering both lentic
and lotic environments or collected from prey
organisms. Annelid specimens should be fixed
in 5010% formalin,  and transferred after 48
hours  to 70%  ethanol  or 5-10%  buffered
formalin for storage. Undesirable  shrinkage is
kept to a minimum with this process. The use
of alcohol as a fixative should be avoided
Table 2C. Pollution Tolerance of Selected
          Freshwater Annelids

              Tolerance to Organic Wastes*
     Taxa                   T      F      I

Eroobdella parva             4
E. ounctata                  4
Mooreobdella microstoma     4

Haemoois grandis                   3
H. marmorata                      3

Alboolossiphonia heteroclita         3
Gloiobdella elonoata         4
Helobdella stagnalis          4
H. triserialis                        3
Glossiohonia  complanata     4
Placobdella multilineata             2
P. ornata                           3
P. papillifera                        3
P. parasitica                        3

MvzobdeHa luoubris                 3
Piscicola punctata                  3

* Ranking from 0 to 5 with 0 being the least
tolerant. T = tolerant; F = facultative.
                    because polychaetes, oligochaetes, and leeches
                    initially preserved in alcohol without first being
                    fixed  in formalin tend  to  deteriorate  and
                    disintegrate. If the specimens of oligochaetes
                    are to be cleared and they have been preserved
                    in 70% alcohol, they should be placed in 30%
                    alcohol and then in water for a short time to
                    leach out the alcohol to enable placement into
                    a  tissue-clearing  solution  (e.g.,  Amman's
                    lactophenol).   Alcohol  retards  the  clearing
                    process of Amman's lactophenol.

Klemm and Hiltunen
To  allow internal  structures to  be  seen,
oligochaete specimens should be cleared before
specific  examination.  Temporary  mounting
media, Amman's lactophenol (lOOg phenol, 100
mL lactic acid, 200 mL glycerine, and 100 mL
water) or CMCP-9, or CMCP-10, can be used
for rapid processing of specimens. Oligochaete
specimens must be cleared  and mounted on
glass slides for examination under a compound
light microscope capable of magnification up to
1000X (oil immersion). An  18 mm diameter,
No. 0 or  1  round cover  glass is appropriate
because it will adequately accommodate nearly
the size range of naidids and tubificids and the
shape  allows for maneuvering the specimens
into  the  most  desired  position  by  gentle
pressure and rotation of the coverglass. When
preparing  a temporary or  permanent  slide
mount, an attempt should be made to place the
specimen on its side, revealing both dorsal and
ventral fascicles of chaetae. Permanent mounts
of oligochaetes can be made following alcohol
dehydration  of specimens and clearing, using
methyl salicylate or xylene, and mounting the
specimens  in a  synthetic  resin,  such as
Harleco's  Coverbond  or   Canada   balsam.
Permanent mounts of oligochaetes are suitable
for systematic study and may last over 20
years. Most leech specimens can be identified
to species by examining the external features
using  a   dissecting  microscope   (450X).
Additional instructions for sorting, processing,
and identifying polychaetes, naidid and tubificid
oligochaetes, and  leeches specimens can be
found  in a number  of  taxonomic  guides
(Brinkhurst 1986;  Hiltunen  and Klemm 1980;
Stimpson, Klemm, and  Hiltunen 1982; Klemm
1985a-c; Klemm at al. 1990).

Literature Cited
Aston, R. J. 1984. Tubificids and water
quality. A review. Environmental Pollution

Bonomi, G. and C. Erseus (eds.). 1984.
Aquatic Oligochaeta. Hydrobiologia 115:1-
Brinkhurst, R.O. 1974a. The Benthos of lakes.
St. Martin's Press, New York. 190 pp.

Brinkhurst, R.O. 1974b. Aquatic earthworms
(Annelida: Oligochaeta). ]n: C.W. Hart, Jr. and
S.L.H. Fuler (eds.). Pollution ecology of
freshwater Invertebrates. Academic Press,
Inc., New York, pp. 3-156.

Brinkhurst, R.O. 1986. Guide to the
freshwater aquatic microdrile oligochaetes of
North America. Canadian special publication
of fisheries and aquatic sciences 84,
Canadian Government Publishing Centre,
Supply and Services Canada, Ottawa,
Ontario, Canada K1A OS9. 259  pp.

Brinkhurst, R.O. 1989. Varichaetadrilus
auoustipenis (Brinkhurst and Cook 1966),
new combination for Limnodrilus auaustipenis
(Oligochaeta; Tubificidae). Proc. Biol. Soc.
Wash. 102(2):311-312.

Brinkhurst, R.O. and D.G. Cook  (eds.). 1980.
Aquatic oligochaete biology. Plenum, New
York. 529 pp.

Brinkhurst, R.O. and R.J. Diaz (eds.). 1987.
Aquatic Oligochaeta. Hydrobiol. 155:1-323.

Brinkhurst, R.O. and B.G.M. Jamieson. 1971.
Aquatic Oligochaeta of the world. Toronto:
Univ. Toronto Press. 860 pp.

Brinkhurst, R.O. and R.D. Kathman. 1983. A
contribution to the taxonomy of the Naididae
(Oligochaeta) of North America. Can. J. Zool.

Carr, J.F. and J.K. Hiltunen.  1965. Changes
in the bottom fauna of western Lake Erie from
1930 to 1961. Limnol. Oceanogr. 16:551-

Goodnight, C.J. and L.S. Whitley. 1960.
Oligochaetes as indicators of pollution. Proc.
15th Ind. Waste Conf., Purdue Univ.,  Indiana,
pp. 139-142.

                                     Great Lakes Region Oligochaeta, Polychaeta, and Hirudinea
Hiltunen, J.K. 1965. Distribution and
abundance of the polychaete, Manavunkia
soeciosa Leidy, in western Lake Erie. Ohio J.
Sci. 65:183-185.

Hiltunen, J.K. 1967. Some oligochaetes from
Lake Michigan. Trans. Am. Microsc. Soc.

Hiltunen, J.K. 1969a. Distribution of
oligochaetes in western Lake Erie. 196I.
Limnol. Oceanogr. 14(21:260-264.

Hiltunen, J.K. 1969b. Invertebrate
macrobenthos of western Lake Superior.
Mich. Academician 1(3-4):123-133.

Hiltunen, J.K. 1969c. The benthic
macrofauna of Lake Ontario. In: Limnological
Survey of Lake Ontario. 1964, pages 39-50.
Great Lakes Fish. Comm. Tech. Rep., No. 14.

Hiltunen, J.K. 1971. Limnological data from
Lake St.Clair, I963 and 1965. Dept.
Commer., NOAA/NMFS, Data Rept. No.
54(CON-71-00644). 54 pp.

Hiltunen, J.K. and D.J. Klemm.  1980. A guide
to the Naididae (Annelida: Clitellata:
Oligochaeta) of North America. U.S.
Environmental Protection Agency, Office of
Research and Development,  Environmental
Monitoring and Support Laboratory,
Cincinnati, Ohio 45268. EPA-600/4-80-031.

Hiltunen, J.K. and D.J. Klemm.  1985.
Freshwater Naididae (Annelida: Oligochaeta).
In: D.J. Klemm (ed.). A guide to the
freshwater annelida (Polychaeta, Naidid and
Tubificid Oligochaeta, and Hirudinea) of North
America. Kendall/Hunt Publ.  Co. Dubuque,
Iowa. pp. 24-43.

Hiltunen, J.K. and B.A. Manny.  1982.
Distribution and abundance of
macrozoobenthos in the Detroit River and
Lake St. Clair, 1977. Great Lakes Fishery
Laboratory Administrative Report No. 82-2,
U.S. Fish & Wildlife Service, Ann Arbor,
Michigan, pp. 87.

Holmquist, C. 1973. Fresh-water polychaete
worms of Alaska with notes on the anatomy
of Manavunkia speciosa Leidy. Zool. Zb. Syst.
Bd. 100,5.497-516.

Howmiller, R.P. and M.A. Scott. 1977. An
environmental index based on relative
abundance of oligochaete species. JWPCF

Klemm, D.J. 1972. The leeches (Annelida:
Hirudinea) of Michigan. Mich. Academician

Klemm, D.J. 1977. A review of the leeches
(Annelida: Hirudinea) in the Great Lakes
region. Mich. Academician 9(4):397-418.

Klemm, D.J. 1985a. Freshwater Polychaeta.
]n: D.J. Klemm (ed.). A guide to the
freshwater Annelida (Polychaeta, Naidid and
Tubificid Oligochaeta, and Hirudinea) of North
America. Kendall/Hunt Publ. Co., Dubuque,
Iowa. pp. 14-23.

Klemm, D.J. 1985b. Freshwater leeches. ]n:
D.J. Klemm (ed.). A guide to the freshwater
Annelida (Polychaeta, Naidid and Tubificids
Oligochaeta, and Hirudinea) of North America.
Kendall/Hunt Publ. Co., Dubuque, Iowa. pp.

Klemm, D.J. (ed.). 1985c. A guide to the
freshwater Annelida (Polychaeta, Naidid and
tubificid Oligochaeta, and Hirudinea) of North
America. Kendall/Hunt Publ. Co., Dubuque,
Iowa. 198 pp.

Klemm, D.J. 1991. Taxonomy and pollution
ecology of the Great Lakes region leeches
(Annelida: Hirudinea). Mich. Academician,

Klemm, D.J., P.A. Lewis, F. Fulk, and J.M.
Lazorchak. 1990. Macroinvertebrate field and

Klemm and Hiltunen
laboratory methods for evaluating the
biological integrity of surface waters.
Environmental Monitoring Systems
Laboratory, U.S. Environmental Protection
Agency, Cincinnati, Ohio 45268.

Krieger, K.A. 1990. Changes in the benthic
macrotnvertebrate community of the
Cleveland Harbor area of Lake Erie from 1978
to 1989. Ohio EPA, Division of Water Quality
Planning and Assessment, Columbus, Ohio

Leaner, M.A. 1979. The distribution and
ecology of the Naididae (Oligochaeta) which
inhabit the filter-beds of sewage-works in
Britain. Water Res. 13:1291-1299.

Leaner, M.A., G. Lochhead, and B.D. Hughes.
1978. A review of the biology of  British
Naididae  (Oligochaeta) with emphasis on the
lotic environment. Freshwater Biol. 8:357-

Mackie, G.L. and S.U. Qadri. 1971. A
polychaete, Manavunkia soecuisa. from the
Ottawa River, and its North American
distribution. Can. J. Zoology 49:780-782.

Milbrink,  G. 1983. An improved
environmental index based on the relative
abundance of oligochaete species.
Hydrobiologia 102:89-97.

Plafkin, J.L., M.T. Barbour, K.D. Porter, S.K.
Gross, and R.M. Hughs. 1989. Rapid
bioassessment protocols for use in streams
and  rivers: benthic macroinvertebrates and
fish.  EPA/440/4-89-001 (printed erroneously
as EPA/444/4-89-001). Assessment and
Watershed Protection Division, USEPA,
Washington, D.C. 20460

Poe, T.P. and D.C. Stefan 1974. Several
environmental factors influencing the
distribution of the fresh-water polychaete,
Manavunkia soeciosa Leidy. Chesapeake Sci.
Resh, V.H. and J.D. Unzicker. 1975. Water
Quality monitoring and aquatic organisms: the
importance of species identification. J. Wat.
Pollut. Control Fed. 47:9-19.
Sawyer, R.T. 1972. North American
freshwater leeches, exclusive of the
Piscicolidae with a key to all species.
Monogr. 46(1 ):1-154.
I. Biol.
Sawyer, R.T. 1974. Leeches (Annelida:
Hirudinea). in: C.W. Hart, Jr. and S.L.H. Fuller
(eds.). Pollution ecology of freshwater
invertebrates. Academic Press, Inc., New
York. pp. 81-142.

Sawyer, R.T. 1986. Leech biology and
behavior. Volume II: Feeding biology, ecology,
and systematics. Clarendon Press, Oxford.
pp. 418-793.

Spencer, D.G. 1976. Occurrence of
Manavunkia soeciosa (Polychaeta: Sabellidae)
in Cayuga Lake, New York, with additional
notes on its North American distribution.
Trans. Am. Microsc. Soc. 95(1 ):127-128.

Stimpson, K.S., D.J. Klemm, and J.K.
Hiltunen. 1982. A guide to the freshwater
tubificidae (Annelida: Clitellata: Oligochaeta)
of North America. U.S. EPA, Environmental
Monitoring and  Support Laboratory,
Cincinnati,  Ohio 45268, EPA-600/3-82-033.
61 pp.

Stimpson, K.S., D.J. Klemm, and J.K.
Hiltunen. 1985. Freshwater Tubificidae
(Annelida: Oligochaeta). ]n: D.J. Klemm (ed.).
A guide to the freshwater Annelida
(Polychaeta, Naidid and Tubificid Oligochaeta,
and Hirudinea) of North America. Kendall/
Hunt Publ.  Co., Dubuque, Iowa. pp. 44-69.

     A Comparison of Macroinvertebrates Collected from Bottom Sediments
                             in Three Lake Erie Estuaries
Philip A. Lewis
Bioassessment and Ecotoxicology Branch
Environmental Monitoring Systems Laboratory
U.S. Environmental Protection Agency
3411 Church Street, Cincinnati, Ohio 45244

Mark E. Smith
Technology Applications Inc.,
C\0 U.S. Environmental Protection Agency,
3411 Church Street, Cincinnati, Ohio 45244

Macroinvertebrates were collected from bottom sediments from three Lake Erie tributaries as a part
of EMSL's Biomarker  Project.  The objective of  this paper is to compare the macroinvertebrate
populations collected from the three water bodies and relate these populations to possible pollutional
stresses and/or habitat characteristics. Three grab samples were collected with either a petite Ponar
or a  standard Ekman-on-a-stick at three different stations at each site. The sampling stations were
chosen randomly from among the nine stations used for collecting fish at each site. In the Black River
above a Coking Plant, 60-80% of the organisms were tolerant oligochaete worms but some pollution
sensitive organisms were also present indicating organic enrichment but not toxic  pollution. All of
the individuals collected from below the plant were oligochaete worms (90%) and other organisms
tolerant of both  organic and toxic pollution. In  Old Woman Creek, over 80% of the individuals
collected were oligochaete worms and blood worms (midges) characteristic of organically enriched
sediments associated with high oxygen demand.  Toussaint Creek samples were characterized by a
variety of midge larvae and many empty mollusk shells but few live mollusks. Less than 50% of the
individuals were oligochaete worms. This may be a reflection of the sediment characteristics which
consisted of gravel and clay  with little of the muck substrate characteristic of the other two sites.
The data indicate that all three sites are effected by organic enrichment and/or agricultural runoff,
but the Black River macroinvertebrate community below the Coking Plant appears to be stressed by
something in addition to organic enrichment.

Key  Words: macroinvertebrates, bottom sediments, pollution, water quality, organic enrichment,
biotic index.
As   part  of  EMSL-Cincinnati's  Biomarkers
Research Program, fish and macroinvertebrates
were collected  from the  Black River,  Old
Woman Creek and Toussaint Creek during April
23-25,  1990.  The  project  objective is to
determine if biomarkers can be used to detect
a wide range of pollutants. The Black River site
was chosen as one  study site because of
polyaromatic hydrocarbons that  have  been
detected in the sediment downstream from a
coking plant and the resulting high incidence of
liver cancer in brown bullhead that inhabit the
site. Old Woman Creek and Toussaint Creek
were sampled as possible control sites.

The purpose of this paper  is to compare the
macroinvertebrate  populations collected from
the three water bodies and attempt to relate
these populations to pollutional stresses and/or
habitat characteristics.

Nine fyke fish nets were set at each site along
both east and west banks  over a distance  of
approximately one  mile. Each net was defined

Lewis and Smith
as a sampling  station for fish. Three of these
stations were randomly chosen  as benthic
macroinvertebrate collection stations on each
water body. Three six inch Ekman or Ponar grab
samples were collected at each station and
preserved  in 70% ethanol. The samples were
returned to  the  laboratory,  sorted  and the
organisms identified to the lowest  taxonomic
level possible. Pollution tolerance value for each
taxa  were  taken   from  EMSL-Cincinnati's
Biological  Methods Manual (USEPA 1990) or
Hilsenhoff (1987). Sediment and water samples
were also collected at each station for chemical

Stations  were analyzed  using  Hilsenhoff's
(1977) modification (HBI) of Chutter's (1972)
Imperial Biotic Index, Shannon-Weaver's mean
diversity (d), and equitability (e). Although the
HBI  was designed to give a measure of the
effects of organic  pollution on the macro-
invertebrate  communities  inhabiting stream
riffles, it should be  useful in analyzing grab
samples collected from soft river substrates if
care is used in interpreting the data. An HBI
score of < 1.75 would indicate excellent water
quality, 1.76 - 2.50 good water quality, 2.51 -
3.75 fair water quality, 3.76 - 4.00  poor water
quality,  and  >4.00 would  indicate grossly
polluted conditions. Mean diversity values of
less than  1.0  are  characteristic  of   gross
pollution, values between 1.0 and 3.0 indicate
fair to poor water quality, and values above 3.0
are common for clean water stations.

Equitability values above 0.5 are indicative of
good water quality, 0.3 • 0.5 fair,  and  values
below 0.3 indicate degradation of water quality.
Stations within each site and the three  water
bodies (between  sites)  were compared  using
the  Community  Loss   Index  and  Jaccard's
Coefficient of Similarity (Plafkin et a). 1989).
Trophic condition index (Howmiller and Scott
1977; Milbrink 1983) was also determined for
each station but the results were not helpful in
interpreting the data, possibly because most of
the oligochaete worms could not be identified
to species.
Station descriptions

Old Woman Creek - Station 1 was located 200
feet north of the RR bridge about 25 feet from
the west shore. Substrate was muck and well
rotted organic material; water depth was about
18 inches. Station 2 was located 100 feet west
of the observation deck about 50 feet from
shore.  Substrate was muck with some clay;
water depth was about 18 inches.  Station 3
was located 500 feet south of the Highway 2
bridge and about 150 feet west of a high gravel
bank  near the  east  shore of  the  estuary.
Substrate was muck and rotting leaves; water
depth was about 16 inches.

Toussaint Creek -  Station 1  was located 200
feet north of Highway 2 bridge about 50 feet
from the east shore. Water depth was about six
feet and the substrate was sandy  and silty.
Station 2 was located about one half mile south
of Highway 2 bridge  about 50 feet  from the
east shore just south  of a small  island. Water
depth was about two feet and substrate was
hard packed  gravel with some clay and silt.
Station  3 was located across the  bay from
Station  2 about 50 feet east of a very small
island. Water depth was about three feet and
substrate was muck,  clay and  hard packed
gravel. A large ditch entered the bay  about 50
feet north of the station.

Black River - Station 1 was located about four
miles upstream from  the  mouth of  the  river
about 30 feet from the west bank across from,
and about  100 feet upstream from, the upper
end of a large island. Water depth was about
seven feet and the sediment was mostly fine
silt. Station  2 was located  about  100 feet
downstream from the lower end of  the large
island  about 20  feet from the bank. Water
depth was about  two feet and  the  sediment
was muck, clay and  leaves.  Station  3 was
about one half mile downstream from a major
discharge from the coking plant at the bend of
the river about 20 feet from the west bank at
about river mile three. Water depth was 21/2
feet and the sediment was mud and silt:

                                         Comparison of Macroinvertebratesin Lake Erie Estuaries
Results and Discussion
Toussaint Creek. The  Ponar samples collected
at Station 1  (Table 1)  yielded 13 taxa, two of
which  (see below) are not generally found in
polluted waters and six  of  which are highly
tolerant of organic pollution.  Just over half
(56%)  of the  individuals were  oligochaete
worms which is an indication  of non-polluted
conditions (Goodnight  and Whitley 1960). The
HBI  of  3.76,  the diversity index  (2.3) and
equitability (0.5)  all indicate fair to marginal
water  quality.  The large number  of  empty
mollusk shells  (123 individuals representing 8
species)  indicate  that conditions  have been
present,  at  least some time  recently,  for a
diverse  population  of  these  organisms  to
develop.   The  midge  larvae  Ablabesmvia
mallochi.  which is  very sensitive  to  metals
contamination,   and  the   gastropod
Somatoovrus. which is not generally found in
organically polluted waters indicate that this
station  is  probably  not,  or  only slightly,
impaired by pollutants. Probably this is the least
impacted  station  sampled during this  study;
therefore it was used  as the reference station
for this study.

The three Ekman samples collected at Station 2
(Table  1)  contained  only  five  living  taxa,
consisting of the midge Polvpedilum scalaenum.
which  is  generally restricted to  unimpaired
waters,  two  pollution  tolerant  oligochaete
worms, and two facultative (wide range  of
tolerance) midge  species. Less than half the
individuals  (36%) collected  were oligochaete
worms which indicates non impaired conditions
(Goodnight and Whitley,  1960).  The HBI  of
3.33, the diversity index (1.8) and equitability
(1.0) all  indicate fair to  good water  quality
conditions at this  station. The presence of the
midge  Polvpedilum  scalaenum  also indicates
that toxic substances are probably not present
in concentrations high enough to effect the
biological integrity of  the  aquatic community.
The lack of a diverse fauna is probably due to
the difficulty of obtaining a sample because of
the hard packed gravel and clay substrate. The
Ekman sampler used was one with a handle so
that it could  be pushed with some force into
the bottom or the samples could not have been
taken here.

The three Ponar samples taken at Station  3
(Table 1) contained only seven living taxa but
14 taxa of empty mollusk shells. It seems odd
that so many shells were collected without any
live animals.  Perhaps something had  recently
occurred that killed all the mollusks,  but it  is
obvious that conditions in the recent past must
have been conducive to the establishment of a
diverse population. Only two taxa present (the
midge  Cryptochironomus  sp.   and  the
oligochaete worm Limnodrilus hoffmeisteri) are
tolerant of pollution, however, these two taxa
make  up  about  80%  of  the   individuals
collected. This station was very near a ditch
that enters the bay here and it is possible that
periodically toxic  and/or  organic  substances
may flow into the bay from this ditch causing
stress on the  benthic community. Our sampling
may have occurred during the beginning  of the
recovery period. The samples were collected  in
the spring soon after planting and storm  runoff
from nearby agricultural lands may have entered
the bay by way of this ditch. Krieger (1989)
reported that agricultural herbicides used with
corn and soybeans reach higher concentrations
in  rivers  of  northwest  Ohio than in  rivers
anywhere else in North America. The  Biotic
Index of 3.73 indicates only a slight impact as
do the diversity index (1.5) and equitability
(0.5).  The  high  percentage of  oligochaete
worms at this station as compared to the other
two may  be  an indication that whatever has
effected the  benthic  community was  not
limiting to the oligochaete worms and probably
was not widespread throughout the bay.

