'“———-—— — ..
"’ . V . ..
S ‘. My},

 

EFFECT or» METAL PLAYING WASTES
014‘ THE ECQLQGY gr A
WARM-WATER :STREAM

This-ts {M “In Dawn 0‘ D“. D.
MICHIGAN STATE UNIVERSITY

Ronald R. Garton
' 1968

' 53‘:
. . . .
-"— -.' '5“;
nu u} 3 c.s;r£;ii-‘I~'.ug!3

 

 

‘

IIIIIIIIZIIIIIIIIIIIIII/ILIIIIIIIIIILIII L ‘ % I7“ “MARY“

t‘ ., r.-

v..';,"'?‘;;'_ Chlte
1 UAA. ‘ Lls‘ty

mvw

 

This is to certify that the ‘ ‘4’

thesis entitled

"£660.01: 06 Metal. 1’an wazsteb on the, Ecofiogy 06

A (Ha/1m Wale/L SULeam"

presented by

Ronald R. Gall/ton

has been accepted towards fulfillment
of the requirements for

ii;— degree in M9 9

Major professor

Date Jufly 24, 1968

0-169

. . .<—_~A - _—._.-—-.._

 

 

 

 

 

ABSTRACT

EFFECT OF METAL PLATING WASTES ON THE
ECOLOGY OF A WARMrWATER STREAM

by Ronald R° Garton

The purpose of this study was to determine the ef-
fects of metal plating wastes upon the ecology of the Red
Cedar River in southern Michigan. Water samples were
analyzed for Cu, Ni, Zn, Cr6, Cr Total and CN' from stations
selected upstream and for 10.3 miles downstream from the
plating plant and numbers and types of invertebrates were
determined at the same spots. Plating wastes could be de-
tected for 10.3 miles below the effluent and strong negative
correlations were found between wastes and all families of
invertebrates except tubificids. Traditional methods of
assessing pollution by indicator types, numbers of families,
etc. were used. In addition, an index derived from the
ratio of insects to tubificids is proposed as a measure of
pollution and an optimum index value is suggested for the
Red Cedar River. By means of a regression equation of

insect-tubificid index on waste concentrations, maximum

limits are suggested for plating wastes in the river.

EFFECT OF METAL PLATING WASTES ON THE

ECOLOGY OF A WARMrWATER STREAM

BY

Ronald fif\Garton

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

1968

ACKNOWLEDGMENTS

I wish to sincerely express my gratitude to Dr.
Robert C. Ball for the important guidance and assistance
which he provided during the course of this study and
during my entire graduate program.

I also wish to thank Dr. Gordon Guyer, Dr. Gerald
Prescott, Dr. Peter Tack, and Dr. Niles Kevern for the time
they gave as members of my guidance committee. And I wish
to thank Dr. Duane Ullrey and his staff in Animal Husbandry
who set such a fine example of interdepartmental cooperation
in making laboratory facilities and equipment available for
chemical analyses. Thanks also to fellow graduate students
for helpful advice and to Alvin Jensen especially for as-
sistance with computer programming.

This study was conducted during tenure of a Pre-
doctoral Research Fellowship (5—Fl-WP526,008) sponsored by

the Federal Water Pollution Control Administration.

ii

TABLE OF CONTENTS

INTRODUCTION .

DESCRIPTION OF STUDY AREA

DESCRIPTION OF PLATING PLANT AND ITS OPERATION .

METHODS

Sampling Program
Chemical Analyses

RESULTS

Hexavalent chromium
Cyanide

Copper

Nickel

Zinc

Total chromium

AND DISCUSSION .

Water Analyses

Copper

Nickel

Zinc

Chromium

Cyanide .
wastes in combination

Invertebrate Sampling

SUMMARY

Traditional methods of analysis
Correlation analysis .
Principal components analysis
Index of pollution

Multiple regression analysis

LITERATURE CITED

iii

Page

11
19

19
26
26
26
26
27
27
27

29

29
31
34
36
39
44
47
51
51
61
64
67
69

78

81

Table

10.

ll.

12.