Community   Loss  Index   and   Jaccard's
Coefficient of Similarity indicate that the three
stations are quite dissimilar with the  greatest
difference between stations 1 and 2. If station
1  is considered the control, station 2  shows a
loss of 2.2 and station 3 shows a  loss of 1.4.
These differences would appear to be signifi-
cant and one might expect that stations 2 and

Lewis and Smith
Table 1. Macroinvertebrates Collected and Pollution Tolerance Values for Toussaint Creek. Intolerant
        taxa denoted by *.
Number of Individuals
Coelotanypus concinnus
Cryptochironomus sp.
Cryptochironomus fulvus gr.
Procladius nr. bellus
Chironomus plumosus gr.
Ablabesmyia mallochi*
Cladotanytarsus sp.
Polypedilum scalaenum*
Tanytarsus guerlus gr.
Other Diptera
Nr. Probezzia sp.
Lirceus lineatus
Urnatella gracilis
Somatogyrus sp."
Limnodrilus hoffmeisteri
L. maumeensis
Branchiura sowerbyi
Unidentified Oligochaeta
Total Individuals
Total Taxa
% Oligochaeta
Biotic Index (HB[)
Mean Diversity (d)
Equitability (e)
Station 1








Station 2



Pollution Tolerance
Station 3












 3  would   be  quite   similar  because  the
 Community Loss Index between them is small,
 but  Jaccard's Coefficient (<0.50) indicates
 otherwise. HBI scores for stations 1 and 2 and
 for 2 and 3 are statistically different from each
other but HBI scores for stations 1 and 3 are
similar.  Probably  most  of  the  differences
observed are ecological and  not caused  by

                                         Comparison of Macroinvertebratesin Lake Erie Estuaries
The data suggest that Toussaint Creek Bay is
not affected to any great extent by pollution,
however,  the  presence  of  many  species  of
empty mollusk shells at stations 1 and 3 leads
me to suspect that occasional instances may
occur, either natural or man caused, that stress
the aquatic community and temporarily affect
the biological integrity of the bay in the vicinity
of the canal that enters the bay on the west
shore. Runoff from  agricultural  lands  during
storms may be a factor here as in  most other
northwest Ohio rivers (Krieger 1989). A total of
17  taxa  consisting   of  200   individuals
(11.8/taxa)  were collected  in the nine  grab
samples taken from Toussaint Creek.

Old Woman Creek. The three Ekman samples
collected at Station  1  (Table 2) yielded nine
taxa,   including  the  bryozoan  Pectinatella
maanifica which  is  known to be  sensitive  to
toxic  contaminants and  five taxa which are
highly tolerant to organic  pollution. Of the 96
individuals collected 83 (86%) wereoligochaete
worms characteristic of organically  enriched
sediments. The HBI  of  4.01  and the  high
percentage of  oligochaete worms  (Goodnight
and Whitley 1960) indicate organic enrichment
but the diversity index  (2.3) and  equitability
(0.7)  indicate fair to good water quality. The
presence of the bryozoan Pectinatella maanifica
at this  station  would  indicate good  water

The three Ekman samples collected at Station 2
(Table 2) yielded ten taxa, none of which are
known to be sensitive to pollution and seven
which are highly tolerant of organic pollution.
Of the 132 individuals collected, 107  (81%)
were  oligochaete worms  characteristic  of
organically enriched sediment. The blood worm
Chironomus Dlumosus. which is highly tolerant
of sediments with high oxygen demand, was
also common at this station. The HBI of 4.03
and the high percentage of oligochaete worms
indicate organic enrichment but the  diversity
index  (1.6) and equitability (0.4) would indicate
fair water quality.
The three Ekman samples collected at Station 3
(Table 2) contained seven taxa, including the
amphipod Gammarus pseudolimnaeus which is
generally not found in waters containing toxic
substances other than organic enrichment and
four  of which  are  highly tolerant to organic
pollution. Of the 31 individuals collected, 23
(74%) were oligochaete  worms characteristic
of organically enriched sediments. The HBI of
3.84   and   the  presence   of  Gammarus
pseudolimnaeus  would  indicate  that  toxic
substances  are probably not  major limiting
factors at  this station. The  percentage of
oligochaete worms was less at this station than
the other  two and  fell in the  moderately
polluted  category of  Goodnight  and Whitley
(1960).  The  diversity  index   (1.8)   and
equitability (0.6) indicate fair to good  water

Community  Loss  Index   and   Jaccard's
Coefficient of Similarity show that stations 1
and 2 are quite similar and that station  3  is
significantly  different  from  the  other  two
stations. The Community Loss Index values also
indicate  that  station  3 has  fewer taxa  in
common with either of the other two stations.
HBI score for station 3 was also lower than for
the other two stations, but the difference  was
not significant (P = 0.051. However, HBI for Old
Woman  Creek  stations  1   and  2  were
significantly  higher  than  for  the  reference
station while station  3 was not significantly
different (P = 0.05).

The  data suggest that Old  Woman  Creek
Estuary may have been temporarily affected by
storm runoff from agricultural lands (Klarer and
Millie,  1989)  and  organic  enrichment  from
decaying  vegetation  and  that station  3  is
slightly less effected than the other two.  The
limiting factor at this site would most likely be
the highly enriched substrate  consisting of
muck  and   decaying  vegetation  probably
accompanied by periods of low DO and  high
levels  of nitrites resulting  from storm runoff
from nearby agricultural lands (Klarer and Millie,
1989). Using Goodnight and Whitley's (1960)

Lewis and Smith

Table 2. Macroinvertebrates Collected and Pollution Tolerance Values for Old Woman
Number of Individuals
Taxa Station 1 Station 2
Coelotanypus concinnus
Cryptochironomus fulvus
Procladius nr bellus
Chironomus plumosus gr
Oicrotendipes sp.
Other Diptera
Nr Probezzia sp.
Dubiraphia sp
Donacia sp.

gr 1





Gammarus pseudolimnaeus*
Limnodrilus maumeensis
L. hoffmeisteri
L. cervix
Branchiura sowerbyi
llyodrilus templetoni
Unidentified Oligochaeta
Pectinatella magnifies*
Total Individuals
Total Taxa
% Oligochaeta
Biotic Index (HBi)
Mean Diversity (d)
Equitability (e)
metric based on percent
present, all three stations



oligochaete worms
would be considered
polluted. A total of 15 taxa consisting of 259


Pollution Tolerance
Station 3 Value






or organically polluted waters.








Two midge
Dicrotendioes neomodestus and
curtilamellata. present
at this station
individuals  (17.3/taxa) were collected in  the
nine grabs taken from Old Woman Creek.

Black River. The three Ponar samples collected
at Station 1 (Table 3) contained 18 taxa most
of which are characteristic of slightly impaired
are usually not  found  under  contaminated
conditions. Of the 137 individuals collected 106
(77%) were oligochaete worms which would
indicate  moderately  polluted  or organically
enriched sediments.  The  HBI of  3.92, the
diversity index (2.4) and equitability  (0.4) all

Comparison of Macroinvertebrates in Lake Erie Estuaries

Table 3. Macroinvertebrates Collected and
Pollution Tolerance Values for Black River.
Number of Individuals
Taxa Station 1 Station 2
Harnischia curtilamellata* 1
Ablabesmyia mallochi*
Tanytarsus guerlus gr
Phaenopsectra prob dyari
Eukiefferiella claripennis
Hydrobaenus pilipes gr*
Cryptochironomus f ulvus gr 7
Cryptochironomus sp 1
Diplocladius cultriger 1
Dicrotendipes neomodestus 1
Polypedilum ophioides 1
Glyptotendipes lobiferus 1
Orthocladius sp.
Cricotopus tremulus gr
Corichapelopia sp.
Procladius nr. bellus
Nanocladius distinctus
Other Diptera
Hemerodromia sp 2
Chaoborus punctipennis 1 0
Nr. Probezzia sp. 3
Ephydridae 1
Unidentified Diptera
Stenelmis sp. *
Dubiraphia sp.
Scirtes sp
Berosus sp
Caenis sp 2
Bactra sp (?)
Argia apicalis














Pollution Tolerance
Station 3 Value

11 3
1 4
1 3
1 3
1 3
9 3

1 3








Lewis and Smith

Table 3. Macroinvertebrates Collected

Asellus communis
Pisidium casertanum
Physella vinosa
Lophopodella carted*
Limnodrilus hoffmeisteri
L. udekemianus
L. cervix
Potamothrix vejdovskyi*
Tubifex tubifex
llyodrilus templetoni
Quistadrilus multisetosus
Oero sp.
Nais elinguis
Nais communis
Unidentified Oligochaeta
Total Individuals
Total Taxa
% Oligochaeta
Biotic Index
Mean Diversity (d)
Equitability (e)

Station 1




and Pollution Tolerance

Number of Individuals

Station 2






Values for Black River.

Station 3












* Intolerant taxa

                                         Comparison of Macroin vertebrates in Lake Erie Estuaries
indicate fair to poor water quality at this station
which  is  located  upstream  from  the  major
effluent from the coking  plant. The  mayfly
Caenis sp., present at this station, is the most
tolerant of the mayflies and is not a good water
quality indicator.

The three Ponar samples collected at Station 2
(Table 3) yielded a surprising 35 taxa, many of
which would not be expected in mud substrate
where tittle current is present. Nine of the taxa
are tolerant  of pollution,  one  (Lophopodella
carter!) is very sensitive to pollution (except in
the statoblast stage) and seven others are not
characteristically  found  in  impaired waters.
Almost  exactly 60%  of  the individuals are
oligochaete worms characteristic of organically
polluted conditions which would indicate some
organic enrichment,  possibly due to decaying
vegetation. It is not likely that toxic pollution  is
present in the sediment at this station because
Potamothrix   vejdovskvi.   an   intolerant
oligochaete,  was  among  the  diverse  worm
fauna present. The diversity of midge and other
insect groups would indicate that conditions are
conducive to the  development of a balanced
benthic community. This may be due, however,
to the  possibility that these  organisms are
drifting into this station from some stream that
may enter the river  behind  the island  just
upstream or  from a spring entering the river
from under the stream bank.  The HBI of 3.62,
diversity index (3.7) and  equitability (0.5) all
indicate fair to good water quality.

The three Ponar samples collected at Station 3
(Table 3) contained 13 taxa, eight of which are
characteristic of polluted conditions. The other
five  species  are  all  facultative  and could be
present in moderately  polluted  waters.  Of the
229 individuals  present,  202 (88%)  were
oligochaete worms indicating grossly polluted
waters. It is interesting that the only deformed
Procladius midge (Warwick, 1989) found during
this study was collected from this station which
is located one half mile downstream of the main
effluent from the coking plant. The HBI of 4.00
indicates  poor  water  quality  as does  the
equitability (0.3) while the diversity index (1.5)
is borderline between fair and poor conditions.

Community   Loss  Index   and   Jaccard's
Coefficient of Similarity  show  that the  three
stations  are  considerably different from one
another. Both stations 1 and 3 show significant
community loss when compared with station 2.
The HBI scores for the three stations are similar
(P  =  0.05), however  the  HBI  scores  for
stations 1  and 3 are both significantly higher
than for the control reference station.

The data suggest that the benthic community
at Station 1, located upstream from most of the
effects of the coking plant, does show  some
stress on the biota. Station 2 samples contain
intolerant organisms (including  riffle  beetles)
which   are  not  characteristic  of   muddy
substrates and some tolerant forms  that are
characteristic of organic pollution.  Station 3 is
noticeably degraded as compared to the  other
stations sampled during this  study. A total of
47 taxa consisting of 530 individuals (7.2/taxa)
were collected in the nine grab samples  taken
from the Black River.

Summary and Conclusions
The  Community Loss  Index  and Jaccard's
Coefficient of Similarity show that Old Woman
Creek and Toussaint Creek macroinvertebrate
communities are more  similar  to each  other
than either one is to the  Black River but  these
similarities are not  very great.  There  is no
significant  community   loss  between  Old
Woman Creek or Toussaint Creek and the Black
River when  composite  data are  compared.
Based on the Community Loss  Index, it would
appear that Old Woman Creek and Toussaint
Creek are both slightly less polluted  than the
Black River, but this is likely due to the diverse
fauna collected from Black River station 2 as
discussed above.

Toussaint  Creek station  1 is considered the
best   control  station   because   of  its
representative  substrate  and  overall quality
based on the combination of metrics used in

Lewis and Smith
this  analysis.  Using  this  as the  reference
station, Community Loss, Jaccard's Coefficient
of  Similarity  and  t-values  based  on  a
comparison of HBI scores for the other stations
sampled are as follows:
Stations     Loss   Jaccard's  HBI
Compared   Index   Coeff. t-values
Toussaint2    2.2    0.13   2.597
Toussaint 3    1.4    0.19   0.040     ns
Old Woman  1  0.3    0.44   4.590
Old Woman  2  0.6    0.47   4.531
Old Woman  3  1.6    0.13   0.241     ns
Black River!  0.6    0.11   2.531
Black River 2  0.2    0.12   0.146     ns
Black River 3  0.7    0.18   3.983
Old Woman  Creek
 Composite    0.4    0.37   3.073     *
Black River
 Composite    0.1    0.10   1.108     ns

•Significant at p < 0.05, df = 4.

These metrics indicate  that Toussaint  Creek
stations 2 and 3, Old Woman Creek Stations 1,
2, and 3 and Black River stations 1  and 3 are
different from  the reference  station. The
differences  between  the  Toussaint   Creek
stations  can   be  explained  by  substrate
differences but the others could be related to
pollution, including agricultural runoff (Krieger
1989). As  might  be  expected  Black River
station 2 did not show a community loss (see
discussion  of  the individual  stations above).
Based on Jaccard's Coefficient all the stations
are significantly different from the reference
station and all but Old Woman Creek stations 1
and 2 are vastly different. All of the HBI scores
differ significantly (P = 0.05) from the reference
station except Toussaint Creek station  3, Old
Woman Creek station 3, Black River station 2,
and the Black River composite. The reason the
Black River composite HBI scores did not differ
from  the reference station is probably because
of the high variability due to station 2 Black
River  samples.   As  mentioned  above,  the
dissimilarities  between  the reference  station
and stations 2 and 3 at Toussaint  Creek and
Black River station 2 may be attributable to
environmental and/or  physical  factors. The
other dissimilarities could be due to organic or
toxic stresses. The Community Loss Index and
Jaccard's Coefficient scores for the composite
data indicate that Old  Woman Creek is more
like  the  reference station  than  is the Black
River. The low Community Loss Index score
and  the  low t-value for the composite Black
River samples as  compared to the reference
station are  mostly due to the diverse fauna
collected at station 2 as discussed  above.

Because   the  sediment  samples  that were
collected for chemical characterization have not
yet been analyzed, it is impossible to reach any
real  definitive  conclusions  based  on  the
macroinvertebrate collections alone. However,
the benthic macroinvertebrate grab collections
seem to  indicate  that all  three Old  Woman
Creek stations were organically enriched, with
oligochaete worms making up over 60% of the
individuals and the remaining taxa characteristic
of  waters  with  high oxygen  demand.  In
Toussaint Creek only Station 3 located near a
drainage ditch showed signs of stress. That
may be  due  to  periodic discharge of toxic
and/or organic pollutants into the bay from this
ditch,  possibly  during storms.   Oligochaete
worms made up 60% or more of the individuals
collected from the  Black  River at all three
Stations indicating organic enrichment  or toxic
substances in  the sediment (Krieger  1990).
However, a few sensitive taxa were found at
Stations 1  and  2 indicating  reasonably good
water quality at these two stations. Station 3
samples contained about 90% worms and no
sensitive taxa indicating a stressed community
of benthic macroinvertebrates.

The complete absence of Hexagenia mayflies,
the   limited  number   of   chironomids  and
gastropods, and the increase in oligochaetes in
Lake Erie  bays  has been correlated with
increased pollution in the lake (Edwards 1990,
Reynoldson et al. 1989). Because no Hexagenia
mayflies and only a few live gastropods were
collected at any of the sites sampled during this

                                        Comparison of Macroinvertebratesin Lake Erie Estuaries
study, it might be reasonable to assume that all
of the sites were polluted.

Neither Old Woman Creek or Toussaint Creek
appear to be good control sites. Grab samples
collected from the Grand and Chagrin Rivers
will be analyzed and the data compared  with
the Black and Cuyahoga Rivers to determine if
either of them might be better control sites for
the benthic phase of this study.

We would  like  to thank Larry Linns,  Eastern
District Office, U.S. EPA Region 5 for providing
a boat  and assisting  with  collection of the
samples. This  work was a  part  of Susan
Cormier  and  Tim   Neiheisel's   Biomarker
Research Project.

Literature Cited
Chutter, P.M. 1972. An empirical biotic index
of the quality of water in South Africa streams
and rivers. Water Research 6:19-30.

Edwards, C.J. 1990. Biological surrogates of
mesotrophic ecosystem health in the Laurentian
Great Lakes. Great Lakes  Science Advisory
Board, Windsor, Ontario.

Hilsenhoff,  W.L. 1977. Use of arthropods to
evaluate water  quality of streams. Technical
Bulletin 100, Department of Natural Resources,
Madison, Wl.

Hilsenhoff,  W.  L.  1987.  An improved  biotic
index of organic stream pollution. Great Lakes
Entomologist. 20:31-39.

Howmiller,  R.P. and M.A.  Scott.  1977. An
environmental   index  based   on  relative
abundance  of oligochaete species.  Journal of
the   Water  Pollution  Control    Federation

Goodnight,  C.J.  and  L.S.  Whitley. 1960.
Oligochaetes   as  indicators  of  pollution.
Proceedings  of  the  15th  Industrial  Waste
Conference, Purdue University Engineering Bull.

Klarer, D.M. and D.F. Millie. 1989. Amelioration
of storm-water quality by a freshwater estuary.
Archives of Hydrobiology  116(3):375-389.

Krieger, K.A. 1989.  Chemical limnology and
contaminants. Pages  149-175 in K.A. Krieger
editor).  Lake Erie estuarine  systems: Issues,
resources, status and management. Estuary of
the Month  Seminar  Series  No.  14, NOAA
Estuarine Programs Office.

Krieger, K.A.  1990. Assessing  lake quality
improvement   using  trends   in   benthic
macroinvertebrate communities: A case study
in Lake Erie.  Presented at the 1990  Midwest
Pollution Control Biologists Meeting,  Chicago,

Milbrink, G.  1983. An improved  index on the
relative abundance  of  oligochaete  species.
Hydrobiologia 102:89-97.

Plafkin, J.L.; M.T. Barbour:  K.D. Porter,  S.K.
Gross   and   R.M.   Hughes.  1989.  Rapid
bioassessment protocols for use in streams and
rivers: Benthic macroinvertebrates and  fish.
U.S. Environmental Protection Agency, Office
of  Water   Regulations  and   Standards,
Washington, D.C. 20460. EPA/440/4-89/001.

Reynoldson, T.B.;  D.W.  Schloesser and  B.A.
Manny.  1989.  Development of  a  benthic
invertebrate objective for mesotrophic  Great
Lakes waters. Journal of Great Lakes Research

USEPA.  1990.  Macroinvertebrate field and
laboratory methods for evaluating the biological
integrity of surface waters. Klemm, D.J., P.A.
Lewis,  F.  Fulk,  and J.M.  Lazorchak.  U.S.
Environmental  Protection  Agency,
Environmental Monitoring System Laboratory,
Office  of   Research   and  Development,
Cincinnati, OH 45268. EPA/600/4090/030

Lewis and Smith
Warwick, W.F. 1989. Morphological deformities
in   larvae  of  Procladius   Skuse  (Diptera:
Chironomidae) and their biomonitoring potential.
Canadian  Journal  of  Fisheries and Aquatic
Sciences 46{7):1255-1271,

     A Comparison of the Results of a Volunteer Stream Quality Monitoring
                   Program and the Ohio EPA's Biological Indices

Mark A. Dilley"
School of Natural Resources
The Ohio State University
2021 Coffey Road
Columbus, OH 43210

Volunteer  stream  quality monitoring  is  increasing in  popularity  around  the  country,  and
organizations involved with the administration of  volunteer stream quality monitoring  programs
are becoming interested  in  the  effectiveness of their  monitoring techniques. This  research
compares  the results of the  Ohio Department of Natural Resources (ODNR)  volunteer-oriented
Scenic  Rivers Stream  Quality Monitoring  Program and the  Ohio  Environmental  Protection
Agency's  (OEPA)  biological  assessments. The volunteer biological monitoring ("kick-seining")
technique was performed on 12 Ohio rivers and tributaries, at 47 different sites, to coincide with
the OEPA's  monitoring agenda for the summer of 1989. Comparisons were made  between the
volunteer stream quality monitoring ratings  and the OEPA's Index of Biotic  Integrity  (IB)) and
Invertebrate Community Index (ICI). Sites which were rated "excellent" using the ODNR volunteer
method tended to meet the OEPA's criteria for attainment of aquatic life uses for both the IBI and
ICI. Sites which  were determined  to be "fair" or "poor"  with the volunteer method corresponded
to IBI and  ICI scores falling in the  non-attainment of aquatic life uses range. Although revisions in
the sampling and rating system for the volunteer  program could improve the predictive value  of
these results as compared to OEPA's indices, the volunteer  technique  assessments  currently
appear to have merit when interpreted in terms of aquatic life use attainment or  non-attainment.

Key  Words: volunteer monitoring, biological indices, stream quality, kick-seining,  Ohio, Scenic
In 1983, the  Ohio Department of  Natural
Resources (ODNR) developed the Ohio Scenic
Rivers Stream Quality Monitoring  Program
with assistance from the Ohio  Environmental
Protection Agency (OEPA). This program uses
volunteers to  conduct  simple  stream quality
assessments   at  designated   monitoring
stations  on the  state's ten  Scenic Rivers.
ODNR's stream quality monitoring technique
involves assessments based on the presence
or absence of 20 taxa of macroinvertebrates
which  are  divided  into   three  categories,
according to  each groups pollution tolerance
level (Fig. 1).
Group One, the pollution intolerant organisms,
includes   mayfly   and   stonefly   nymphs,
dobsonfly, caddisfly and water penny beetle
larvae, riffle beetles, and gill-breathing snails.
Group  Two  macroinvertebrates,  with  inter-
mediate pollution tolerances, include dragon-
fly and damselfly nymphs, beetle and cranefly
larvae, scuds, crayfish, sowbugs, and clams.
Group Three, the pollution tolerant organisms,
consists  of  aquatic  worms,  pouch  snails,
black fly and midge larvae,  and  leeches.
Many of the taxa used in the program encom-
pass  entire  orders  (i.e. mayflies  - Order
Ephemeroptera,  caddisflies  - Order  Trichop-
tera) so identification is not refined.
      Current address: Metcalf & Eddy, Inc., 2800 Corporate Exchange Drive,
      Suite 250, Columbus, Ohio 43231

                 GROUP 1  IThe&e oigonum* HAS. gznvuiUy poLtu£u>n-±itCotuian£.   TheAJi. dominance
                           gtnvuULLu Ug>u.fc
                                                  Volunteer Monitoring Program Comparison
In this  Stream  Quality Monitoring Program,
volunteers  collect  macroinvertebrates from
riffles   using   the   "kick-seine"  technique.
Riffles, with little or no vegetation and stones
up to 15 inches in  diameter, are the  type of
habitat best suited to this method of sampling
(Frost   et   al.   1971).   The   "kick-seine"
technique  involves  disturbing  the substrate
upstream   of   the  seine  to  dislodge   the
macroinvertebrates  which cling  to, and  hide
under the rocks and debris in the  riffle. Once
freed of the substrate, the macroinvertebrates
are carried  by the current into the seine. After
a  sample  has  been collected,  the seine is
taken  to  the  stream  bank   where   the
organisms  are hand-picked from the net and
identified  on site.  Macroinvertebrates often
exhibit patchy distributions in streams (Rabeni
and   Minshall   1977,   Schwenneker  and
Hellenthal  1984). Therefore,  volunteers  are
encouraged to take  samples from a variety of
habitats until they feel that no new taxa are
represented in  their samples. No set  number
of samples has ever been established for the
program, however.

After all the samples have been collected,
volunteers  fill   out  an   assessment  form
indicating  the  station sampled  (according  to
OEPA river miles),  water  conditions such  as
clarity, algal bedgrowths,  and odor, and  the
macroinvertebrate groups  found (Fig.  2).  For
each macroinvertebrate taxon  group  located
at the station, an estimated count letter code
is entered on the assessment form. The letter
codes A, B, and C represent 1 to 9, 10 to 99,
and  100  or  more  individuals,  respectively.
Using  estimated counts  allows  the  ODNR
staff to get an  idea of population  sizes while
not  placing  the burden  of  counting   the
organisms on the program volunteers.

The final assessment score, referred to as the
Cumulative Index Value, or CIV, is based only
on the diversity of  macroinvertebrates found
and not the quantity.  In the scoring system,
each Group One taxon in the sample receives
a point value of three, each Group Two taxon.
a point value  of two, and each Group Three
taxon, a point value  of  one.  The CIV is the
sum of the points given to each category. The
final step in the stream quality assessment is
determining a  qualitative  rating for the station
based on the CIV.  Cumulative Index Values of
over  22  are  given  an "excellent"  rating.
Scores between 17 and  22  are rated "good."
"Fair"  is 11-16, and  a "poor" rating  is given
to scores of less than 11.

Once  completed,  assessments  are  sent  to
ODNR  where  they  are  entered   into   a
computer database. The database allows for a
quick  review  of  the   history  of  a  given
monitoring station  to  determine if the site has
experienced any significant  impacts over the
years it has been monitored. It is believed that
this   may  allow   for   early  detection  of
degradation on the Scenic  Rivers. The Ohio
Department of  Natural  Resources  stresses
that  the   procedure   "is  not  intended  to
pinpoint subtle changes  in water quality, but
rather the general  condition  of the river," and
that "information  which indicates  potential
decreases in water quality will be coordinated
with   the  Ohio  Environmental   Protection
Agency."  (ODNR  n.d.).  The program  is not
intended to completely assess the source or
degree of degradation, but  rather provide  an
inexpensive and enjoyable way for the public
to   flag   the  attention   of   responsible
enforcement  agencies  in  the  event that
further study may be warranted.

The   simplicity  and  accessibility   of  the
program has made it  popular among  schools,
conservation  groups,  scouts,  and  families.
Since its beginning, the Stream Quality
Monitoring Program has  grown quite  rapidly.
During  the   1990  monitoring   season,
approximately  3,000 volunteers  monitored
Ohio's  Scenic  Rivers.  In  addition  to this
considerable volunteer force,  many Soil and
Water Conservation  Districts in  Ohio  have
expressed  interest  in   ODNR's  method  to
develop volunteer  stream quality monitoring
programs for streams  within their counties




                                                                            SAMPLE i
                                                                     NO. OF PARTICIPANTS
               SURFACE SCUM. ETC.)
                                                   HACH KIT RESULTS  (If used) AND
                                                   OTHER OBSERVATIONS
                                                     USE BACK OF FORM  IF NECESSARY
               WIDTH OF RIFFLE

               WATER DEPTH
               WATER TEMP.  (°F)
                        BED COMPOSITION OF RIFFLE (X)

                        SILT L_I         SAND LJ
                                        COBBLES (2"- 10")
          GRAVEL (V- 2"

BOULDERS (> 10") f~]
                                                ESTIMATED COUNT
                                                LETTER CODE
                   B ^0 to 99
                   c > 100 or more
                       NUMBER OF TAXA
                        INDEX VALUE 3
                                   NUMBER OF TAXA
                                    INDEX VALUE 2
              NUMBER OF TAXA
               INDEX VALUE 1

                         INDEX VALUE
                                  STREAM  QUALITY  ASSESSMENT

                                  EXCELLENT (> 22) O       SOOD (17-Z2) f""l

                                  FAIR (11-16)     Q       POOR « 11)  O
                PLEASE SEND THIS FORM TO:
                                             Mr. John S. Kopec, Planning Supervisor
                                             Division of Natural Areas and Preserves
                                             Ohio Scenic Rivers Program
                                             1889 Fountain Square Court
                                             Columbus. Ohio  4322*      Phone: (614) 265-6458
Figure 2. Ohio Department of Natural Resources Scenic Rivers Stream Quality Monitoring Program
assessment form.

                                                 Volunteer Monitoring Program Comparison
(Kopec   1989).   Other   states  and  private
organizations  are also  patterning  programs
after ODNR's technique.

Generally,  identification  of  invertebrates to
only  the  order  level  of  classification is
considered   to   have   limited   ecological
meaning.   Species  level   identification  is
necessary for a  more sensitive measure of
water quality. (Resh and Unzicker 1975). This
fact, and the increasing interest in ODNR's
Stream  Quality  Monitoring  Program,  caused
OEPA and  ODNR staff  to  question  quality
assurance and quality control for the program.

To  examine the accuracy of ODNR's stream
quality  monitoring  technique,  a  source of
reliable stream health information was needed
for  comparison.  James  Karr  (1981) stated
that it would be impossible, because of the
complexity  of  stream ecosystems,  to  ever
recognize all the potential factors that may
impact biological communities. Although no
techniques exist which can fully acknowledge
all  the  processes  at  work  in an  aquatic
ecosystem,  biological  monitoring  integrates
the effects  of many processes that occur in
streams.  To assess stream health, the OEPA
uses  biological   indices  which have  been
closely  studied  and  tested,  making  the
OEPA's  methods  the best available source of
stream   health  and   biological   integrity
information in Ohio.

The OEPA monitors rivers and streams using.
three  primary   indices   as  criteria  for
assessment. The  Index of Biotic Integrity, or
IBI, originated by Karr (1981), is based on fish
populations. Invertebrate  samples are used to
compile the  Invertebrate Community Index, or
ICI. The third index, the  Index of Well-being,
or  Iwb,  was not examined  in this study.
These indices are used  to  rate the  relative
quality of  Ohio's rivers and are translated  into
ratings of "exceptional, good,  fair, poor,  and
very poor."  The reason for  the use of more
than  one   organism   group  (fish   and
invertebrates)  is  explained  in the   OEPA
publication   Biological   Criteria   for  the
Protection of Aquatic  Life:  Volume  I.  which
states "The  need  to  use  both  groups  is
apparent   in   the   ecological   differences
between them,  differences  that  tend  to be
complementary   in   an  environmental
evaluation" (Ohio EPA 1988).

In   order   to  address   the   quality
assurance/quality control  issue  for ODNR's
Stream   Quality  Monitoring  Program,  this
research examined  the correlation  between
the OEPA's indices  (IBI and  ICI) and ODNR's
CIV  and also the agreement between  ODNR
staff- and volunteer-generated stream  quality
assessments. The  general  objective  of this
paper  is  to  illustrate  how  accurately the
results  of  ODNR's   simple   approach  to
biological monitoring can reflect stream health
assessments  based on  more sophisticated

Methods and Materials
Over the summer of 1989 (late June to mid-
September),  the   standard   ODNR  stream
quality  monitoring  technique  (as  described
above) was performed  on  12 of Ohio's rivers
and   tributaries  which   were   also   being
monitored by the OEPA (Fig. 3). The sites on
these rivers represented a variety of habitat
and  impact types.  With  the help  of the
OEPA's  staff,   a   sampling  schedule  was
arranged  which  closely  adhered   to  their
agenda. This was  done to  help reduce the
effects of seasonal  or temporary variations in
stream  quality. All  ODNR  stream   quality
assessments were made within 0.8 river miles
of the area sampled  by the OEPA and all
ODNR assessments were made  within two
weeks of OEPA's testing.

A 9  in. high,  18 in. wide rectangular frame
1732 in. mesh dip net was substituted for the
seine to allow for solo collections.  This type
of net and the  standard 1/16  in. mesh seine
are used interchangeably in  ODNR's program
to   allow  stream   quality  monitoring
coordinators to make collections alone. At

Figure 3. Ohio rivers monitored in study.
each site sampled, four regular samples were
collected  from  areas  approximately  9  ft.
square and a search was conducted along the
stream's edge for macroinvertebrates such as
dragonfly  naiads,  which may  prefer slower
water velocity or vegetation. An index  value
was calculated for each sample and  a CIV
was calculated for the riffle. The CIV was
then  translated  into   a  qualitative   rating.
Assessments were made on 37 different sites
for comparison  with  the IBI,  and many of
those sites were monitored twice, resulting in
56 assessment records. For the comparison
with the ICI, 30 assessment records from 30
sites were collected.  The data were entered
into a FoxBase Mac database.

In the  spring of  1990, the  OEPA finished
processing all of  its  1989 data,  and their
assessments were merged into  a  master
database.   The   study  sites   were  then
examined for correlations   between ODNR's
stream  quality  monitoring  results  and  the
indices of OEPA, the IBI and ICI.
To  examine  volunteer  accuracy,  ODNR's
volunteer  monitoring database was searched
for sites which were monitored both by staff
members and volunteers within a three month
period of time.  Matched records in which one
sample  was taken in the  early months of
spring  and  the other in  the summer  were
discarded,  due  to the notable  changes in
benthic   community  composition  between
these time periods. Spring CIVs are typically
higher  and  are usually  not comparable to
summer  CIVs.  Over  200  usable matched
records  were located in the database.

Results and  Discussion
Comparison  of the IBI and ODNR's CIV
For  sites  rated   "excellent"   by  ODNR's
method, corresponding IBIs ranged from 30 to
57 (Fig. 4). This range includes IBIs which the
OEPA would consider indicative  of  "fair" to
"exceptional" conditions. The CIV ratings did
not  match  exactly  those  of   the  OEPA.
However, 86% of the corresponding  IBIs did
fall at  a  value of 40 or  above,  indicating
attainment of aquatic life uses, as designated
by  the  OEPA. For  sites  rated   "good,"  the
corresponding   IBIs  again   showed  a  wide
range,   with approximately  half  the  values
indicating that sites did  attain  aquatic  life
uses, while  the other half indicated that sites
did not attain life  uses. All "fair" and "poor"
ratings  were  observed at  sites where  IBIs
were less than 40, indicating non-attainment
of life uses.

A  primary  reason   for  lack  of  complete
agreement  between   the  CIV  and   IBI
qualitative ratings is the inherent differences
between  the  indices. The  IBI   is an  index
based  on fish collected from  a  200 meter
reach of stream and ODNR's CIV is based on
macroinvertebrates sampled from a riffle only.
However, another factor, drainage area, was
found  to affect the correlation. The OEPA
designates  sampling  sites as   headwater,
wading, and boat sites, based on the drainage
area. When  the boat sites were eliminated

                                                  Volunteer Monitoring Program Comparison
Good           Fair
 Figure 4. Notched box plots of Cumulative  Index Value (CIV) qualitative ratings versus Index of
 Biotic Integrity (IBI) scores, 25th and 75th percentiles, IBI range, and IBI outliers (>2 interquartile
 ranges from median). IBI qualitative ratings  (exceptional, good, fair,  poor, and very poor)  appear
 on  the  right vertical  axis. Shading indicates  approximate boundaries between ratings and the
 variability of the index.
from  the  comparison,  the  definition between
ODNR's   qualitative  ratings   and   the
corresponding  IBIs  increased  (Fig. 5).  The
median IBI  score  for  each  corresponding
ODNR rating  fell in the  correct  qualitative
range for  the IBI, and the IBI ranges for the
"excellent"   and   "good"   ratings   were
shortened and more defined. The IBI range for
sites  rated  "good"  was  still   considerably
large, but  it  was  centered in the correct IBI
qualitative range.  For "fair" sites, all IBIs fell
in the non-attainment  range of less than 40.
A box plot of those  sites with drainage areas
greater than  200  square  miles  further
illustrates  the impact of drainage area. (Fig.
6). Notice that, for sites of larger  drainage,
there  is no detectable definition between sites
rated   "excellent"   and  "good"  and   the
corresponding IBIs. The IBI  ranges for the CIV
ratings are notably similar.
          Comparison of the ICI and ODNR's CIV
          For   sites  rated  "excellent"   by  ODNR's
          method, ICIs  ranged from 41 to  57 (Fig. 7).
          This  range includes  ICIs  which  the  OEPA
          would consider  "good" to "exceptional."  As
          in the IBI comparison,  the CIV ratings did not
          exactly match the ICI ratings of the  OEPA.
          However, all of  the ICIs corresponding  to the
          "excellent" rating did fall at a value of 35 or
          above,  indicating attainment of  aquatic  life
          uses, as designated by the OEPA for the  ICI.
          For sites rated "good," the corresponding ICIs
          showed  a wide  range,  with approximately
          62% of the values indicating attainment of
          aquatic life uses, while the other 38%  of the
          values  indicated non-attainment.  ICI  values
          were less than 35 at sites where "fair"  ODNR
          results   were  observed,   indicating   non-
          attainment of life  uses. No  "poor" sites for

Good           Fair
Figure  5. Notched box plots of Cumulative Index Value (CIV) qualitative ratings versus Index of
Biotic Integrity (IBI) scores, 25th and 75th percentiles, IBI range, and IBI outliers (>2 interquartile
ranges from median)  for sites  with drainage  area  <.  200  sq. mi.  IBI  qualitative  ratings
(exceptional, good, fair, poor, and very poor) appear on the right vertical axis. Shading indicates
approximate boundaries between ratings and the variability of the index.
 Good           Fair
Figure 6. Notched box plots of Cumulative Index Value (CIV) qualitative ratings versus Index of
Biotic Integrity (IBI) scores, 25th and 75th percentiles, IBI range, and IBI outliers (>2 interquartile
ranges  from median)  for sites  with  drainage  area >  200 sq. mi.  IBI  qualitative ratings
(exceptional, good, fair, poor,  and very poor) appear on the right vertical axis. Shading indicates
approximate boundaries between ratings and the variability of the index.

                                                 Volunteer Monitoring Program Comparison

Figure 7.  Notched  box  plots  of  Cumulative  Index  Value  (CIV)  qualitative  ratings versus
Invertebrate Community Index (ICI) scores, 25th and 75th percentiles, ICI range, and ICI outliers
(>2 interquartile ranges from median). ICI qualitative ratings (exceptional, good, fair, poor, and
very poor)  appear on the  right vertical axis. Shading indicates approximate boundaries  between
ratings and the variability of the index.
comparison  with the ICI were present in the
data  set.  Overall,  there  was   a  closer
correlation (the ICI ranges for the CIV ratings
were more  defined)  between ODNR's  CIV
ratings and  the ICI than ODNR's ratings and
the   IBI.   The   "good"   CIV  rating   still
encompassed a large range of Ids, however,
and  the  actual CIV and ICI  ratings did not
always match.

The  differences between  ODNR and  OEPA
macroinvertebrate  assessments may be due,
in part, to the fact that the OEPA retains its
collections for microscopic investigation and
they are  better able  to locate  and identify
small early instar forms of these organisms.
Another factor  is that  the  OEPA researchers
always make an attempt to sample a riffle, a
         run, and  a  pool  area  when  performing  their
         qualitative collection  procedure. This  could
         result  in  a  higher  diversity  of  organisms in
         their   samples  as  compared  to   ODNR's
         samples,  which  are taken  only  from  riffle
         areas.  In  addition, both the  IBI  and ICI
         incorporate  a correction factor  to adjust for
         drainage   area  impacts,   while  ODNR's
         technique  does   not.   For   the   ICI/CIV
         comparison,  drainage area impacts were not
         found to noticeably affect the correlation.

         Comparison   of  Volunteer  and   Staff
         Volunteer ratings  tended to be higher  than
         assessments made by ODNR staff members
         (Fig. 8). For sites rated  "excellent" by  staff,
         approximately 80% of volunteer CIVs fell in








                      Excellent         Good            Fair
                                STAFF-COLLECTED CIV RATING
Figure 8. Notched box plots of OONR staff ratings versus volunteer-generated Cumulative Index
Values (CIVs), 25th and 75th percentiles, CIV range, and CIV outliers (>2 interquartile  ranges
from median). CIV qualitative ratings (excellent, good, fair, and poor) appear on the right vertical
axis. Dashed lines indicate boundaries between ratings.
the excellent range, showing agreement. For
sites rated "good" or "fair" by staff, the range
of corresponding volunteer  CIVs was wide,
including CIVs which would  be  rated "fair" to
"excellent."  Differences between  staff  and
volunteer   ratings   may   be  due   to
misidentification of organisms  by volunteers,
a  misconception among  program volunteers
that water  quality is  always  "excellent"  in
Ohio's Scenic  Rivers  (potential  bias), or  a
greater  sampling   effort  by   volunteers  as
compared to staff  members, who  may rush
through many reference sites in a day. In the
Central  Ohio  area  (ODNR's   headquarters),
stream quality  monitoring coordinators have
received   better  instruction   on  sampling
strategy   through  frequent   contact  with
program  administrators and, as a result, the
volunteer and   staff  assessments  for  this
region  showed  closer   agreement.  This
suggests  that  part of the reason  for the
                                   general lack of agreement may  be due  to
                                   insufficient  sampling by  the ODNR stream
                                   quality monitoring coordinators,  although  all
                                   of  the   aforementioned  factors  probably
                                   contribute  to the  high variability  of  these

                                   The qualitative ratings  of ODNR's  volunteer
                                   monitoring technique do not necessarily agree
                                   with  the qualitative ratings  of  the  OEPA.
                                   However,  ODNR's  CIV ratings  do tend  to
                                   reflect the  attainment  ("excellent"  CIVs)  or
                                   non-attainment ("fair"  and "poor"  CIVs)  of
                                   aquatic life uses,  as designated  by the Ohio
                                   EPA,  for both the IBI and the ICI. Hence, the
                                   assessments may  be useful in screening sites
                                   at a basic level.

                                   CIV ratings tend  to reflect  IBI ratings  more
                                   accurately in streams and rivers  with smaller

                                                 Volunteer Monitoring Program Comparison
drainage areas. Drainage area did not appear
to have a marked  effect on  the  correlation
between  the  CIV   and  ICI,  but  further
collection of data could amplify an  otherwise
undetectable  effect.  Adequacy  of sampling
with the use of ODNR's technique may also
affect  the correlation.  The  results  of this
research suggest that larger  drainage  areas
may require  a modified approach, although
determining  exactly  what  that  approach
should  entail  is  beyond the  scope  of this

A review of  ODNR's database revealed that
program  volunteers  tend  to  overrate the
health of Ohio's Scenic Rivers as compared to
staff assessments. This is probably due to a
lack of  standardization  in  the  number of
samples collected and misidentification of the
organisms. These  problems could be solved
through  more  thorough  training  and  better
communication between ODNR, the regional
coordinators,  and the volunteers.  To improve
on the program, a measure of sampling  effort
and  better quantitative  estimates  could  be

It should be noted that the range of observed
CIV ratings used  for the comparisons in this
paper is  constricted. There  were  relatively
few sites which were rated "fair"  or "poor"
using  ODNR's  "kick-seine"  method.  Further
collection of  data will  be  necessary before
suggestions of revisions to the scoring criteria
or rating system can be made.

This research was supported by the School of
Natural  Resources, The Ohio State University,
through an Undergraduate  Honors  Research
Scholarship. This  research  could  not  have
been  completed   without   the  input   and
assistance of my field assistant/secretary (and
fiancee)  Chris  McKinney; Ed Rankin,  Chris
Yoder,   Dennis  Mishne,  Jeff  DeShon,  Mike
Bolton,   and   many  others   at   the  Ohio
Environmental   Protection  Agency;   Stuart
Lewis   and  John   Kopec   of  the   Ohio
Department of  Natural  Resources;  and  my
faculty   advisor  from   The   Ohio   State
University  School  of Natural  Resources,  Dr.
David L. Johnson. My sincere thanks goes out
to each of these individuals.

Literature Cited
Frost, S., A.  Huni, and W.E. Kershaw. 1971.
Evaluation of  a kicking technique for sampling
stream bottom  fauna.  Canadian  Journal  of
Zoology 49:167-173.

Karr,   J.R.   1981.  Assessment  of   biotic
integrity  using  fish  communities.  Fisheries

Kopec, J.S.  1989.  The Ohio  Scenic Rivers
Stream Quality  Monitoring Program: Citizens
in action,  pp.  123-127. ]n W.S. Davis and
T.P. Simon (eds). Proceedings  of the  1989
Midwest  Pollution Control Biologists  Meeting,
Chicago,  IL.  USEPA Region V, EPA 905/9-

Ohio Department of Natural Resources, n.d.
Ohio's Scenic River Stream Quality Monitoring
Program   -  A  citizen  action   program.
Columbus:   Ohio   Department   of   Natural

Ohio Environmental  Protection Agency. 1988.
Biological criteria for the  protection of aquatic
life: Volume  I.  The role  of  biological data in
water quality  assessment. Columbus, Ohio.

Rabeni,  C.F.  and  G.W.   Minshall.  1977.
Factors affecting microdistributton of stream
benthic insects. Oikos 29(1):33-43.

Resh, V.H. and J.D. Unzicker.  1975. Water
quality monitoring and aquatic organisms: The
importance of species identification. Journal
of  the Water   Pollution  Control  Federation

Schwenneker,  B.W.  and   R.A.  Hellenthal.
1984.  Sampling  considerations   in   using

stream insects  for monitoring  water quality.
Environmental Entomology 13:741-750.

             The Effects of Sediment Deposition on Insect Populations
                    and Production in a Northern Indiana Stream

 Gary W. Kohlhepp and Ronald A. Hellenthal
 Department of Biological Sciences
 University of  Notre Dame
 Notre Dame,  Indiana 46556

 In 1986 the  St. Joseph County, Indiana,  Drainage Board began conducting routine maintenance
 operations in  and along Juday Creek, a third-order tributary of the St. Joseph River. These activities,
 which include debris and snag removal from stream channels, have led to a large increase in sediment
 deposition into the lower reaches of the stream. Monthly benthic invertebrate samples were collected
 from June 1989 to June 1990 from a riffle area in Juday Creek and insect densities and secondary
 production  rates during this time were compared to those from a previous study at the same site in
 1981-82. Invertebrate density and  production rate responses varied based on functional feeding
 group. Among filter-feeders, two species showed significantly lower mean annual densities in 1989-
 90 compared to 1981-82, two species showed significantly lower densities during several months
 in 1989-90 versus corresponding months in 1981-82, and only one species (Hvdropsvche morosa)
 showed significantly greater density in 1981-82. Among collector-gatherers, mean annual densities
 were significantly  higher  for five of six species collected in 1989-90. Shredders showed mixed
 responses,  with two species having significantly higher mean annual densities in 1981-82, and one
 species, Taeniootervx nivalis. having higher densities in 1989-90. While production rates of three of
 the five species for which production rates were calculated increased  in 1989-90, the net effect of
 the increased sediment deposition was  a reduction in the combined production rates of the five
 species from  2765.1 mg/m2/year in 1981-82 to 653.8 mg/m2/year in  1989-90.

 Key Words: sediment, water quality, secondary production, functional feeding group, filter-feeders,
 collector-gatherers, shredders, benthic macroinvertebrates
The effects of anthropogenic disturbances such
as clear-cut logging, modification of riparian
vegetation, and changes in land-use practices
on streams and aquatic  organisms are well
documented (Dance and Hynes 1980, Swanson
et al. 1982, Ward 1984). Physical and chemical
attributes  of  streams  affected   by  these
activities include light penetration (Mclntire and
Colby 1978), water temperature (Hall and Lantz
1969, Holtby  1988), nutrient concentrations
(Chauvet and Decamps 1989), and inputs  of
woody debris  and sediments (Bryant 1983,
Swanson et al. 1987). Such changes have been
shown  to  affect  algae  (Minshall   1978),
macrophytes  (Hynes  1970),  invertebrates
(Newbold  et al. 1980), fish  (Thedinga  et al.
1989, Ward and  Stanford 1989)  and  other
vertebrates (Hawkins et al. 1983,  Spencer  et
al. 1991). However, impacts associated with
more subtle or  routine  activities such  as
removal  of  snags  and woody  debris  from
streams  are less understood. Because  such
operations   likely   represent  common   and
widespread  management   practices  in  the
midwestern  United States,  quantifying  their
impact on the biota is essential.

In 1986, the  St.  Joseph  County (Indiana)
Drainage Board  began to perform maintenance
operations in and along Juday Creek,  a third-
order tributary of the St. Joseph River. Included
in these activities are removal of snags, debris,
and trees that block water flow and that might
result in flooding or pooling along the  stream.
Coincident with these practices in Juday Creek
has been a large increase in sediment transport
and  subsequent  deposition  in  downstream

Kohlhepp and Hellenthal
locations.  Because data on insect production
rates and populations were collected from this
stream in  1981-82  (Schwenneker 1985) and
1985-86 (M.B. Berg pers. comm.),  prior  to
drainage  board  activities,   we   had   an
opportunity to document changes in population
densities  and   production  rates  of  stream
benthos that may have resulted from stream
maintenance operations. Our objectives were to
consider 1) the effects of increased sediment
deposition   on   benthic   macroinvertebrate
populations; 2) whether changes in population
densities  were   reflected  by  changes  in
invertebrate secondary production rates; and 3)
the implications of these results for predicting
invertebrate responses to sediment deposition.

Study Site
This study was conducted in a riffle area of the
creek (41°43'N, 86°16'W, elevation  - 206m)
on  land owned and maintained  by the  St.
Joseph Co. Chapter of the Izaak Walton League
of America. Juday Creek is important from a
conservation perspective because it is one of
only a few streams in Indiana known to support
breeding trout populations. The upper segment
flows through  flat, agricultural land, and then
makes  its way  through primarily residential
areas. The lower segment, which includes the
study   location,   flows   through   natural,
deciduous  woodlands and has a gradient of
1.3%. The site is heavily shaded from the late
spring through the early fall, and the substrate
is a mixture of sand, gravel, and cobble. Some
of the physical and  chemical data collected at
this  site  in  1981-82  and   1989-90   are
summarized in Table 1.

A silt trap  is located  approximately 100 m
upstream  from the site. It was built to protect
the  lower  segment of Juday Creek  from
excessive sediment deposition. Prior to drainage
board  operations,  approximately 18 yd3 of
sediment were removed from the silt trap every
1.5-2 years. Since 1988 the silt trap has been
dredged once a year. In 1989, approximately
90 yd3 of sediment were removed from the trap
(J. Moore, pers. comm.). However, during the
1989-90  sampling  period,  the  trap  was
completely filled after 4 months. Therefore, it
provided little or no protection to the lower part
of the stream for 8 months. During March and
early April 1990,  a  large pulse of sediment
entered the study  area, resulting in  extensive
coverage of the gravel and cobble  substrate
with sand. This pulse was apparently a result of
an   unusually   high  number   of   stream
maintenance  activities during  this  period
compared to previous years.

Materials and Methods
Benthic samples were collected monthly for a
period of thirteen months during the course of
two separate studies.  The  first  was  from
September   1981  to   September   1982
(Schwenneker 1985)  and the second from June
1989 to  June 1990. In  both studies,  ten
random benthic bottom samples were collected
each month from  the  riffle using a 0.09 m2
Hess sampler with a mesh size of 333 fjm.
Because  of  extremely  high  water levels,
samples could not  be collected in January and
March  of  1982, and these two months were
omitted  from comparisons  between  years.
Samples were preserved in 80%  ethanol and
transported   back   to  the   laboratory  for
processing and analysis.

In the  laboratory, invertebrates were  sorted
from  the  substrate  using  sugar  flotation
(Anderson 1959) and then identified to species
and instar or size  class. Instar determinations
were based on head  capsule width, except for
stoneflies and mayflies. These organisms were
divided into size classes based on body length
from the front of the head to the base of the
cerci. Population  densities were recorded for
the   most   abundant  species   (excluding
chironomids), and each was assigned to  a
functional feeding  group (Merritt and Cummins
1984). Mean annual densities were compared
between  years  using a  repeated measures
analysis of variance on log transformed data

                                                      Sediment Deposition Effects on Insects
Table 1. Comparison of physical and chemical
characteristics of Juday Creek in 1981-82 and

Temperature (°C)
Current Velocity (m/s)
Depth (cm)
Alkalinity (mg/l CaC03)
Nitrate (mg/l-N)
Orthophosphate (mg/l)
Conductivity (t/mho/cm)
In addition, densities of each  species were
compared month by month between years (i.e.,
January  1982 versus January 1990) using a
one-way  ANOVA.  Production  rates  were
calculated for all species in the 1981-82 study,
and  for five  species  in the 1989-90 study,
including   Hydropsvche   soarna  Ross,
Hydropsvche   betteni  Ross,   Qptioservus
fastiditus   (LeConte),   Baetis   vaoans
McDunnough, and Taenioptervx nivalis (Fitch).
Dry  weights  for each instar or size  class of
these organisms were obtained by drying at
70°C for 48  hours.  Production  rates were
measured using either the instantaneous growth
method for those species with distinguishable
cohorts, or the size-frequency method for those
with cohorts  that could not be  distinguished.
Instantaneous growth calculations were made
from   computer   programs  developed   by
Schwenneker (1985)  and Berg  (1989). Size-
frequency production  calculations were made
using the Aquatic Ecology-PC software package
(Ekblad 1986). Because early instars often are
inefficiently  sampled, apparent  increases in
population densities are sometimes seen during
cohort  development.  For   production  rate
calculations,  densities were back-calculated
using the catch-curve method of  Waters and
Crawford (1973). The result of this correction
was  that if densities were  lower  at sampling
time  T-1  than at time T, densities at time T-1
were set equal to those at  time T. Given the
typical log-type  decline exhibited  by stream
invertebrates,  this  correction  represents  a
conservative  estimate   of  densities
(Schwenneker 1985).

Population responses of individual species to
sedimentation varied depending on functional
feeding group. Among the six species of filter-
feeders, four  had lower mean annual densities
in  1989-90,   although  only two of  these
differences were statistically significant (Table
2). Hvdropsvche soarna and Chimarra obscura
(Walker)  showed  the most  dramatic density
reductions in   1989-90 (over 80% and  95%
respectively).   Mean   annual  densities  of
Hvdropsvche   betteni  and   Cheumatopsvche
petiti  (Banks)  were  not  significantly different
between years, but  each did show significant
density differences in eight of the ten monthly
comparisons  (Fig.  1).  During  six of  these
months, ]H- betteni showed  higher densities in
1981-82 compared  to 1989-90, and £.  petiti
had significantly higher  densities in 1981-82
during five months. Hvdropsvche morosa (Ross)
mean annual density was significantly higher in
1989-90 compared to 1981-82, and Simulium
sp. had similar densities between years. All six
filter-feeders had lower densities, four of which
were  significant (p<.05), in April and May of
1990 versus 1982.

The reduction in secondary production rates of
filter-feeders  in  1989-90  was much   more
pronounced   than  changes  in   population
densities (Fig. 2). The  production  rate of ]H.
sparna dropped more than  90% in 1989-90
from the 1981-82 rate. From 1981-82 to 1989-
90, the  production rate of JH. betteni  declined
by more than  50%,  while mean annual density
was reduced about 30% in 1989-90.

Population  densities   of  collector-gatherers
showed different responses to  the increased
sediment deposition than the of filter-feeders.
Five  of the six species  of collector-gatherers
showed  significantly greater  mean  annual
densities in 1989-90 compared to 1981-82

Kohlhepp and Hellenthal

Table 2. Mean annual densities (N/M2) and p values of individual taxa in
= p < 0.05; •• = p < 0.01.
1981-82 1989-90
Mean (1 SE) Mean (1 SE)
Cheumatopsvche petiti
Chimarra obscura
Hvdropsyche betteni
Hvdropsvche morosa
Hvdropsyche sparna
Simulium sp.
Antocha so.
Baetis vaoans
Macronvchus olabratus
Qptioservus fastiditus
Stenelmis crenata
Stenonema spp.
Arnphinemura delosa
Taenioptervx nivalis
Tipula abdornjnalis
(Table 2). The elmid beetle

44.2 (4.0)
258.8 (23.0)
26.3 (2.4)
1648.7 (136.9)
45.9 (5.4)

43.9 (3.3)
17.3 (2.4)
22.1 (1.7)
3.9 (0.7)

66.6 (8.6)
3.3 (0.5)
larvae Qptioservus greater

2.2 (0.4)
293.7 (35.6)

57.6 (5.2)
34.0 (4.5)
155.1 (13.0)
91.3 (10.7)

3.4 (0.7)
1.3 (0.3)
in 1 989-90.
1981 -82 and 1989-90. «



Densities of some species
fastiditus and Stenelmis crenata (Say) had the
largest  increases  in  mean annual densities
between years  (650% and  almost  4000%,
respectively I. Macronvchus glabratus (Say) was
collected so rarely in 1981-82 that its density
was  not reported  by Schwenneker (1985).
Therefore,  we could not  compare densities
between years, although densities were clearly
of collector-gatherers dropped in April and May
of 1990,  although some declines  could  be
explained partly by life history characteristics.

We  compared  secondary   production  rates
between years for two species of collector-
gatherers, Qptioservus  fastiditus and  Baetis
vagans. Production rates showed changes

                                                      Sediment Deposition Effects on Insects
        **     •
       Apr  May  Jun  Jul  Aug  8«P Oel  Nov  Dec

             •I Wat-62  ^ 19M-BO
                                                   Combined  H. iparna H. bttltnl O tutldllui B. v«gtrn  T. nlvtlli

                                                            •11981-82  £211989-80
   ftt  Apr M«y Jun  Jul  Aug S*p  Oet  Nm Dec
Figure 1. (A)  Mean densities (N/M2) by month
for Hvdropsvche betteni in 1981-82 and 1989-
90.  (B) Mean densities  (N/M2)  by month of
Cheumatopsvche oetiti in 1981-82 and 1989-
90. * = p<0.05; ** = p<0.01).
similar to those of the population densities for
these species. Production rates of Q. fastiditus
and  B.  vaaans were more than 300% and
200% greater, respectively, in 1989-90 than in
1981-82 (Fig. 2).
                                               Figure  2.  Combined  (all  5  species)  and
                                               individual species secondary production rates
                                               (mg/m2/yr) in 1981-82 and 1989-90.
Population densities of shredders showed mixed
responses  between  years (Table 2).  Mean
annual  densities  of  Amohinemura   delosa
(Ricker) and  Tioula  abdominalis  (Say) were
significantly lower in 1989-90 than in 1981-82,
Mean annual densities of Taeniootervx  nivalis.
on the other hand,  increased significantly in
1989-90. However, the secondary production
rate of T. nivalis was only 20% higher in 1989-
90  (Fig.  2),  because  of  a  greater  mean
individual biomass in 1981-82.

The predator functional feeding group is  not
represented because  the only  predaceous
invertebrates at this location besides f latworms
were   the     filter-feeding   hydropsychid
caddisflies.   The   only  obligate   scraper,
Glossosoma intermedium (Klapalek), had greater
densities in 1981-82 than in 1989-90 (8.91 /m2
vs. 6.6/m2), but differences were not significant
(p = 0.389).

Three  of   the  five  species  for which we

Kohlhepp and Hellenthal
Collector -gatherers
          Collector-gatherers   4 25
                 1.29       ^
Figure 3. Percent contribution by functional feeding  group to  (A) 1981-1982 total densities, (B)
1989-90 total  densities,  (C)  1981-1982 secondary production  rates, (D) 1989-90  secondary
production rates. Relative values are calculated using only five species in Figure 2.
calculated  secondary  production rates  had
higher rates in 1989-90 compared to 1981-82.
However, summing the production rates for all
five species resulted in a substantial decline in
production  rates  in  1989-90  compared to
1981-82 (Fig. 2). Of these five species, the
two  filter-feeders  in  1981-82  contributed
88.48% of the  numbers  and 94.46% of the
combined secondary production rate (Fig. 3). In
1989-90, these filter-feeders declined in both
densities and production  rates, resulting  in a
large decline in total  secondary production
rates. The three species that  increased in
production rate in  1989-90 versus  1981-82
were those  that contributed  little  to  total
production rate. The result is  a shift from a
community dominated by filter-feeders in both
numbers and production rate in 1981-82 to a
community  in  1989-90  in  which  collector-
gatherers    and  shredders   increased  in
                   importance in terms of relative contribution to
                   both numbers and production (Fig. 3).

                   It seems likely that  the shift in invertebrate
                   community structure and decreased secondary
                   production rates observed in 1989-90 was the
                   result of increased sediment deposition into the
                   study area. Differences cannot be explained by
                   changes in the physicochemical characteristics
                   of Juday Creek between years (Table 1). Other
                   authors have noted that reduction or elimination
                   of the overhead canopy can result in abrupt
                   changes in abundances of stream invertebrates
                   (Behmer  and Hawkins 1986, Newbold et al.
                   1980). However, the riparian vegetation in this
                   part of the stream has remained undisturbed.
                   Attributing  differences   between  years  to
                   increased fish  predation also is unsupported.
                   Furthermore, data collected in 1985-86 showed

                                                       Sediment Deposition Effects on Insects
 that insect species abundances were similar to
 those  from 1981-82 in this  section of  the
 stream (M.B. Berg, pers. comm.). The county
 drainage board began maintenance operations
 along  Juday Creek soon after the  1985-86
 study. Almost immediately, sediment deposition
 into the lower portion of Juday Creek increased
 substantially.  The result  has  been  severely
 reduced population densities for some species
 and higher densities for others.

 Functional feeding group classification proved
 to be a good predictor of a species' population
 response  to the sediment deposition. Filter-
 feeder  population densities were the  most
 severely reduced (Table 2, Figure 1). Reduction
 of filter-feeder densities due to greater sediment
 deposition is consistent with their biology and
 habitat requirements. These organisms require
 a  solid and somewhat  stable substrate  on
 which to spin their nets (Hynes 1970, Minshall
 1984), and are considered intolerant of heavily
 silted and sandy  areas where  they lose their
 attachment to the substrate (Marsh and Waters
 1980, Tebo 1955). In addition, large  amounts
 of suspended inorganic sediments can clog nets
 and interfere with the feeding mechanisms of
 the net-spinners  (Nuttall  and  Bielby 1973).
 These  results  generally  are consistent with
 those found in other studies. The net-spinning
 caddisflv Arctoosvche orandis (Banks) is highly
 sensitive to sedimentation due to the  filling of
 interstitial spaces of rocks (Cline et al. 1982,
 McClelland and Brusven  1980). Barton (1977)
 found   reduced   densities  of  Hvdropsvche
 slossonae Banks (48%) and Cheumatopsvche
 sp. (66%) in an  Ontario stream immediately
 downstream from a highway construction site
 compared to upstream locations. He suggested
 that damage to  benthos  only occurs when
 stones are buried  by sediment,  which  was the
 case in our study. Gurtz and Wallace (1986)
and Benke et al. (1984) found lower production
rates of net-spinning caddisflies on sand when
compared with more stable substrates. Other
studies also have found that filter-feeders are
intolerant of sediment additions (Cherry et al.
 1979,  Nuttall  and Bielby 1973,  White and
 Gammon 1976).

 In contrast to the filter-feeders, mean annual
 densities of  all collector-gatherers increased
 significantly  in  1989-90 from the  1981-82
 levels  (Table  2).  Although   most  of  the
 deposited sediment was inorganic sand during
 the  1989-90 study,  the amount  of organic
 material deposited in the riffle area probably
 increased as well. This would result in a greater
 food supply for the collector-gatherers, which
 forage along the substrate for food particles
 (Berkman et al. 1986, Merritt and Cummins
 1984). Elmid beetles showed the greatest rise
 in numbers  (Table 2). Although studies  have
 shown that  Qptioservus  and Stenelmis prefer
 larger, solid  substrates  (Cummins and  Lauf
 1969, Marsh and Waters  1980, Rabeni  and
 Minshall 1977), these organisms are known to
 tolerate fine sediments to some extent (Brown
 1987, White and Gammon 1976). The reason
 for the increased  density  of Macronvchus
 glabratus in  1989-90 is unclear,  since this
 species is almost always associated with wood
 substrate  (Brown  1987,  Hynes 1970). The
 amount of woody debris in the stream was not
 quantified   in   1981-82  or   in   1989-90.
 Stenonema also is somewhat tolerant of silt
 (Dance and  Hynes  1980, Jones  and  Clark
 1987), and  could benefit from an  increase in
 deposition of organic matter. Investigators have
 found that Baetis nymphs generally prefer larger
 substrates and drift  in the presence of  large
 quantities  of  sediment  (Culp  et  al.  1986,
 Wagner 1989, White and  Gammon 1976).
 However,  numbers of the European  species
 Baetis  rhodani  (Pictet) increase as sediment
 deposition increases (Nuttall and Bielby 1973,
 Scullion and Edwards 1980). Wallace and Gurtz
 (1986) found that although production rate of
 Baetis sp. was greatest on cobble and gravel,
 individuals  of  this   genus were   found  in
moderate numbers in  sandy areas.  Culp et al.
 (1986) showed that  while sediment transport
reduced B.  tricaudatus  Dodds  densities by
67%, sediment deposition actually decreased
drift  rates of this species.  As with the other
collector-gatherers in this study, the potential

Kohlhepp and Hellenthal
increase in food material transported into the
study  area  may have  offset  any negative
effects on B. vaaans associated with the loss of
stable  substrate. However, densities  of  Q.
fastiditus. Antocha sp.f and Stenonema spp. all
decreased substantially  in  April and May of
1990, coinciding with the sediment pulse that
covered the  study  site.  It is  possible that
continued heavy sediment  deposition could
eventually cause population declines in many of
these collector-gatherers.

The  population  responses  of  the shredder
functional feeding group varied depending on
the  species  (Table 2).  Taeniootervx  nivalis
appeared  to  benefit  from  the  increased
sediment  deposition.  T. nivalis nymphs are
generally found  in leaf  packs or other debris
along the  stream margins (Sephton and Hynes
1984,  Stewart and  Stark 1988). This species
may not be affected by sediment that covers
the cobble and gravel substrate as long as there
are sufficient leaf packs in the stream. During
this  study,  there  were  many  leaf  pack
accumulations both in the main channel and
along the margins and backwater areas. Like
the  collector-gatherers, T. nivalis may have
benefirted from a potential increase in available
detritus in the  study  area. Because adults
emerge in February and the nymphs diapause
deep in the substrate until September (Stewart
and  Stark 1988), this species would not have
been affected by  the  sediment  pulse  that
occurred  in  the spring.  In contrast, Tioula
abdominalis had significantly lower densities in
1989-90  compared to  1981-82 (Table  2).
Cummins and Lauf (1969) found that Tioula
calootera Loew preferred coarse substrates, but
showed a wide tolerance for finer sediments,
including  silt. They suggest that  these larvae
are  probably found in  microhabitats  of finer
sediments and organic  material between and
behind coarse sediments. If J_. abdominalis has
similar tolerances, then this species should be
moderately   affected   by  the  increased
deposition. Amphinemura delosa densities also
were significantly lower in  1989-90 (Table 2).
Another  congeneric  species,  A. sulcicollis
(Stephens), is known to move from leaves to
stone as it develops (Hynes 1976). If A., delosa
exhibits a similar habitat shift, then a reduction
in numbers would be expected with increasing
sediment deposition.  However,  Scullion and
Edwards (1980) found that A. sulcicollis was
tolerant to mine discharge siltation. The lack of
tolerance  of  A.  delosa   in  our study  may
represent   species-specific   differences   in
tolerance or may be related to the amount of
material deposited.

Most studies  that examine the  effects of
sediment deposition on benthic  invertebrates
look only at changes in population densities or
relative changes in species composition (Barton
1977,  Nuttall and Bielby  1973, Scullion and
Edwards 1980). The  few studies that have
looked at changes  in biomass found it to be a
better  measure  of  benthos   response to
sedimentation   than  density and  diversity
(Letterman and Mitsch   1978,  Marsh and
Waters 1980). We have found no studies that
examine the effects of sediment  deposition on
the  secondary  production  rates  of  stream
invertebrates.   Studies   have  compared
invertebrate  secondary   production   rates
between  logged   and unlogged  areas (eg.
Wallace and Gurtz 1986), but have difficulty
separating  the  relative effects of  sediments,
increased  algal   growth,   and   water

There  are many  advantages  to  calculating
secondary production  rates rather than looking
only at changes  in  abundances.  Secondary
production incorporates a measure of individual
growth as well  as population  density,  and
provides   a   measure   of   the   functional
importance of a species to stream energy flow
and nutrient  processing  (Short  et al.  1987).
From a management perspective, invertebrates
are an important component of  fish diets and
may limit fish production (Benke 1984, Tebo
 1955). Our results suggest that  changes in
secondary production rates between years give
a  clearer picture of what is happening to the
benthos in Juday  Creek  than changes in

                                                      Sediment Deposition Effects on Insects
abundance. Because eight species show greater
densities in 1989-90 compared to six species
with greater densities in 1981-82, examination
of numbers atone suggests that the only impact
of the sediment on the invertebrate community
was a change in relative abundance of species.
However,  the species  that  had  reduced
densities in 1989-90 are those that contributed
the most to invertebrate secondary production
rates in  1981-82. Those that  had  increased
densities  in   1989-90  contributed  little  to
secondary production in  1981-82 (Figure  2).
The result was that the combined production
rates for five species decreased from  2765.1
mg/mz/year in 1981-82 to 653.8 mg/mz/year in
1989-90. This 78% reduction in secondary
production rate represents a substantial decline
in the amount of food  available to  support
higher trophic level organisms such as fish. The
response of Taeniootervx nivalis populations
also represents a good example of the value of
calculating production rates. While densities of
this species increased approximately  6 times in
1989-90  from  1981-82,  production  rates
increased only 20% (Fig. 2). The large increase
in mean  densities of this species was mostly
offset by lower individual growth rates.

Our results suggest that a  functional  feeding
group  classification  of  organisms  offers a
method for predicting the downstream impacts
of  stream   maintenance  activities.   If   a
community is dominated by filter-feeders, then
substantial impacts associated with an increase
in sediment deposition may  be expected. If
collector-gatherers contribute the majority of
invertebrate   production,   then   moderate
increases in sediment transport and deposition
may actually enhance population densities and
production  rates through increased  import of
organic material.  However, heavy,  sustained
sediment  deposition  will  probably  have  a
negative  impact on many of these species as
well. In addition,  secondary production  rates
are  a better measure of invertebrate community
response than diversity  and population density.
Finally, many studies have documented that the
benthos recovers  rapidly when the  source of
sediment  is eliminated and  high flows  are
allowed to wash the  sediment downstream
(Barton 1977, Cherry et al. 1979, Tebo 1955).
If silt traps are properly maintained and cleaned
out before filling with sediment, then it is
possible that negative impacts associated with
instream  maintenance  operations  can  be

We would like to express our appreciation to
the St. Joseph Co. Chapter of the Izaak Walton
League of America for  allowing access to  the
stretch of Juday Creek that runs through their
property,  and especially  the caretaker, John
Moore, for all of his help during the study.  We
also are grateful to Barb  Hellenthal for  her
comments on   an   earlier  version  of  this
manuscript. This research was supported by a
grant from the Indiana Academy of Science.

Literature  Cited
Anderson,  R.O. 1959. A  modified  flotation
technique for sorting bottom fauna  samples.
Limnology and Oceanography 4: 223-225.

Barton, B.A.  1977. Short-term  effects of
highway constuction on the limnology of a
small stream in southern Ontario.  Freshwater
Biology 7: 99-108.

Behmer, D.J. and C.P. Hawkins. 1986. Effects
of  overhead  canopy  on  macroinvertebrate
production in  a Utah  stream.   Freshwater
Biology 16: 287-300.

Benke, A.C. 1984.  Secondary production of
aquatic insects.  Pages  289-322 in V.H. Resh
and D.M.  Rosenberg (editors). The ecology of
aquatic insects. Praeger Press, New York.

Benke,  A.C.,  T.C.  Van  Arsdall, Jr.,  D.M.
Gillespie, and F.K. Parrish. 1984. Invertebrate
productivity in a subtropical blackwater river:
the importance  of  habitat  and  life  history.
Ecological Monographs  54(1): 25-63.

Kohlhepp and Hellenthal
Berg,  M.B. 1989.  The role of Chironomidae
(Diptera) in stream insect secondary production.
Ph.D.  Dissertation, University of Notre Dame,
Notre  Dame, Indiana. 239 pp.

Berkman,  H.E., C.F. Rabeni, and T.P. Boyle.
1986.  Biomonitors  of  stream  quality  in
agricultural areas:  fish versus invertebrates.
Environmental Management 10(3): 413-420.

Brown, H.P. 1987. Biology of riffle  beetles.
Annual Review of Entomology 32: 253-273.

Bryant, M.D. 1983. The role and management
of woody debris in west coast salmonid nursery
streams.  North American Journal of Fisheries
Management 3: 322-330.

Chauvet,  E. and H. Decamps. 1989. Lateral
interactions in a fluvial landscape: the River
Garonne, France. Journal of the North American
Benthological Society 8(1): 9-17.

Cherry, D.S., S.R. Larrick, R.K. Guthrie, E.M.
Davis, and F.F. Sherberger. 1979. Recovery of
invertebrate and vertebrate populations in  a
coal ash stressed drainage system. Journal of
the Fisheries Research Board of Canada 36:

Cline, L.D., R.A. Short, and J.V. Ward. 1982.
The influence of highway construction on the
macroinvertebrates and epilithic algae of a high
mountain  stream. Hydrobiologia 96: 149-159.

Gulp,  J.M., F.J.  Wrona, and R.W.  Davies.
1986. Response of stream benthos and drift to
fine sediment deposition versus transportation.
Canadian  Journal of Zoology 64:1345-1351.

Cummins, K.W. and  G.H.  Lauff. 1969. The
influence  of  substrate particle  size  on the
microdistribution   of  stream  macrobenthos.
Hydrobiologia 34: 145-181.

Dance, K.W. and H.B.N. Hynes.  1980. Some
effects of agricultural land use on stream insect
communities. Environmental Pollution (Series A)
22: 19-28.

Ekblad, J. 1986. Aquatic Ecology-PC. Oakleaf
Systems, Decorah, Iowa.

Gurtz,   M.E.   and  J.B.  Wallace.  1986.
Substratum-production  relationships  in  net-
spinning caddisflies (Trichoptera) in  disturbed
and undisturbed hardwood catchments. Journal
of the  North American Benthological Society
5(3): 230-236.

Hall, J.D. and R.L. Lantz. 1969. Effects of
logging  on the habitat  of  coho salmon  and
cutthroat trout in coastal streams. Pages 355-
376 in T.G.  Northcote (editor). Symposium on
salmon and  trout in streams.  February 22-24,
1968. University of British Columbia.

Hawkins,  C.P., M.L. Murphy, N.H. Anderson,
and M.A.  Wilzbach. 1983. Density of fish and
salamanders in relation to riparian canopy and
physical habitat in streams of the northwestern
United States. Canadian Journal of Fisheries
and Aquatic Sciences 40(8): 1173-1184.

Holtby, L.B.  1988, Effects of logging on stream
temperatures  in   Carnation   Creek,  British
Columbia, and associated impacts on the coho
salmon  (Oncorhvnchus  kisutch).  Canadian
Journal of Fisheries and Aquatic Sciences 45:

Hynes,  H.B.N. 1970.  The ecology of running
waters. Liverpool  University Press,  Liverpool,
Great Britain.

Hynes, H.B.N. 1976.  Biology  of Plecoptera.
Annual Review of Entomology 21: 135-154.

Jones, R.C. and C.C. Clark.  1987.  Impact of
watershed  urbanization  on  stream  insect
communities. Water Resources Bulletin 23(6):

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

                                                      Sediment Deposition Effects on Insects
Marsh, P.C. and T.F. Waters. 1980. Effects of
agricultural drainage development on benthic
invertebrates  in   undisturbed   downstream
reaches. Transactions of the American Fisheries
Society 109: 213-223.

McClelland, W.T., and M.A.  Brusven.  1980.
Effects  of sedimentation on the behavior and
distribution of  riffle insects  in  a  laboratory
stream. Aquatic Insects 2:161-169.

Mclntire,  C.D. and J.A.  Colby.  1978. A
hierarchical   model  of   lotic   ecosystems.
Ecological Monographs 48:167-190.

Merritt, R.W. and K.W. Cummins.  1984. An
introduction to the aquatic insects. Second ed.,
Kendall/Hunt  Publishing Company, Dubuque,

Minshall, G.W. 1978.  Autotrophy  in  stream
ecosystems. BioScience 28(12): 767-771.

Minshall,   G.W.   1984.   Aquatic  insect-
substratum relationships.  Pages 358-400 in
V.H. Resh and D.M. Rosenberg (editors).  The
ecology of aquatic insects.  Praeger Press, New

Newbold,  J.D., D.C. Erman,  and K.B.  Roby.
1980. Effects of logging on macroinvertebrates
in  streams with  and without  buffer strips.
Canadian  Journal  of  Fisheries and  Aquatic
Sciences 37: 1076-1085.

Nuttall, P.M. and G.H. Bielby. 1973. The effect
of china-clay wastes on stream invertebrates.
Environmental Pollution 5:77-86.

Rabeni, C.F. and G.W. Minshall. 1977. Factors
affecting microdistribution  of stream benthic
insects. Oikos 29: 33-43.

Schwenneker, B.W. 1985.  The contribution of
allochthonous  and  autochthonous  organic
material to aquatic insect secondary production
rates  in  a  north  temperate  stream.  Ph.D.
Dissertation, University of Notre Dame, Notre
Dame, Indiana. 363 pp.

Scullion,  J.  and R.W. Edwards.  1980.  The
effects of  coal  industry  pollutants on  the
macroinvertebrate fauna of a small river in the
South Wales coalfield. Freshwater Biology 10:

Sephton, D.H. and H.B.N. Hynes. 1984.  The
ecology  of   Taeniootervx   nivalis  (Fitch)
(Taeniopterygidae; Plecoptera) in a small stream
in southern  Ontario.  Canadian  Journal  of
Zoology 62: 637-642.

Short, R.A.,  E.H. Stanley, J.W. Harrison,  and
C.R. Epperson. 1987. Production of Corvdalus
cornutus (Megaloptera) in four streams differing
in size, flow,  and temperature. Journal of the
North American  Benthological Society 6(2):

Spencer,  C.N.,  B.R. McClelland,  and  J.A.
Stanford.  1991.   Shrimp stocking, salmon
collapse,  and eagle displacement. BioScience
41(1): 14-21.

Stewart, K.W. and B.P. Stark.  1988. Nymphs
of North American stonefly genera (Plecoptera).
The Thomas Say Foundation, Vol. 12. 460 p.

Swanson, F.J., S.V. Gregory, J.R. Sedell, and
A.G. Campbell. 1982. Land-water interactions:
the riparian  zone.  Pages 267-291 in  R.L.
Edmonds (editor) Analysis  of coniferous forest
ecosystems  in the  western  United States.
US/IBP Synthesis Series 14.  Hutchinson Ross
Publishing, Stroudsburg, PA.

Swanson, F.J., L.E. Benda, S.H. Duncan, G.E.
Grant,  W.F. Megahan, L.M. Reid,  and R.R.
Ziemer.   1987.  Mass  failures  and  other
processes  of sediment production  in Pacific
Northwest forest  landscapes.  Pages 9-38 in
E.O. Salo and T.W. Cundy (editors). Streamside
management: forestry and fishery inter- actions.
University  of  Washington, Institute  of Forest

Kohlhepp and Hellenthal
Resources, Seattle.

Tebo,  L.B.,  Jr.  1955.  Effects  of  siltation,
resulting from improper logging on the bottom
fauna of a small trout stream in the southern
Appalachians.  The Progressive  Fish-Culturist

Thedinga,  J.F., M.L. Murphy, J. Heifetz, K.V.
Koski,  and S.W. Johnson.  1989. Effects of
logging on size and age composition of juvenile
coho  salmon  (Oncorhvnchus  kisutchl   and
density  of  presmolts  in  southeast  Alaska
streams.   Canadian Journal  of  Fisheries  and
Aquatic Sciences 46: 1383-1391.

Wagner,  R.  1989. The  influence of  artificial
stream bottom siltation  on  Ephemeroptera in
emergence traps. Archive  fur  Hydrobiologie
115(1): 71-80.

Wallace, J.B. and M.E. Gurtz. 1986. Responses
of  Baetis mayflies   (Ephemerop-  tera)  to
catchment logging. American Midland Naturalist
115(1): 25-41.

Ward, J.V. 1984. Ecological perspectives in the
management of aquatic insect habitat.  Pages
558-577  in  V.H. Resh  and O.M. Rosenberg
(editors)   The  ecology  of  aquatic  insects.
Praeger Press, New York.

Ward, J.V. and J.A. Stanford. 1989. Riverine
ecosystems: the influence of man on catchment
dynamics and fish ecology. Pages 56-64 in D.P.
Dodge (editor). Proceedings of the International
Large River Symposium.  Can. Spec. Publ. Fish.
Aquat. Sci. 106.

Waters, T.F. and G.W. Crawford. 1973. Annual
production of a  stream mayfly population.
Limnology and Oceanography 18: 286-296.

White, D.S. and J.R. Gammon. 1976.  The
effect of suspended solids on macroinvertebrate
drift in an Indiana creek. Proceedings of the
Indiana Academy of Science 86: 182-188.

            Agricultural Impacts on the Fishes of the Eel River,  Indiana

 James R. Gammon and Clifford W. Gammon
 Department of Biological Sciences
 DePauw University,
 Greencastle, Indiana 46135

 The Eel River of northern Indiana is a major tributary of the Wabash River. It is approximately 177
 km (110 mi) in  length with an average rate of descent of 0.457 m/km (2.41  ft/mi). Approximately
 79% of its 210,800 ha (814 m2) drainage basin is devoted to row-crop agriculture. The fish
 communities and habitat were studied during the summer of 1990. Fish were collected from 25 sites
 located throughout the Eel River and some of its tributaries. A 3/16 inch mesh, 30 ft by 4 ft seine
 was effective in collecting small fish including darters. Backpack electrofishing was also used at most
 stations on two separate dates. Historic records of the fish communities were examined and, when
 possible, converted into  Indexes  of Biotic Integrity values so that changes over time could be
 estimated. Habitat evaluation included a mainstem reconnaisance,  a habitat survey (HEP) conducted
 at all collecting  stations, and a synoptic turbidity survey on July 16 and 17,1990. Estimates of the
 amount of woodland  were made from conventional  analysis of enlarged infrared photographs.
 Analyses of existing suspended sediment data were used to evaluate possible impacts of nonpoint-
 source influence from agricultural  fields as well as historic records of fish kills and chemical spills
 within the Eel River watershed. The 1990 fish community was generally better than the community
 found in 1982. However, this improvement is probably temporary and the result of a series of recent
 years when both river discharge and suspended sediment concentrations were lower than normal.
 From a  longer time perspective the fish community is degraded, with many species which were
 common 50 years ago now either absent or very severely reduced.  Rainbow darter, orangethroat
 darter, bluebreast darter, and stonecat were not collected at all. Sculpin, greenside darter, blackside
 darter, silver shiner, rosyface shiner, longear sunfish,  and smallmouth bass were very restricted in

 Key Words:  non-point  source, habitat, suspended sediment, fish community, Indiana
In  1982  populations  of smallmouth bass
(Micropterus  dolomieui)  were found  to be
virtually  lacking  by  Braun and Robertson
(1982) who collected from the same sites as
Taylor (1972). Exerting roughly equivalent
effort and similar  methods, Taylor  (1972)
collected 98 smallmouth bass while Braun and
Robertson (1982) found only  3. During  the
1980's  smallmouth bass populations  in  the
lower part of the Eel River were augmented by
stocking fin-clipped  fingerlings  (5130 fish on
10-28-83; 5000 on  9-17-85; and 6960 on 4-
17-86).  A  limited  number  of  stations were
more  intensively  sampled  and additional
tributaries were also investigated.

The current study  was  planned  to provide
information about the fish communities at all
of Taylor's sites and a few additional sites. It
included an evaluation of  instream and near-
stream  habitat  from  the  standpoint   of
agricultural nonpoint sources of pollution  and
their  possible  influence  on  those   fish
communities.  This report is a condensation
and extension of a larger report to the Indiana
Department of  Environmental  Management
(Gammon  and Gammon  1990).

The Study Area
The  Eel  River is  a major tributary of  the
Wabash River in northern Indiana.  Originating
in northwest  Allen county near Ft. Wayne, it
flows southwest for  approximately 177  km
(110  miles)   through  Kosciusko,  Whitley,
Wabash and Miami counties into the Wabash

Gammon and Gammon
River at Logansport in Cass county. Its rate of
descent is approximately 0.457  m/km (2.41
ft/mile) with a lower rate in the upper third and
a slightly higher rate in the lower  20 km.

This area originally contained glacial lakes and
swampy  wetlands,  but  it was  extensively
ditched  and   drained  prior  to  1900  for
agricultural use.  Approximately  79% of its
2,148  km2 (814 m2)  drainage basin area
(Hoggatt 1975) is devoted to rowcrop agricul-
ture, primarily corn and soybeans. Most of the
smaller tributaries and the upper river have
been channelized to facilitate drainage. Low
mill  dams have  been  constructed at various
locations, many of which are currently in a
state of disrepair except  in Logansport. That
dam severely restricted the movement  of
Wabash  River fishes into the Eel River and
facilitated evaluations of impacts.

Materials and Methods
The study included  a) a reconnaisance float
trip  of the entire river,   b) sampling each
station twice  by electrofishing,  c)  sampling
most of these same stations once by seining,
and d)  a habitat  survey (HEP) at each station.
Secchi transparency and temperature were
routinely  measured  on  each  occasion.  In
addition,  synoptic  short-term  profiles   of
turbidity, temperature, and dissolved oxygen
concentration  were  determined  on  three
separate dates.

Single  stations were located on lower Twelve
Mile, Paw Paw, Squirrel, Beargrass, and Sugar
Creeks, and also upstream and downstream of
Columbia City on Blue River. The remaining 16
stations were  located on  the matnstem of the
Eel River. A few mainstem stations (Taylor's
2B, 2, and 3) and Squirrel Creek  were not
seined  because  of   inappropriate   seining

Seining was  conducted  with a 30-foot  by
4-foot seine having 3/16 inch mesh weighed
down  by a  heavy steel chain tied to  the
bottom. This method was very effective at
capturing darters and minnows. Three seining
passes  along   20   meters   of   shoreline
constituted each seine sample.

Electrofishing utilized  a  Safari Bushman 300
backpack shocker carried in a canoe or while
wading, depending on place and depth.  Each
electrofishing sample was about 20 minutes in
duration along approximately 400 meters of
shoreline.   This  method  was  effective  in
capturing  larger fish  such as  redhorse and
suckers and species which prefer nearshore
cover such as sunfish and bass.

All captured fish were  identified to species,
weighed  and measured, then released  back
into the river. Those fish not easily identified in
the  field  were  preserved  in  formalin  and
returned  to the laboratory for identification
(Trautman 1981).

Fish data were analyzed using the Iwb and the
IBI. The 1990 Iwb values were based upon the
average of two electrofishing catches at each
station.  The  rationale  of this  community
parameter is presented  by Gammon  (1980),
who recommended multiple collections at each

The Iwb was calculated as:
Iwb =  0.5 In  N
+  0.5 In W  + Div.no. +
Where, N = number of fish captured per km;
W  = weight in kg of fish captured per km;
Div.no.  =  Shannon  diversity  based  on
numbers; Div.wt.  = Shannon diversity based
on  weight.

The IBI methodology has  been  thoroughly
discussed by Karr (1981 and 1987), Karr et al.
(1986 and  1987), and Angermeier and Karr
(1986). Regional applications are summarized
by  Miller et al. (1988).

                                                          Agricultural Impacts in the Eel River
Table 1. Scoring criteria used to determine IBI
for Eel River fish collections.
1 (worst)
3    5(best)
Fish species (total)  0-9   10-19
Darter species      0-1    2-3
Sunfish Species     0-1    2-3
Sucker Species     0-1    2-3
Intolerant Species   0-1    2-3
No. Individuals     0-100 101 -200 & 201
Percent individuals as:
  Green sunfish   11-100
  Omnivores      45-100
  Insect, cyprinids  0-20
  Top carnivores    0-2
  Hybrids          4-10
  Diseased         6-10
           6-10   0-5
          21-44  0-20
           3-10  £11
           2-3    0-1
           2-5    0-1
The original criteria for determining IBI (Karr,
et. al., 1987) were modified slightly for the Eel
River  (Table 1). The scaled metrics are those
used  in studies  of  the  Sugar Creek system
(Gammon et al.  1990a) and an  agricultural
analysis  of  several streams  in west-central
Indiana (Gammon et al. 1990b). They differ in
some details from the criteria used in  other
studies. The 1990 IBI values were based upon
the combined catches from electrofishing and
seining. The IBIs calculated on data from earlier
Eel  River studies may  be influenced to an
unknown  degree by the somewhat different
methodologies used to  collect fish.  Taylor
(1972) used a combination of electrofishing
and  rotenone,  while  Braun  and Robertson
(1982) used more intensive electrofishing. We
have   elected  to  use  the  same  criteria
regardless of stream order.

Habitat was quantitatively evaluated at each
mainstem collecting site, except for the most
downstream site near the Logansport dam and
Taylor's site 1, using a habitat evaluation pro-
Table 2. Habitat assessment scoring criteria

Parameter         Excellent Good Fair Poor

 Substrate and Instream Cover
1. Substrate/cover  16-20 11-156-10  0-5
2. Embeddedness   16-20 11-156-10  0-5
3. Water velocity   16-20 11-156-10  0-5

 Channel Morphology
4. Channel Alter    12-158-11  4-7  0-3
5. Scour/Deposition 12-15  8-11  4-7  0-3
6. Pool/Riffle Ratio  12-158-11  4-7  0-3

 Riparian and Bank Structure
7. Bank stability     9-10   6-8  3-5  0-2
8. Bank vegetation   9-10   6-8  3-5  0-2
9. Bank cover       9-10   6-8  3-5  0-2
                                 cedure (HEP; Plafkin et al. 1989) adapted from
                                 Platts et al. (1987).  HEP quantifies 9 habitat
                                 characteristics summarized in Table 2.   The
                                 total score for each site was based upon data
                                 from  10 transects at each site spaced 25, 50,
                                 or 100 feet apart.

                                 In   addition,  several   other   physical
                                 measurements were taken  whenever  fish
                                 collections   were   made   during   special
                                 longitudinal  surveys. Stream turbidity  was
                                 measured with a secchi disc or a Minispec20
                                 nephelometer.   Water  temperature   and
                                 dissolved oxygen were measured using a YSI
                                 meter. Water velocity was measured with a
                                 Gurley  pygmy  meter.  ALI  distances  were
                                 measured optically using a Leitz rangefinder.

                                 Estimates  of the amount of  woodland  were
                                 based on  conventional  analyses of enlarged
                                 LandSat infrared photographs taken on May 2,
                                 1981. These were  obtained  from the  U.S.

Gammon and Gammon
Geological Survey (ESIC), EROS Data Center,
Sioux Falls, SD.

The drainage area perimeter was determined
using  topographic maps  of  tributaries. This
scaled map  was  superimposed  over  the
infrared photographs on a light table. Plots of
land with permanent tree cover were outlined
on the topographic map.

Using  a light table, the marked topographic
map was  traced onto a fine grid. Individual
grids  with more than  50%  woodland was
calculated. Land use in a few tributaries was
not   determined  because  of  insufficient
coverage of LandSat infrared photographs.

A total of 6,635 fish comprising 46 species
were  captured by electrofishing and seining.
Forty species and 4154 individuals (63%) were
taken by seining alone.  Electrofishing catches
also   yielded 40 species,  but only  2481
individuals or 37% of the total.

Bluntnose minnow (Pimeohales notatus) was
very common comprising 40.9% of  the total
number seined,  while sand shiner (Notroois
stramineusl.  spotfin  shiner  (N. soilopterus).
striped shiner (N. chrvsocephalus).  silverjaw
minnow (Ericvmba buccata). and creek chub
(Semotilus atromaculatusl together contributed
another 37%.

The  electrofishing catch  was more  evenly
distributed with common shiner (N. cornutus)
and white sucker (Catostomus commersoni)
each  contributing about  15% to the  catch.
Substantial numbers of the following were also
collected:  creek chub  (9.3%), bluntnose
minnow   (9%),  rock   bass  (Ambloolites
ruoestris:  7.4%), and  northern hog  sucker
(Hvpentilium nioricans:  7.1%).

Smallmouth  bass  (Microoterus  dolomeiui)
adults and subadults were mostly found in the
lower 50 miles of the Eel River and only in Paw
Paw and Twelve  Mile Creeks.  Catch  rates
were higher in the lower 30 miles of river and
attenuated from RM 30 to RM 51.7. Three of
12 smallmouth bass 250 mm and longer were
fin clipped, indicating that they were stocked
fish.  Two   of  these  were  collected  by
electrofishing at RM 37.8(1) near  Roann and
the other at RM 27.3(68) near Chili. Young-of-
the-year smallmouth bass were taken only in
the extreme lower part of the Eel River and in
Paw Paw and Twelve Mile Creeks.

Largemouth  bass (M. salmoides) formed a
minor component of the catch. Fair numbers of
small  spotted  bass  (M-  punctulatusl  were
scattered throughout the mainstem Eel and in
Paw Paw and Twelve Mile Creeks. This species
had not been present since they could easily
have  been misidentified as small  largemouth
bass.  Spotted bass young-of-the-year  were
found even in the otherwise poorer habitat of
the upper 30 miles above South Whitley. This
species has been shown to be tolerant of high
turbidity and sedimentation (Gammon 1970).

Rock  bass was taken at all stations except
Squirrel  Creek.  Longear  sunfish (Lepomis
meaalotis) were most common at the upper
mainstem stations and in the Blue River and
were sporadic in the lower river. Green sunfish
(L-  cvanellus) also occurred at most sites, but
was more abundant in the upper mainstem and
in  the Blue River. Substantial numbers of
bluegill (L. macrochirusl were also  taken more
regularly in the upper mainstem Eel from RM
63.5 to RM 80.

The most abundant catostomid  was  white
sucker with greatest  numbers  in the  upper
mainstem from RM 63.5 to RM  80 and  in the
Blue River, Sugar Creek, and Beargrass Creek.
They were uncommon in the lower 60 miles of
the mainstem. Northern hogsucker was widely
distributed throughout the mainstem and most
tributaries.   Spotted  sucker   (Minvtrema
melanops) was found in good numbers only in
the pool above the Logansport dam.

                                                            Agricultural Impacts in the Eel River











     IBI and Iwbxtt
a   a
       0  10   20  30  40  50  60  70  80

                   River Mil*
  IBI baaad upon Sugar Craak erllarla and
  Inoludn aakilng data.
  Mod. Iwb daMaa 'lotoraitt' apaatoa
                                                D  O
                                                        0  10  20   30  40  50  60  70   80

                                                                    River Mil*
                                                          ••°- 1972IBI  -*- 1982IBI  -&- 1990IBI
                                                   Baa«d upon Sugar Craak erltarla.
Fig. 1. IBI, Iwb, and modified Iwb values for
1990 Eel River fish communities.
                       Fig. 2. IBI profiles of mainstem Eel River fish
                       communities for 1972, 1982, and 1990.
Golden redhorse (Moxostoma ervthrurum, was
the  most  common  of the three redhorse
species, but it was not all  that abundant. It
was absent between RM 56.5-80, as well as,
from all tributaries including Blue River. Black
redhorse (Moxostoma duauesnei) was almost
as  common  as golden  redhorse, but  was
mostly restricted to the lower 30 miles of the
mainstem.   Greater   redhorse   (Moxostoma
valenciennesil  is  a rare species throughout
Indiana and most of its range,  but a  healthy
population  thrives in the Eel River system. It
was particularly abundant  in the  lower 20
miles of river,  but was also found in Paw Paw
and Squirrel Creeks.

The distribution of smaller species of minnows
and darters is best illustrated by the seining
catches.    Bluntnose   minnow   (Pimeohales
notatus) was the dominant species, occurring
throughout  the mainstem and  tributaries.
Common shiner was  even more frequently
encountered by electrofishing and was also
                       widely distributed throughout the Eel River

                       Spotf in shiner and sand shiner mostly occurred
                       in the lower 50 miles of the mainstem. Creek
                       chub was common only  in  the tributaries.
                       Redfin  shiner  (Notropis   umbratilus)   and
                       rosyface shiner (Notroois rubellus) were most
                       common in the lower river, but also occurred in
                       Sugar  and Twelve Mile Creeks.  River chub
                       (Nocomis micropoaonl was regularly taken by
                       seine and electrofishing mostly downriver from
                       RM 65. A few bigeye chub (Hyboosis amblopsl
                       were also present in the lower river.

                       Among  the  darters,   only  johnny   darter
                       (Etheostoma  nigrum)   was  common  and
                       widespread.   Blackside  darter   (Percina
                       maculata). greenside darter (E. blennioides) and
                       eastern sand  darter (Ammocrvpta pellucida)
                       were found  only  in the lower river.  Dusky
                       darter  (P. sciera) was taken  only from upper
                       Eel River (RM 88.0) and  Beargrass Creek.

 Gammon and Gammon
110 •
  HEP tcor*
   0  B  12  18 24 30 36  42 48 54 BO 66  72 78
                    - IBI IcorM   * TributariM
 Fig. 3. IBI and HEP values for the mainstem Eel
 River and tributaries.
      10   20   30   40   SO   60   70  10  90  100
CIS Tributaries
Fantail darter (E. flabellare) was collected only
at RM 63.5). Mottled sculpin (Cottus bairdi)
was taken only  at RM 56.5 and RM 63.5).

Important  community   index   values   are
summarized in  Table 3. IBI values were also
calculated on less extensive data sets provided
by Braun et al. 1984,1986) on five collections
of fish from each of three stations;  2B (RM
3.3), 3B (RM 8.3) and 3 (RM 46.4) during the
years 1984 and 1985. The mean IBI values at
stations 2B, 3B, and 3 were 39.6, 42.0, and
43.6  in  1984  and  43.2,  41.2,  and  42.9 in
1985, respectively.

The IBI  and Iwb profiles  for the Eel River
mainstem are shown in Figure 1. An additional
modified Iwb is  also  shown,  wherein four
tolerant  species   were  deleted prior  to
calculation, carp, bluntnose  minnow,  creek
chub, and green sunfish.

All three profiles indicate somewhat depressed
fish communities  in the lower river, probably
because of the ponding effect  of the dam,
followed by relatively good communities from
RM 8 to RM 25. From RM 30 to RM 80 there
is considerable  variation from place to place,
but the communities are generally depressed,
especially at RM 70.

In Figure 2, the  1990 IBI profile is repeated and
compared  to IBI  profiles based  on  Taylor's
(1972)  and Braun  et  al.'s (1982) series of
collections. The  1990  fish communites are
clearly much better than they were in 1982.
However,   both    profiles  indicate   better
communities in the lower river  than in the
upper river. In  1972 there was less difference
in  the  lower  mainstem but equal  variation
between stations.
  Fig. 4. Turbidity (NTU) for the mainstem Eel
  River and its tributaries on July 16 and 17,
 Habitat Evaluatipn
 Habitat scores  were generally  lower in the
 upper part of the watershed and higher

                                                               Agricultural Impacts in the Eel River
      0   5  10  15  20   25  30  35  40
                   % woodland
   Turbidity deteratkiee' en 7-M ( IT. 1MO.

Figure  5.  Turbidity  (NTU)   of  Eel  River
tributaries in relation to  percent woodland in
their drainage basins.
     Rainbow darter r-
     Roayface ahmer IE
       Silver ahiner i
     Blackaide darter Si
    Qreenaide darter
        Rock baaa
    Smallmouth baaa
     Longear aunfiah ai
          Sculpin •
         Stonecat F
  Orangetnreat darter p
    bluebreaat darter p
                                                                              40     60
                                                                              % frequency
                                                                          EH3 1940-41
1940-41: 6 aialnetaei • 4 trlbutarlee
19*0: 12 aialnalen • e tributaries
  Figure  6.  Frequency of occurrence of some
  species collected by seining in 1990 compared
  to 1940-41.
downstream. Upstream from South Whitley,
habitat features were uniformly low in quality
and   homogeneous   because   of   past
channelization and recent deforestation of both
banks.  Habitat  scores  of  tributaries  were
generally  higher than the  mainstem reaches
into  which   they  flowed  (Figure  3).  An
excpetion  was Paw Paw  Creek  which  was
somewhat lower.

Turbidity and Landuses
Turbidity  determination  during the  synoptic
surveys on July  16-17, 1990 are  portrayed
graphically in Figure 4. Scattered showers fell
throughout northern  Indiana during  the week
previous to the turbidity determinations. It is
not known to what extent these  results may
be affected by differential  rainfall. Tributaries
which were distinctly more turbid than others
included Squirrel Creek, Otter Creek, Simonton
Creek,  Hurricane  Creek,  Blue River, Solon
Ditch, and Johnson Ditch. A bridge was under
construction in the Simonton watershed, but
       301 I  I I I  I I  I I  I I I  I I  I I  I I I  I I  I I  I
             1975    1980    1985    1990     1995

              •If HM0**n Crfe
          0  Eel River

          +  Sugar Creek
O Big Raccoon crk.
     IBI Meed uvea Sneer Craek criteria.
   Figure 7. Changes in the IBI values for the Eel
   River compared  to Big  Raccoon  Creek and
   Sugar Creek.

Gammon and Gammon
Table 3. Fish community indicies for Eel River stations.
Station 1 990
. Soec. elec.

                                    Tributary Stations
          Twelvemile Creek
          Paw Paw Creek
          Squirrel Creek
          Beargrass Creek
          Sugar Creek
          Blue River - upstream from Columbia City
          Blue River - downstream from Columbia City
animals were also pastured in the stream and
some corn fields extended to stream banks.

The turbidity of mainstem water was high in
the upper river mainly because of highly turbid
Johnson ditch. The water cleared considerably
after passing through two mainstem gravel pits
at RM  84 then again  became  progressively
more turbid as it flowed downstream.

During this same period the turbidity gradually
increased in the mainstem from the upper river
to  the  lower   river, although there  were
localized  sharp   increases   in   turbidity
downstream from  both Johnson ditch and
South Whitley. Earlier in the summer (June 12,
1990)  when  water  levels  were  higher the
turbidity (NTU) was 45 in the lower 60 km (40
mi) of river and between 46-48 in the  upper
river. In some streams lateral erosion can be a
major source of sediment and turbidity, but
scoured banks were a very limited component
of  the  lower  portions  of  the  Eel  River
mainstem. However, they were highly evident
in the channelized upper portions. The  entire
upper 50 km (33 mi) of the Eel River has been
stripped of its trees and bushes along both
banks. During this  study the trimmings had

                                                           Agricultural Impactsinthe Eel River
been removed from the river and were piled
along the shore for burning.

Woodlands were readily determined from the
infrared  photographs,  but  other  kinds  of
permanent vegetation  such  as brushlands,
pastures,   and  winter   wheat   were
indistinguishable from one another, estimates
of woodland ranged from only 7.0% in the
Beargrass Creek watershed to 40.9% in the
Weesaw Creek watershed. There was a greater
percentage of landuse in  agriculture south of
the mainstem and  in the upper two-thirds of
the  Eel  River watershed than  north  of the
mainstem and in the lower third. There was an
inverse relationship between the percentage of
tributary watersheds  in  woodland  and  the
measured turbidity (Figure 5).


The  Eel River in 1990 was found to- support
fairly diverse fish  communities  throughout
most of  the  watershed,  although the  upper
reaches  had  depressed  populations  and
reduced numbers of species. Many species of
juvenile fish were caught, with larger numbers
at stations 1B and  Twelve Mile Creek. This is
an  indication that reproduction for  many
species was successful during the past couple
of years.

Several usually  common species which Braun
and  Robertson  (1982)  did  not collect were
found in  good numbers in 1990: river chub,
bigeye  chub,  several   species  of  shiners
including silver shiner, spotfin shiner, rosyface
shiner, redf in shiner, and blackside and johnny

Some species present in 1972 were found only
rarely or not at all in  1990. These species
included  mottled  sculpin,  blacknose  dace,
unidentified  madtom species,  suckermouth
minnow,  largemouth bass, and carp.

It is difficult to evaluate long-term changes in
abundance of  any single  species of fish
because   of  the   different  collecting
methodologies. The comprehensive study of
Gerking (1945) used the seine as the primary
collecting gear and our effort  in 1990 was
comparable.   Gerking  collected  from  five
mainstem  sites and  four tributaries.   We
collected  from 12 mainstem sites  and six

A  comparison  of  percent  frequency  of
occurrence from these studies indicates rather
drastic   reductions   for   many   species
populations of sediment sensitive fish (Figure
6). Rock bass, johnny darter, and eastern sand
darter appear to be distributed much as they
were 50 years ago. However, many species
have suffered  drastic  declines   including
rainbow  darter (E. caeruleum),  orangethroat
darter (E. soectabile). and bluebreast darter (E.
camuruml  which  may  have  been  totally
eliminated from the river.

Changes over time in populations of clams and
mussels  parallel  those  of fish.  Henschen
(1988) concluded that while the Eel River once
supported  a  diversity  of  mussel  species
throughout its length, its  currently  reduced
population is mostly confined to the lower river
in Cass and Miami Counties.

Changes in the Fish Community over Time
The IBI offers one way of addressing questions
about how the overall  fish community has
changed over time and how it compares to fish
communities in other  streams. The  mean IBI
values for the Eel River  mainstem  stations
declined from 40.7 in  1972 to 36.9 in 1982.
This increased substantially to 43.1  in 1990.
The  IBI values estimated from  data of the
studies of Braun et al (1984,1986) and Braun
(1990) generally corresponded  to  improving
trend noted in  the 1980's.

The overall Eel River fish community appears
to have improved  rapidly from  the  degraded
community found in 1982.

Gammon and Gammon
1 1 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1 — 1— 1 — 1 — 1 — 1 — 1 —
o "*"
O D 1990 O
0 + 1988 ^ '
H" Toflfi

                                                           Agricultural Impacts in the Eel River
Nevertheless, a series of years with high runoff
and  increased non-point source pollution can
depress fish communities to equally low levels.

Other factors which may modify the recovery
of fish  populations in the Eel River  system
include the absence of high quality tributaries
to serve as refugia for sensitive species during
unfavorable years and the dam blockage at
Logansport which reduces recolonization by
species from the Wabash River.

The Potential Influence of Habitat and Turbiditv
Much of the upper Eel River is characterized by
low HEP values, e.g. channelized stream beds,
poor riffle/pool development,  and a  lack of
instream structure. In addition, riparian trees
have been removed recently from many older
previously channelized sections of the  river.

The  bottom substrate usually  included much
fine  sediment,  as  indicated  by  the   low
embeddedness scores for almost all  of the
mainstem  stations  and   most  tributaries.
Turbidity  was high  for  virtually the entire
summer. At Roann we saw a layer of mud  two
centimeters deep  on top of a flat boulder after
higher water had subsided in a pool.

The lower 48 km (30 mi) of Eel River contained
much  better   habitat than  the upstream
reaches.  Beds of water  willow  (Dianthera)
were mostly limited to the lower 64 km (40
mi) of the mainstem Eel. This section also had
fairly good  riparian  protection  and good
instream habitat.

Habitat  in the tributaries generally  scored
higher than the mainstem. Twelve Mile Creek,
with  26.5%  of  its watershed  in  forest,
contained the best habitat, followed by Squirrel
Creek.  The  Blue  River is approximately  the
same size as  the Eel  River where  the  two
streams converge. With only 11 % permanent
vegetation cover, its turbidity readings were
among the highest recorded.  Fish from  this
stream, and Paw Paw Creek, were commonly
infected with blackspot disease (Simon 1989).
Potential Negative Effects from  Point-Source
Agricultural  point-source pollution in  Indiana
often occurs because of accidents or careless
handling of animal wastes and farm chemicals.
Spilled  materials, animal  wastes applied to
fields, and  the contents of  waste  holding
lagoons may  be flushed into  ditches  and
streams following  rain storms.  Fish  kills
reported  to  the   Indiana   Department  of
Environmental Management (IDEM) since 1969
include  five  incidents on Paw Paw Creek and
single kills on Twelve Mile, Pony, Beargrass,
and Clear  Creeks.

There were 39 additional reports of  spilled
materials  which  are  not  known  to  have
resulted in  fish kills,  but which may have
exerted sublethal  damage.   Most  of  the
materials  were fertilizer and  animal  wastes,
which include wastes generated by chickem,
turkey,  veal, and swine rearing operations.

All of the known causes of ifsh kills and most
of the  spills  reported  within the Eel River
watershed are agriculturally based. The actual
number of fish kills and spills is unknown, but
would certainly far exceed  the number of
reported cases.

In the decade  following passage of the Clean
Water Act of  1972, it was estimated  that
municipal  BOD loads decreased  by 46%  and
industrial BOD loads decreased at least 71%
(U.S. Environmental Protection Agency, 1982).
Most of the communities in  the area have
improved waste treatment and reduced BOD
concentrations by   at least  50%.   Some
previously unsewered communities now have
a central treatment system. It is likely that any
negative influences  from these point sources
are masked by the  magnitude  of non-point
source impacts.

Weather and Nonooint Source Pollution
Unlike point source pollution, nonpoint sources
of influence such as occurs from plowed fields
are most  severe during storm  events.  The

Gammon and Gammon
discharge of rivers is roughly proportional to
the  amount  of  rainfall,  hence,  non-point
sources are most severe when rainfall is great
and river discharge is high. Conversely, non-
point sources are reduced during periods of dry
weather.  Fish  populations  are  negatively
affected  by  non-point  sources  during  the
reproductive  period  and  in  the  months
immediately  after  hatching,  spring   and

From October 1974 through September 1980
the U.S. Geological Survey Water Resources
Division determined daily sediment loads for
the  Eel River near Logansport  (Anonymous
1974 through 1980). Data from the Eel River
and  rivers throughout Indiana is analyzed and
discussed by Crawford and  Mansue (1988).
They estimated that for the Eel River the mean
annual suspended sediment yield  was  178
tons/square mile/year and the flow-weighted
mean  annual   suspended   sediment
concentration  was  89 mg/l  (median  = 53
mg/l). These values are high for the northern
moraine/lake  portion   of  Indiana  which
Crawford  and  Mansue found to  have the
lowest sediment yield. Only that part of the Eel
River watershed north of the mainstem resides
within the moraine area. The portion  of the
watershed situated south of the  mainstem is
located in the Tipton  Till Plain  where  both
parameters were generally much larger.

Monthly data from May through August for the
years  1974 through  1980 was   used for
regression   analysis  of  suspended  solids
concentration on discharge. The regression
equation obtained was then  used to estimate
the  suspended solids  concentration for the
months May through  August for  the years
following  1980 (Figure 9). Suspended solids
concentrations were highest during May and
June when relatively high discharges occurred
during half  of the years since  1974. "Wet"
summers  of relatively  high suspended solids
concentration include the years 1974, 1975,
1980,1981,1982, and 1986. "Dry" summers
when Eef River water was  relatively clear
include the period from 1976 through 1979,
1983 through 1985, and 1987 through 1988.

During  "dry"  summers the effects of point
sources of pollution such as from population
centers  would  theoretically  increase, but
nonpoint source pollution should be less than
normal. For streams influenced mostly by NFS
the fish communities following a sequence of
"dry" summers should improve. The Eel River
fish communities did improve somewhat, but
less than might have been expected compared
to fish communities in Big Raccoon Creek.

The 1990 fish communities may be as good as
the Eel  River is able  to support considering
present land use. The summers of 1989 and
1990  were   relatively  "wet".  Therefore,
reproductive success and survivorship through
the first year of life would  be expected to be
lower than normal. It is likely that the 1991
fish communities will be poorer than they were
in 1990 and the prognosis for improvement in
the future is bleak unless changes in land-use
are implemented.

The Eel River is essentially a linear stream. Its
drainage basin is  long  and  narrow and its
tributaries are generally small first and second
order streams.  Improving  landuse in these
tributaries will be necessary in order to improve
the  mainstem  of  the Eel River. Thorough
surveys of all tributary watersheds should be
conducted using both Geographic Information
System (GIS) technology and  ground study.

Twelve Mile  Creek,  Paw  Paw  Creek, and,
possibly,  Squirrel  Creek appear  to be less
influenced by agriculture than other tributaries.
These  tributaries  may act  as  refugia  for
sensitive species during periods of stress and
serve as species reservoirs to  replenish the
mainstem during more benevolent times. They
should receive special attention to ensure that:
 a) the streamside  riparian  buffer  zone  is

                                                           Agricultural Impacts in the Eel River
 maintained, b) tilled fields do not impinge on
 the stream itself, c)  hogs and cattle are not
 pastured directly in the streams, d) appropriate
 forms of conservation tillage are encouraged,
 e) animal wastes are  properly disposed.

 Several other tributaries appear to be more
 environmentally  degraded than others. Otter
 Creek, Simonton Creek, Hurricane Creek, Blue
 River,  Solon  Ditch,  and  Johnson  Ditch
 delivered higher than average sediment loads
 to the Eel River during the survey of July 16
 and 17, 1990. While this survey is only a brief
 "snapshot" in time, it nevertheless suggests
 that these streams  may  have greater  than
 average  negative impacts on the  Eel River
 system. They should also receive the same
 items of attention listed above.

 Streams in the upper watershed are referred to
 and used as drainage ditches. Nevertheless,
 these streams  are permanent "creeks"  and
 should  support  normal  aquatic  life.  Their
 rehabilitation  could   contribute   positively
 toward   the   improvement  of   the  lower
 mainstem.  The  creation  of a  "green belt"
 riparian corridor would also contribute toward
 a greater ecological diversification.

 The Eel River study was funded by the Indiana
 Department of Environmental  Management.
 Undergraduate students D. Wallace and M.
 Giesecke acted  as the  primary  field crew.
 Assistance in collecting fish was provided by
 J. Riggs, C. Hansen, J. Hecko, and N. Masten,
 who were supported by grants from Eli Lilly
 and  Company   and  PSI  Energy.  The  Big
 Raccoon Creek data was collected through a
 grant  from Heritage  Environmental  Services,

The  study benefitted  significantly through
consultation with J.  Ray, C.L.  Bridges, and
R.J. Gammon of the  Indiana Department of
Environmental Management  (IDEM)  and  E.
Braun and T.  Stefanavage  of the Indiana
Department of Natural Resources (IDNR). This
contribution is dedicated to the  successful
completion of the first year of life of Robert
Wayne Pitman-Gammon, who is considerably
more alert, lively, and curious than he was a
year ago and to his parents.

Literature Cited

Anonymous.   1975-1981.  Water resources
data for Indiana. U.S. Geological Survey Water-
Data Reports for Water Years 1974 through

Braun, E.R. 1990. A survey of the fishes of the
Eel  River in Wabash  and  Miami Counties,
Indiana  1989. Indiana Department of  Natural
Resources, Division of Fish and Wildlife, 607
State Office  Building,  Indianapolis,  Indiana
46204.  30 pp. mimeo.

Braun, E.R. and R. Robertson. 1982. Eel River
watershed fisheries investigation 1982. Indiana
Department of Natural Resources,  Division of
Fish and Wildlife, 607 State Office Building,
Indianapolis, Indiana 46204. 60 pp. mimeo.

Braun, E.R., R. Robertson, and T. Stefanavage.
1984. Evaluation of smallmouth bass stocked
in the Eel River 1984 progress report.  Indiana
Department of Natural Resources,  Division of
Fish and Wildlife, 607 State Office Building,
Indianapolis, Indiana 46204. 47 pp. mimeo.

Braun, E.R., R. Robertson, and T. Stefanavage.
1986. Evaluation of smallmouth bass stocked
in the Eel River 1985 progress report.  Indiana
Department of Natural Resources,  Division of
Fish and Wildlife, 607 State Office Building,
Indianapolis, Indiana 46204. 84 pp. mimeo.

Crawford,  C.G.  and  L.J.  Mansue.  1988.
Suspended sediment characteristics of Indiana
streams, 1952-84. U.S. Geological  Survey,
Open File Report 87-527. 79 pp.

Gammon, C.W. and J.R. Gammon.  1990. Fish
communities and habitat of the Eel River in
relation to agriculture. A report for the  Indiana

Gammon and Gammon
Department  of  Environmental Management,
Office of Water Management, Indianapolis, IN

Gammon, J.R. 1970. The effect of inorganic
sediment on  stream biota. Water Pollution
Control Research Series 18050DWC 12/70:1-

Gammon, J.R. 1980. The use of community
parameters derived from electrofishing catches
of river fish  as indicators of environemntal
quality,  pp.  335-363 in  Seminar on Water
Quality  Management   Tradeoffs.  U.S.
Environmental Protection Agency, Washington,
D.C. EPA 905/9-80-009.

Gammon, J.R. and J.R. Riggs.  1983. The fish
communities of Big Vermilion River and Sugar
Creek.  Proceedings   Indiana  Academy   of
Science 92: 183-190.

Gammon,  J.R.,  C.W.  Gammon,  and M.K.
Schmid. 1990.  Land  use influence  on  fish
communities  in  central Indiana streams,  pp.
111-120. in W.S. Davis (editor). Proceedings
1990 Midwest  Pollution Control Biologists
Meeting.  U.S.   Environmental   Protection
Agency, Region V, Environmental Sciences
Division, Chicago, IL. EPA 905/9-90-005.

Gammon,  J.R.,  C.W. Gammon, and  C.E.
Tucker. 1990. The fish communities of SUgar
Creek.  Proceedings   Indiana   Academy  of
Science 99: in press.

Gammon, J.R. 1990. The fish communities of
Big Raccoon Creek 1981 -1989. A report for
Heritage  Environmental   Services,  One
Environmental Plaza, 7901 West Morris Street,
Indianapolis, IN 46231. 120 pp.

Henschen, M. 1988. The freshwater mussels
(Unioinidae)   of the  Eel  River  of  northern
Indiana. Indiana DNR, Division  of Fish  and
Wildlife,   607   State  Office   Building,
Indianapolis, IN  46204. 73 pp. mimeo.
Hoggart, R.E. 1975. Drainage areas of Indiana
streams.   U.S.   Geological  Survey,  Water
Resources Division, Indianapolis, IN. 231 pp.

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

Karr, J.R.  1987. Biological  monitoring  and
environmental  assessment:   a  conceptual
framework. Env. Management 11: 249-256.

Karr, J.R., K.D.  Fausch, P.L. Angermeier, P.R.
Yant, and  I.J.  Schlosser. 1986.  Assessing
biological integrity in running waters: a method
and its rationale. Illinois Natural History Survey
Special Publications  5, Urbana.

Karr, J.R., P.R. Yant, K.D. Fausch, and I.J.
Schlosser.   1987.  Spatial   and   temporal
variability  of the  index of biotic integrity  in
three  midwestern   streams.   Transactions
American Fisheries Society 116: 1-11.

Miller, D.L., P.M. Leonard, R.M. Hughes,  J.R.
Karr,  P.B.   Moyle,  L.H.   Schrader,  B.A.
Thompson,  R.A.  Daniels,  K.D.  Fausch,   A.
Fitzhugh,  J.R. Gammon,  D.B. Halliwell,  P.L.
Angermeier,  and  DJ. Orth.  1988.  Regional
applications of an index of biotic integrity for
use in water resource management.  Fisheries
13: 12-20.

Plafkin, J.L., M.T. Barbour, K.D. Porter,  S.K.
Gross,  and  R.M.  Hughes.  1989.  Rapid
Bioassessment  Protocols  for  use in streams
and  rivers:  benthic macroinvertebrates  and
fish. EPA 444/4-89-001.

Platts, W.S., C. Armour, G.D. Booth,  M.
Bryant,  J.L.  Bufford,  P.  Cuplin, S. Jensen,
G.W.  Lienkaemper,  G.W.   Minshall,  S.B.
Monsen, R.L. Nelson, J.R. Sedell,  and  J.S.
Tuhy. 1987. Methods for evaluating riparian
habitats  with  applications to management.
U.S. Department of Agriculture, Forest Service,
Intermountain  Research  Station,   General
Technical Report INT-221.177 pp.

                                                           Agricultural Impacts in the Eel River
Simon, T.P. 1989. Biological Survey  of the
instream fish and water quality  evaluation of
Wayne Reclamation  and Recycling, Whitley
County, Indiana. U.S. Environmental Protection
Agency, Central Regional Laboratory, Chicago,
IL. 60605. 19 pp. mimeo.

Taylor, M. 1972. Eel River watershed fisheries
investigations report 1972. Indiana Department
of Natural  Resources,  Division  of Fish and
Wildlife,  607   State  Office   Building,
Indianapolis, Indiana 46204. 65  pp. mimeo.

Trautman, M.B.  1981.  The fishes of  Ohio.
Ohio State University Press, Columbus. 782

U.S. Environmental Protection Agency. 1982.
National water quality inventory: 1982 report
to Congress. Washington, D.C. 63 pp.

Selenastrum Algal Growth Test: Culturing and Test Protocol at the Illinois EPA

Michael J. Charles and Greg Searle
Illinois EPA
Office of Ecotoxicology, if 31
2200 Churchill Road
P.O.Box 19276
Springfield. IL 62794-9276

The availability of quality test organisms is of fundamental concern in conducting  any regularly
scheduled biomonitoring activities. The Selenastrum algal assay requires a continuous supply of pure
log-phase algae. The most convenient means to meet this demand is through the establishment of
in-house cultures. Upon receipt of a pure Selenastrum "starter" culture from an outside source,
laboratory stock cultures are initiated. The algae is aseptically transferred to a series of culture flasks
containing synthetically prepared algal medium. Once pure algal cultures have been established, a
regime of routine cell transfers will provide the laboratory with a steady supply  of log-phase  algal
cells suitable for testing purposes. Back-up reserve cultures are stored on agar  slants and plates.
Testing of municipal and  industrial  effluents at the Illinois EPA using Selenastrum  algae follows
USEPA test  protocol. Through experience running the test and repeated attempts to get confident
results, the testing has been refined and the integrity of the analysis is ensured. Various techniques
are employed in the testing that serve to tighten the USEPA protocol and may be of interest to other
regulatory bioassay personnel.

Keywords: Selenastrum algae, log-phase, aseptically, agar slant, agar plate.
At the present time, the Illinois EPA (IEPA) is
the only state run bioassay laboratory in USEPA
Region V conducting the  Selenastrum  caori-
cornutum algal growth test. The algal growth
test is conducted in the Toxicity Testing Unit
(TTU) which is one of two units in the Office of
Ecotoxicology (OE). OE serves  as a support
laboratory for the various control  divisions
within the IEPA (air, land, water, and  public
water  supplies). IEPA uses USEPA protocol.
Short Term Methods for Estimating the Chronic
Toxicity  of Effluents and Receiving Waters to
Freshwater  Organisms  (USEPA 1989) as  a
guideline for  culturing  and conducting  tests.
TTU maintains a continuous supply of in-house
stock cultures for  use  in  bioassays. Various
techniques  are employed  in culturing  and
testing that serve to tighten USEPA protocol.

Establishing and Maintaining Selenastrum Stock
The Selenastrum algal  assay  calls for healthy
log-phase-growth  cells which are harvested
from resident in-house stock cultures. The TTU
laboratory maintains a  continuous supply of
these  log-phase  Selenastrum  cultures  in
quantities  sufficient  to  meet   all   testing
demands. The  culturing process is relatively
straight-forward  and consists of six  basic
components:   1)  general   culture  setup/
conditions; 2) algal nutrient culture medium; 3)
aseptic technique; 4) routine cell  transfers; 5)
back-up/reserve  cultures;  and,   6)  quality
assurance/quality control considerations.

1. General Culture Setup/Conditions.
The Selenastrum cultures are maintained in an
environmental chamber at 25  ±  1°C under a
continuous "cool-white" fluorescent illumination
of 400 ± 40 ft-c (4306 ± 431 lux). The algal
cells are  kept in a constant state of suspension
through  the use of a  mechanical shaker at
approximately  100 cpm  (cycles  per minute).
The culture flasks are  arranged on the shaker
table and allowed to incubate anywhere from
four to seven  days, depending  upon current
testing schedules. This incubation interval

                                                 IEPA Selenastrum Culturing and Test Protocol
Table   1.   Nutrient  Stock  Solutions   For
Maintaining Algal Stock Cultures (adapted from
USEPA 1989).
Stock    Compound
Amount dissolved
in 500 ml
Distilled H20

6.08 g
2.20 g
92.8 mg
208.0 mg
79.9 mg
0.714 mgb
3.63 mg°
0.006 mgd
1 50.0 mg
12.75 g
7.35 g
0.522 g
7.50 g
•ZnCI2 - Weigh out 164 mg and dilute to 100
ml. Add 1 ml of this solution to Stock #1.
bCoCI2-6H2O - Weigh  out 71.4 mg and dilute to
100 mL. Add 1 mL of this solution to Stock #1.
cNa2MoCy2H20 - Weigh out 36.6 mg and dilute
to 10 mL. Add 1 mL of this solution to  Stock
dCuCI2'2H2O - Weigh  out 60.0 mg and dilute to
1000 mL. Take 1 mL of this solution and dilute
to 10 mL. Take 1 mL  of the second dilution and
add to Stock #1.
provides plenty  of  viable,  log-phase
suitable for bioassay purposes.
2. Algal Nutrient Culture Medium.
Upon receipt of a Selenastrum "starter" culture
from an established  outside source, in-house
stock  cultures  are  initiated  by  aseptically
transferring a  portion of the cells  to freshly
prepared algal  nutrient medium. The  culture
medium consists of a mixture of various macro-
and micronutrients  prepared in  four separate
stock nutrient solutions using the reagent grade
chemicals listed in Table 1.

The nutrient medium is prepared by adding 1
mL of each of the four stock solutions, in order
as listed in Table 1, per liter of distilled water.
The  solution  is  mixed  well and  then  pH-
adjusted to 7.5 ± 0.1 by dropwise addition of
0.1 N NaOH   or  HCL, as appropriate.  The
medium is then immediately filtered through a
pre-washed 6.2 /jm pore diameter membrane at
a vacuum pressure of approximately 8 psi. The
algal nutrient   medium  is  then  ready to be
dispensed into  the  various culture flasks and
inoculated as needed. Any leftover portions of
the sterile  medium  may  be  stored  in  a
refrigerator at 4°C until needed. Care should be
taken, however, to  seal off the storage vessel
well  so  as to  prevent  loss  of  water  by
evaporation. Evaporation losses  will alter the
concentration  of macro-micronutrients in the
final medium, thus compromising its quality for
use in culturing purposes.

3. Aseptic Technique.
Extreme   care   is  exercised   to   prevent
contamination   of  the  cultures  by  other
microorganisms. All glassware products used in
the culturing process are  thoroughly cleaned,
sealed  with aluminum foil, and sterilized at
121°C in an autoclave. All  pipet tips used in
handling  the algal  cells  during routine  cell
transfer procedures are of the disposable type,
and they too are autoclaved at 121°C. The algal
nutrient culture medium is cold-sterilized before
use by  passing  it through  a 0.2  jjm  pore
diameter membrane filter,  as described above.
Despite these efforts, contamination problems
do occur from time to time. Contaminated
cultures are either discarded or used as food for
Ceriodaphnia cultures.

4. Routine Cell  Transfers.
To meet scheduled  testing demands, the TTU
laboratory maintains a  continuous  source of

Charles and Searle
log-phase Selenastrum cells. This is achieved
through a series of routine cell transfers from
existing stock cultures to various afiquots of
fresh algal nutrient medium.  An inoculum is
prepared from a four - to seven -  day stock
culture by concentrating the cells of the culture
through a centrif ugation process. The algal cell
concentrate is then diluted with distilled water
to provide an initial density of approximately
10,000 cells/mL in the culture flasks. A Coulter
Counter" (model ZM)  is utilized in cell density
determinations for both the stock cultures and
the final inoculum.

Once prepared, 1 mL of inoculum is aseptically
transferred to each of three 500 mL Erlenmeyer
culture flasks containing 250 mLs of fresh algal
medium each. After inoculation, the flasks are
situated in  the environmental chamber on a
mechanical shaker for incubation purposes. An
incubation period of four to seven days renders
plenty of  healthy, log-phase Selenastrum cells
ready for harvest and  use in testing  and/or
other  purposes.  Routine  cell transfers  are
carried out twice per week, with each transfer
staggered 3*4 days apart. This arrangement will
provide a continuous  supply of log-phase cells
suitable for biomonitoring  purposes.

The volume of stock cultures required depends
on the test  loads  involved  and  any other
targeted uses for the algae (ie., food source for
Ceriodaphnia   cultures,   etc.).   The   TTU
laboratory meets  all of  its current algae
demands by inoculating  1.5 - 2 L  of fresh
culture  medium weekly.

5. Back-Up/Reserve Cultures.
As  mentioned above,  contamination  of the
stock cultures seems to be inevitable from time
to time. It is therefore essential to have in place
some type of a back-up/reserve system for
storing  clean,  pure Selenastrum stocks  that
may be called upon  to rejuvenate "dirty" or
"fouled" cultures. The TTU laboratory meets
this objective through the use of a system of
agar slants and plates. The  agar  medium is
prepared  with the same stock nutrients, in the
same amounts, as the  standard  liquid algal
medium. The only difference is that the stock
nutrients are dissolved in a 1-2% BactoRAgar
solution. The agar nutrient medium is mixed up
in an AgarMatic^bench top agar sterilizer, which
in turn is linked up to a PourMatic™automatic
plate dispensal  system. Thus, the agar medium
is mixed, sterilized, and poured into plate form
all in one process. Any excess medium is then
hand-poured into test tubes for use as slants. A
large batch of plates and slants are poured all at
once, the bulk of which is  then  stored  in a
refrigerator at 4°C until needed.

At scheduled intervals of approximately  once a
month, several  fresh agar plates are "streaked"
with  Selenastrum cells  from  existing  stock
cultures. A 10 /j\ inoculating loop is used to
transfer the cells from the liquid stock cultures
to the agar, where they  are  streaked out into
quadrants on the plated medium. The plates are
then arranged on a rack situated in a partially
enclosed glass  box shelter in the environmental
chamber for incubation  purposes. The glass
box, along with rubber bands used to seal the
lids on the petri dishes, serves to break up the
airflow  patterns  of the  chamber around the
immediate  vicinity  of  the  plates, thereby
minimizing dessication problems of the  media.
The plated  cultures  need  air  exchange for
proper growth, but too much airflow will only
serve  to  dry  out  the plates  completely,
rendering them useless  for storage purposes.
An incubation period  of  1-2  weeks yields
several distinct Selenastrum colonies that may
then  be targeted for transfer to fresh liquid
nutrient medium, thereby  rejuvenating active
stock cultures.

Agar slants are also streaked up from time to
time  as needed,  but  serve  primarily  in  a
secondary backup role.  After incubation, the
slants displaying  healthy Selenastrum colonies
are pulled from the environmental chamber and
stored in a refrigerator at 48C for up to  several
months. In  this  way,  they may  serve  as  a
 "backup" to the backup  cultures.

                                                 IEPA Selenastrum Culturing and Test Protocol
 6.   Quality   Assurance/Quality   Control
 At  each cell transfer, the stock cultures are
 examined   microscopically   for   species
 verification purposes and to look for any signs
 of microbial contamination. This information,
 along with general observations on the overall
 condition of the  cells themselves,  is  then
 recorded  in  an algal culture  logbook.  This
 enables the TTU laboratory to keep a running
 history  of  culture activities and any special
 problems/ solutions encountered. Stock cultures
 are also subjected to monthly NaCI reference
 toxicant tests.  EC 50  point estimates  are
 calculated for each reference test, and these
 values are then plotted on  a standard reference
 toxicant control   chart   for quality   control
 purposes (Figure 1). These steps are taken to
 ensure  the  quality  and  suitability   of  the
 Selenastrum   stock  cultures   for   use  in
 biomonitoring activities.

 Selenastrum Algal Growth Test Protocol
 Samples received by TTU for algal bioassays
 consist of municipal and industrial effluents and
 their ambient receiving waters (upstream of the
 effluent outfall). For the purposes of the IEPA,
 Illinois is divided into seven regions. All regions,
 except Region 4 (Field Operations Services in
 Champaign, IL) and Region 5 (Field Operations
 Services in Springfield, IL), ship samples to OE
 via  bonded  courier (e.g.,  Emery Worldwide).
 Regions 4 and 5 hand deliver samples to OE. All
 samples are  received in  the laboratory  and
 testing started within 28 hours of sampling. A
 chain of custody is maintained by field  and
 laboratory personnel to ensure the samples are
 not tampered with.

 When samples arrive in TTU they are logged in,
 warmed  to the proper temperature,  aerated,
and  initial  water chemistries are  performed.
 Initial water chemistries consist of alkalinity,
hardness,  chlorine,    and   ammonia
determinations.   Measurement   of   these
parameters helps resolve the cause of toxicity.
When the  samples  have  been  warmed  and
aerated,  dilutions are poured. A 0.5  dilution
series is used. Temperature, pH, conductivity,
and  dissolved oxygen are measured  on each
dilution to determine if these parameters are
within  the  range  for   normal   growth  of
Selenastrum. A 250 ml portion of each dilution
is then poured off for the algal bioassay.

Each 250 mL dilution is enriched with 250 fjL
of each of the four nutrient  stock solutions
(with  EDTA).  To reduce the possibility  of
contamination  in the  algal bioassay, aseptic
techniques are  employed throughout the test.
All glassware is  washed  with non-phosphate
detergent and rinsed with tap  water, acetone,
hydrochloric  acid,  tap  water,  and  distilled
water. Glassware and pipet tips are autoclaved
at 121°C.

Each dilution (with nutrients) is filtered through
a  0.2  fjm membrane filter.  This filtration
removes  any  indigenous  algae   from  the
dilutions. Following filtration, 150  mL of each
dilution is measured in each of three 125 mL
Erlenmeyer test flasks (three flasks per dilution
with 50 mL of diluent per flask). The entire test
consists of three test flasks  in each  of the
following concentrations; control, 0%, 6.25%,
12.5%, 25%, 50%,  100%.

The  test flasks are  inoculated with 1  mL of
log-phase-growth Selenastrum (4 to 7 days old)
to provide an  initial cell density  of 10,000
cells/mL (±10%). At IEPA the algal cells are
not "washed" prior to inoculation  (there is no
need to remove EDTA from the test cells since
the test nutrients contain EDTA). The required
volume of stock culture needed to inoculate the
test flasks is calculated as follows:

number of   x   volume  of test   x   10,000
test flasks      solution/flask       cells/mL
 cell density (cells/mL) in the stock culture

=  volume (mL) of stock culture required

Test flasks are covered with aluminum foil for
autoclaving. After the flasks are inoculated, the

Charles and Searle


• •




9 A





a Da
o a D a
a °



Figure 1. Illinois EPA Reference Toxicity Test (NaCI).

                            Front row moves to back
 Figure 2. Rotational pattern.

                                                IEPA Selenastrum Culturing and Test Protocol
aluminum foil  is removed and  replaced with
100 mL plastic beakers for incubation. Cell
density is checked in at least three flasks within
two hours of  inoculation. The test flasks  are
randomized  on  the  holding  tray  prior  to
incubation. The flasks are rotated  at 24,  48,
and 72 hours using a standard rotation pattern
(Figure 2) to help ensure that all test containers
receive equal amounts and intensity of light  and
even  temperature throughout  the 96 hour

The algal incubator  is kept  at a  constant
temperature of 25 ±  1°C. Lighting is turned on
approximately four hours before the start of the
test to  allow  the lights to reach equilibrium.
During  the   test the flasks  are  rotated
mechanically  at  100  cycles  per  minute
continuous rotation.  Light intensity during  the
test is 400 ±  40 ft-c (4306 ± 431 lux). Light
lux is measured at the beginning  of the test  and
at the end of the test. The pH of the 0%  and
the 100% is also measured at the beginning
and at the end of the test.

Test termination is at 96  ± 2 hours. Algal cell
density  in  each  flask is1 measured using  a
Coulter  Counter" Model ZM. Test cultures are
diluted   with   Isoton" (a  sodium  chloride
electrolyte solution) and counted directly on the
Coulter  Counter". Three cell counts are taken
for each aliquot and the  mean  cell volume is
averaged for the three counts. Each test flask
is  mixed   thoroughly  following   US EPA
procedure. For IEPA purposes, the counts from
each  test culture must have less than 10%

Test results are considered acceptable  if  the
average  cell counts  in the control flasks  are
greater  than  2  x. 106  cells/mL and control
variability does not exceed 20%. When using
stock   nutrient   solutions  without   EDTA,
obtaining average cell counts  greater than
200,000 cells/mL was not a problem, however
keeping  control  variability below 20% was
difficult. Without EDTA,  cell  counts  in  the
controls ranged from 400,000 to  800,000
cells/mL. Variability in the three control flasks
was as high as 79%. variability in control flasks
inoculated with stock nutrient solutions contain-
ing EDTA consistently remained below 20%.

EDTA  can  lower toxicity  of  a  sample  by
complexing heavy metals. EDTA facilitates algal
growth   by  increasing  the  availability  of
micronutrients. Based on the control  flask
variability (when EDTA is not used) the decision
was made at IEPA to conduct the Selenastrum
algal bioassay with EDTA in the stock nutrient
solutions. Adverse effects on Selenastrum cell
growth, expressed in  LOEC and NOEC  values,
are  obtained  using   Dunnett's  Procedure.
Statistics are  analyzed   using  an  in-house
written computer program.

The Selenastrum algal  bioassay  is a useful
aquatic toxicity  test, and is an important
component  of  lEPA's  testing  program.  In
addition to detecting phytotoxic contaminants,
the bioassay could identify wastewaters which
are  nutrient  rich  and   biostimulatory.  By
incorporating  a  freshwater primary producer
(Selenastrum) into the bioassay regime, toxicity
could be detected  which is not  detected by
tests  using  primary consumers (Ceriodaohnia
dubia.  at   IEPA)  or  secondary  consumers
(Pimeohales promelas. at IEPA).

Literature Cited

USEPA.   1989.   Short-Term  Methods   for
Estimating the Chronic Toxicity of Effluents and
Receiving Waters to Freshwater  Organisms.
Environmental Monitoring Systems Laboratory,
U.  S.   Environmental   Protection  Agency,
Cincinatti, Ohio, EPA/600/4-89/001.

Effects of Acute Sublethal Levels of  pH on  the Feeding Behavior of Juvenile
Fathead Minnows

Robert D. Hoyt1 and Hanan Abdul-Rahim2
Graduate Center for Toxicology
The University of Kentucky
Lexington, KY 40506

This study was conducted to determine the impact of acute sublethal pH levels on the feeding
behavior of juvenile fathead minnows. Eighteen to 24 day-old juveniles were fed live or dead brine
shrimp under  light or  dark  conditions in  order  to  identify  the  role of the senses  of  vision,
chemoreception, and mechanoreception in feeding at different pH's. Feeding trials were conducted
at various pH combinations; 5.0, 7.0, 10.0; 4.5, 7.0, 11.0; and 3.5, 7.0, 11.5. Total mortality was
observed at  pH 3.5 and 11.5. Appetitive behavior was present at all pH levels as evidenced by
frequencies of occurrence of feeding ranging from 93-100%. No relationship was observed between
pH and the number of fish feeding. The fathead minnow is chiefly a visual feeder and vision was not
affected at any pH level as 99.9% of all brine  shrimp, live and dead,  were consumed  in the light.
Significantly fewer brine shrimp were consumed in the dark than in the light and significantly fewer
brine shrimp were consumed in the dark at the lower  pH's than  in the dark at pH's 7.0+ .   No
measurable impact on feeding behavior was observed at pH 7.0 and 10.0 + . Chemoreception was
stressed at low pH levels. The ability of chemoreception and mechanoreception to successfully
function in consort in capturing living prey in the dark at low pH levels was noticeably impacted. The
effect of low pH on mechanoreception was not determined.

Keywords: pH, fathead minnow, feeding behavior
The science of behavioral  toxicology is a
recently  developed diagnostic  approach to
measuring  and  recording  observations  of
behavior  that   reflect   biochemical   and
ecological   responses  of   organisms  to
environmental contamination (Little 1990).

The science of behavioral  toxicology is a
recently  developed diagnostic  approach to
measuring  and  recording  observations  of
behavior  that   reflect   biochemical   and
ecological   responses  of   organisms  to
environmental contamination  (Little 1990).
Behavioral  activities  are  rapidly  becoming
recognized  as highly sensitive indicators of
sublethal toxicity (Diamond et al.  1990, Little
and Finger 1990). A variety of behaviors has
been  used   to  study  sublethal  toxicities
including ventilation and cough frequencies,
feeding  activities, temperature  preference,
predator avoidance, swimming performance,
schooling  behavior, and pH detection and
avoidance (Hill 1989). However, while it is
readily acknowledged by investigators that
differing behavior activities involve a diversity
of sensory-motor pathways and physiological
processes (Sandheinrich and Atchison 1990),
little attention has been given to the impacts
that toxicants selectively impart to specific
senses  or sensory pathways. Although  the
    1  Present Address: Department of Biology, Western Kentucky University, Bowling Green, KY

    2 Present Address: Salem College, Salem, NC 27108

                                                                       Sublethal pH Effects
embryo-larval-juvenile life cycle stages are
accepted  as being the  most  sensitive for
toxicity   tests   (McKim   1977),   little
consideration has been given to the impacts of
sublethal toxicants upon the sensory systems
of these life cycle stages, many  of  which
exhibit a gradient of sense organ development
from the time of hatching until the completion
of successful behavior formation (Noakes and
Godin 1988).

Although  acid  stress  and  depressed pH
conditions have been studied at length from
many different  perspectives  (Zischke et al.
1983,  Leino et  al. 1987, Mills et  al.  1987,
Jansen and Gee 1988, among others), and are
receiving much local press in  relation to acid
precipitation, little attention has been given to
the effects  of  acid stress  on fish behavior
(Jones et al. 1985). Lemly and Smith (1985)
summarized the literature supporting fathead
minnows as being among the most acid sensi-
tive fishes. Jones et al. (1985), in pursuing the
effects of sublethal pH levels on the behavior
of arctic char, reported acid stress to suppress
chemoreception.   Lemly and  Smith (1985,
1987)  found acidification to significantly
affect the ability of fathead minnows to detect
or respond to chemical  stimuli. Jones et al.
(1985) described this chemo-suppression to
likely result  from the  reduced stimulatory
nature of amino acids at reduced pH's and the
damage of  epithelial tissues  (olfactory epi-
thelium) by acidic conditions. Lemly and Smith
(1987) suggested that   increased  olfactory
mucous thickness in response to lowered pH
prevented normal stimulus-receptor interaction
and/or  that  chemical interaction at the sub-
cellular level was impaired because of stearic/
charge  changes  at the   receptor   cells.
Whatever the explanation, these observations
are of potentially profound  importance to
environmental biologists in that they represent
avenues  for  unrecognized  massive  larval
mortalities among those  fish species that are
dependent   upon  chemoreceptors   of
chemoreceptors/mechanoreceptors   in   the
formation of exogenous feeding behavior.
 The purpose of this study was to determine
 the effects of different acute pH levels on the
 senses  of  vision,  mechanoreception  and
 chemoreception  in the  feeding  behavior  of
 juvenile   fathead   minnows,   Pimeohales

 Methods and Materials
 Test Fish
 Fathead minnows used  in the  project were
 obtained within 12 hours of hatching from the
 U.S. EPA  Newtown Fish facility, Newtown,
 Ohio,  on 3 July,  1990. Fish were maintained
 in 2 1  finger bowl in ASTM water at 24 + /-1C
 and were  fed freshly hatched  brine  shrimp
 twice  daily, 0800 and 1700 h. At the onset of
 the pH trials, the minnows were 18 days old
 and averaged 11.58 mm in total length (range
 10.1 mm- 13.0mm). All experimentation was
 conducted   at   the  Graduate  Center  for
 Toxicology at the University of Kentucky,
 Lexington, KY.

 pH Test Solutions
 pH solutions in 3.0 1 aliquots were  prepared
 using  ASTM water with Nitric acid to produce
 low pH levels  and  Sodium  Hydroxide  to
 produce high pH levels. A  pH of  7.0 was
 achieved  by adding  either  Nitric  acid  or
 Sodium Hydroxide as required.  An Orion pH
 meter was used to determine  pH  levels in
 producing the desired pH concentrations. Acid
 pH's  tested were  5.0,  4.5, 4.0, and 3.5.
 Basic  pH's were 10,10.5,11.0, and 11.5. An
 acidic, a basic, and a pH 7.0 concentration,
 was used  daily for four consecutive days in
 the following sequence: Day 1 - pH's 5.0, 7.0,
 10.0;  Day 2 - pH's 4.5, 7.0, 10.5; Day 3 -
 pH's 4.0,  7.0, 11.0; Day 4  - pH's 3.5. 7.0,

 Sense Organ Isolation
 Each  daily combination  of pH test  solutions
 was applied to four different feeding regimes,
 living  brine shrimp fed in the light, living brine
 shrimp fed in the dark, dead  brine shrimp fed
 in the light, and dead brine shrimp fed in the
' dark (see Table 1 for design).

Hoyt and Abdul-Rahim
Table 1. Frequency of occurrence of juvenile fathead minnows eating at least one brine shrimp during
live-dead, light-dark, feeding trials at varying pH levels.
pH Level

pH Level

pH Level

pH Level
Test Procedure




5/5 4/5
5/5 5/5
5/5 5/5
5/5 5/5
5/5 5/6
4/5 4/5
5/5 5/5
5/5 5/5
5/5 55
5/5 5/5
5/5 5/5
5/5 4/4






5/5 45
55 55
5/5 55
5/5 3/3
4/5 55
55 55
5/5 55
5/5 45
55 55
55 55
55 55
5/5 55


The fish




4/5 45
5/5 55
5/5 55
5/5 55
5/5 55
5/5 55
5/5 55
5/5 55
5/5 55
5/5 55
4/5 55
4/4 4/4


were allowed to acclimate in the
A total of  180 minnows  were selected at
approximately 1600 h the day before a trial.
The fish were separated into 12 groups of 15,
each group  of which was placed in a  500 ml
beaker  containing  water of a  specific  pH
concentration   (daily  test  combinations
described above). The fish were then placed in
an environmental chamber at 25.0 C with an
8 hour dark period (2200 to 0600 h)  for
acclimation until approximately  1300 h the
following  day.   Five  fish  from  each  pH
concentration were then placed in  each of
three 150 ml finger  bowls containing  fresh
mixtures of the test pH's for replicate trials.
finger bowls for ten minutes in either light or
dark before food was added. Light feeding
trials  with  live  and dead  brine shrimp were
conducted prior to similar dark feeding trials.

Brine  shrimp (Salt  Lake City variety) were
raised in the laboratory and fed immediately
following  24 hours  incubation. The brine
shrimp used in  the feeding  trials were the
same variety and  size  used to  raise  and
maintain the minnows. Average  brine shrimp
length was 0.7 mm.  Fifty live or dead brine
shrimp for each fish subsample were selected

                                                                        Sublethal pH Effects
Table 2. Number and Percent of brine shrimp remaining following juvenile fathead minnow live-dead,
light-dark feeding trials at varying pH levels.
pH Level


0 (2.7%)
8 (20.7%)

pH Level


23 (41.3%)
21 (27.3%)

pH Level


7 (7.3%)

pH Level




0 (7.3%)


3 (3.3%)
0 (6.0%)


1 (1.3%)





3 (8.7%)







with a  10 cc syringe and counted using a
dissecting microscope. Each of nine syringes
was loaded immediately  prior to the feeding
exercise and the fifty brine shrimp  added to
each  group of  five fish following the  ten
minutes acclimation to the test pH's. Feeding
time for all  tests  was  ten  minutes. Brine
shrimp for the dead feeding trials were killed
by treatment in an ultrasonic  bath for two to
four minutes. Fish  were aspirated  from the
test dishes immediately following  the feeding
trial and isolated in holding dishes. While the
number of brine  shrimp  ingested  by  each
individual  fish could not be determined, the
number of fish having consumed at least one
brine shrimp was recorded using a dissecting
microscope.  Each feeding test  dish  was
examined   with the  aid  of  a  dissecting
microscope and the number of brine shrimp
remaining  following the  feeding  trial  was
counted. All  fish used in  a feeding  exercise
were excluded from further feeding.

Hoyt and Abdul-Rahim
All fish survived every trial except those at pH
3.5 and 11.5 in which  100 percent mortality
was observed.

Frequency of Occurrence of Feeding
The number of fish ingesting at least one brine
shrimp during the feeding trials ranged from
93 to 100 percent for all pH levels (Table 1).
No relationship was observed between number
of fish feeding and light or dark or live or dead
food  conditions.  No  relationship between
frequency of occurrence of fish feeding and
decreasing   or  increasing  pH  levels  was
observed.   The   average   frequency  of
occurrence of feeding by fishes in pH 7.0
water, 97.7%, was the same as that (97.7%)
for  fish in  the increasing pH level trials and
only slightly greater than that (96.1 %) for fish
in the decreasing pH level trials (Table  1).

Light-Dark Feeding versus pH Level
With the exception of three brine shrimp at pH
4.5, all food organisms (99.9%). live or dead,
were  consumed during light feeding trials, at
all pH levels (Table 2). In the dark, however,
live  or  dead brine  shrimp  remained after
feeding in 15 of 18 trials. The number of brine
shrimp remaining in each  dark  trial  ranged
from  1.3% to 49.3% (Table  2).  At pH  7.0,
slightly more dead brine shrimp  remained in
the dark than live  brine shrimp, 5,0% and
1.1%, respectively.  At the  lower pH's 5.0,
4.5, and 4.0 combined, more live brine shrimp
(31.1%) remained  in  the dark  than dead
shrimp (18.4%). The number of live and dead
brine shrimp remaining at the higher pH's was
generally similar to that of pH 7.0 (Table 2).
Significantly fewer  live brine  shrimp were
consumed  at pH  4.5 and 4.0  than  at the
higher levels while significantly fewer dead
brine  shrimp were consumed at pH 5.0 and
4.5 (Table  2).

Based upon the high frequency of occurrence
of  feeding  (93 + %)   by  juvenile  fathead
minnows  at all  pH  levels during  all  test
regimens, appetitive behavior was concluded
to be present to the lethal pH levels of 3.5 and
11.5. Hill (1989), in a chronic study of low pH
effects  on feeding behavior of smallmouth
bass,  also observed  no  loss  of  appetitive
behavior at pH 4.2. Mortality at pH 3.5 in this
study was  consistent  with the report by
Mount  (1973)  that  most lethal pH values
recorded from  laboratory data occur below
4.0. No similar data were found regarding high
pH  mortality,  although  Carlender (1969)
reported the fathead minnow to have a broad
tolerance for pH. The persistence of appetitive
behavior through  all  pH  levels was further
supported by the minnows  eating all except
three  (99.9%) live and dead brine shrimp fed
during the light feeding trials. Consequently,
since  appetitive behavior persisted throughout
the study, any observed reductions in feeding
activity were considered to be the result of the
selective impairment of sense organs by the
different  pH   levels,  or  the  behavioral
inactivation of the feeding response during
certain  environmental conditions of  the test  (
i.e..dark), or possibly a combination of  both
these features.

According to the feeding patterns observed in
this  study,   the   fathead   minnow  is
predominantly a  visual, daylight feeder. The
removal of all except three brine shrimp, live
and dead, in the light at all pH levels identified
the eyes, or the eyes in conjunction with the
senses  of  chemoreception and  mechano-
reception, as the major sense organs involved
in early life stage feeding. Klemm (1985)
reported the fathead  minnow to be primarily
omnivorous  and  to  possess  large   black
eyes,presumably  functional   if   heavily
pigmented,  at the time of hatching. The eye
must play a strategically  greater role in early
life   stage  feeding   as  indicated  by  the
recommendations by  Birch et al. (1975) and
Klemm (1985) that fathead minnow fry at
least 6  days to 28 days old be fed live, freshly
hatched (small) brine shrimp during the day,
 while older individuals may be fed frozen brine

                                                                       Sublethal pH Effects
shrimp or varying  forms of  dry  chow. This
recommendation suggested  a  greater visual
feeding  success  on living  prey by  larval-
juvenile individuals and more successful visual-
chemoreceptive  feeding  behavior in  older

The eyes did not  appear to be  functionally
impacted by different pH levels in this study.
All brine shrimp, living and  dead except for
three  individuals, were consumed in the light
at all pH levels.  However, the  small test
chambers (150 ml) with concomitant  short
reaction distance and number of brine shrimp
per trial  (50) might have alleviated any eye
stress that was present or  developing. Hill
(1989) reported low pH levels to impair visual
acuity, coordination, and agility, subsequently
resulting in lower growth in chronic trials with
smallmouth  bass. The  question  of  acute
versus chronic study interpretations is brought
into   focus  at  this  point.  Hill's  (1989)
suggestion that acute bioassays  may be too
short  to detect certain biological  parameters,
such as growth and survival,  is  not contested.
However, in meeting the objectives of studies
such as this, acute sublethal tests, especially
feeding, may be more sensitive than chronic
growth  studies (Sandheinrich  and Atchison
1990). That the fathead minnow is  not an
adept nocturnal feeder was supported by the
number of uneaten live and dead brine shrimp
in dark feeding trials at all pH's. The presence
of more  dead (5.0%) than live (1.1%) brine
shrimp following dark feeding trials at pH 7.0
indicated a slightly greater ability or behavioral
preference by the  fathead minnow to select
living  food  over  dead  food  in the  dark.
However, at the lower pH's  during the dark,
feeding  was greatly reduced and more dead
brine shrimp were consumed  than live  shrimp.

Chemoreception   was   considered   to  be
impaired at pH 5.0 when 20.7% of the dead
brine  shrimp remained, worsened at  pH 4.5
when 27.3% remained, and  then inexplicably
improved at pH 4.0 when 7.3% of the shrimp
remained. The marked improvement in dark-
dead feeding at the lowest pH could not be
explained. Feeding on dead food in the dark
was considered to be entirely a function of
chemoreception since stimuli for visual and
mechanoreceptive senses  were not present.
Consequently, since appetitive behavior was
known to exist, this  decreased feeding  on
dead  food in the dark strongly suggested
chemoreceptive inhibition. Yoshii and Kurihara
(1983)   reported   that   bluegill   without
functional  lateral  lines   did  not  produce
successful feeding strikes in the dark based on
chemoreception   alone.   The   omnivorous
feeding capability of certain  species such as
bluegill sunfish and fathead minnow might
employ the sense of chemoreception at levels
not yet described and different than  other
species. However, the findings by Jones et al.
(1985)  that  pH   5.0 suppressed  chemo-
orientation in the Arctic char, and Lemly and
Smith  (1985, 1987) that  pH 6.0 eliminated
fathead minnow responses to chemical stimuli
supported the initial conclusions drawn in this
study that chemoreception was  impaired by
the lower pH levels.

Living  prey in the dark seemed to represent
the maximum sensory challenge  in feeding,
especially for  the  visual feeding  fathead
minnow.  Consequently, low pH exhibited its
greatest impact on feeding  on live brine shrimp
in the  dark.  Although  live  brine shrimp were
successfully preyed upon in the dark at pH 7.0
and higher, only slightly more than 50% were
captured at the lowest pH's.  Live food under
dark   conditions  would   suggest   the
involvement  of the  combined  senses  of
chemoreception  and   mechanoreception  in
successful feeding. Enger et al. (1989) and
Montgomery  (1989)  presented   evidence
substantiating the role of mechanoreceptors in
detecting  moving prey. Montgomery (1989)
further described the role of mechanoreceptors
as  operating  synergistically  with vision  in
daylight   planktivory  and  singly  in  total
darkness.  No  mention   was   made  by
Montgomery of any receptor involvement with
mechanoreceptors  in  dark  feeding.  Hara

Hoyt and Abdul-Rahim
(1986) summarized the extensive  literature
reviews on  the role of  chemoreception in
feeding behavior and identified the first step in
the feeding  sequence  as arousal to  the
presence of food which is primarily  mediated
by olfaction. Atema (1980) excepted the most
visual fish species, i.e., anosmic sticklebacks,
from the above behavior  and suggested that
accompanying   senses  such  as
mechanoreception and vision may also be
involved  in  initial prey  recognition.  Should
these sensory  assumptions  be  correct  and
chemoreception was impaired to a reduced
level  of  effectiveness  in  establishing  the
presence  of   living  food,  then  several
explanations  regarding   the   role  of
mechanoreception in  dark  feeding become
likely.   First,   mechanoreceptors   act
synergistically with chemoreception in dark-
live feeding and the impairment of one sense
automatically  reduced  the success  of  the
other; secondly, mechanoreceptors  also were
impaired by the lowered pH rendering them
incapable of successfully  detecting  and/or
locating   the  moving   prey;   or  thirdly,
mechanoreceptors alone  are  incapable of
successfully detecting and locating living prey
in complete darkness  in  strongly visually
directed species.

Fathead  minnow feeding behavior was not
observed to be affected by any non-lethal pH
in the light. The interaction of  vision with
chemoreception,  or mechanoreception, or
both, produced  successful  feeding  on  live
brine shrimp at all pH levels. Likewise, in the
dark at pH's of 7.0 and higher, the senses of
chemoreception   and   mechanoreception
operated  successfully in feeding on live food
(98.9%) in the dark. However, at low pH's in
the   dark,  chemoreception  and
mechanoreception were impaired in detecting
and locating living prey. While chemoreception
was observed to  be stressed at low pH's, no
evidence   of  such   an   effect   on
mechanoreception  was  detected.   Future
studies  using  streptomycin  sulfate
(Montgomery 1989)  or  cobalt (Karlsen  and
Sand 1987) to ablate mechanoreceptors might
provide  valuable  insights  into  the role of
mechanoreceptors  in initiating  as  well as
concluding the feeding response.

This study was supported by funding from the
National Science Foundation Kentucky EPSCoR
Program  to  the  first author  and the 1990
Undergraduate Summer Research Experience
for Minorities, Graduate Center for Toxicology,
University of Kentucky, the second author. We
thank W.J. Birge of the Departments of
Biological  Sciences  and   Toxicology,  the
University of Kentucky, for serving as project
host  and  providing  the  physical  facilities
necessary for this investigation. Special thanks
go to T. Short and  W.A. Robison for their
technical expertise and assistance. Additional
thanks go to the numerous other members of
Dr. Birge's laboratory who contributed to the

Literature Cited

Atema, J.  1980. Chemical senses, chemical
signals, and feeding behavior in fishes, pp. 57-
94. In:  Bardach,  J.E.  et al.  (eds.),  Fish
behavior and its use in the capture and culture
of fishes.   International  Center for Living
Aquatic Resources Management, Manila.

Birch, T.J., J.P. Abrams, and  G.L. Martin.
1975. Design and operation of a static rearing
unit for fathead minnows. Ohio Environmental
Protection  Agency,  Division  of  Surveillance
and Laboratory Services, Columbus, 8 pp.

Carlender, K.D. 1969. Handbook of freshwater
fishery biology.  Vol. 1. Iowa State University
Press, Ames, I A.

Diamond, J.M., M.J. Parson, and D. Gruber.
 1990. Rapid detection  of sublethal toxicity
using fish ventilatory behavior.  Environ. Tox.
Chem. 9:3-11.

                                                                      Sublethal pH Effects
Enger, P.S.. AJ. Kalmijn, and 0. Sand. 1989.
Behavioral investigations on the functions of
the lateral line and inner ear in predation, pp.
575-587.  In: Coombs, S. et al. (eds), The
Mechanosensory Lateral Line. Springer-Verlag,
New York.

Hara, T.J.  1986. Role of  olfaction in fish
behavior, pp. 152-176. In: Pitcher, T.J. (ed.).
The behavior of teleost fishes. Croom Helm,

Hill, J.  1989. Analysis  of  six  foraging
behaviors as toxicity indicators, using juvenile
smallmouth   bass   exposed  to   low
environmental pH. Arch. Env. Contamination
Tox. 18:895-899.

Jansen, W.A.,  and  J.H.  1988. Effects of
water acidity on  swimbladder function  and
swimming in the fathead minnow, Pimephales
promelas. Can. J. Fish. Aquatic Sci. 45:65-77.

Jones, K.A., T.J. Hara, and E. Scherer. 1985.
Behavioral  modifications   in  arctic  char,
(Salvelinus aloinus)   chronically exposed to
sublethal pH. Phys. Zool. 58(4):400-412.

Karlsen, H.E., and 0. Sand. 1987.  Selective
and reversible blocking of the lateral line in
freshwater fish. J. Exp. Biol. 133:249-262.

Klemm,  D.J. 1985.  Distribution, life cycle,
taxonomy,  and culture  methods.  Fathead
minnow (Pimephales promelas). pp. 112-125.
In:  United  States  Environmental Protection
Agency,  Methods for Measuring the Acute
Toxicity of Effluents to Freshwater and Marine
Organisms.  U.S. Environmental  Protection
Agency,   EMSL,  Cincinnati.   EPA/600/4-

Leino, R.L., P. Wilkinson,  and J.G. Anderson.
1987. Histopathological changes in the gills of
pearl dace, Semotilus maroarita. and fathead
minnows,   Pimephales   promelas.  from
experimentally acidified Canadian lakes. Can.
J. Fish. Aquatic Sci. 44 (Suppl. 11:126-134.
Lemly, A.D., and R.J.F. Smith. 1985. Effects
of acute  exposure to  acidified water on the
behavioral  response  of  fathead  minnows,
Pimephales promelas.  to chemical  feeding
stimuli. Aquatic Tox. 6:25-36.

Lemly, A.D., and R.J.F. Smith. 1987. Effects
of chronic exposure  to  acidified water  on
chemoreception of feeding stimuli in  fathead
minnows (Pimephales promelas): mechanisms
and ecological implications. Env. Tox. Chem.

Little,  E.E.  1990.  Behavioral  toxicology:
stimulating challenges for a growing discipline.
Env. Tox. Chem. 9:1-2.

Little, E.E. and S.E. Finger. 1990. Swimming
behavior as an indicator of sublethal toxicity in
fish. Env. Tox. Chem. 9:13-19.

McKim, J.M. 1977. Evaluation of tests with
early life stages of fish for predicting long-term
toxicity.  J. Fish.  Res. Board  Can. 34:1148-

Mills, K.H., S.M. Chalanchuk, L.C. Mohr, and
I.J.   Davies.   1987.  Responses  of  fish
populations in Lake 223 to 8  years of
experimental   acidification.  Can.  J.  Fish.
Aquatic Sci. 44 (Suppl. 1) 114-125.

Montgomery, J.C. 1989. Lateral line detection
of planktonic prey, pp. 561-574. In: Coombs,
S. et al.  (eds.). The Mechanosensory Lateral
Line. Springer-Verlag, New York.

Mount, D.I. 1973. Chronic effect of low pH on
fathead   minnow  survival,   growth   and
reproduction. Water Res. 7:987-993.

Noakes,  D.L.G.  and J.G.J.  Godin. 1988.
Ontogeny  of   behavior  and   concurrent
developmental changes in sensory systems in
teleost fishes, pp. 345-395. In: physiology of
developing  fish,   Part  B:   Viviparity  and
posthatching juveniles. Academic Press.

Hoyt and Abdul-Rahim
Sandheinrich, M.B. and G.J. Atchison. 1990.
Sublethal toxicant effects  on fish foraging
behavior:   empirical   vs.   mechanistic
approaches. Env. Tox. Chem.  9;107-119.

Yoshii, K. and  K. Kurihara.  1983. Role  of
cations  in  olfactory  reception.  Brain  Res.

Zischke, J.A., J.W. Arthur, K.J. Norlie,  R.O.
Hermanutz, D.A. Standen,  and T.P.  Henry.
1983.  Acidification   effects   on
macroinvertebrates  and  fathead  minnows
(Pimeohales oromelas) in outdoor experimental
channels. Water Res. 17:47-63.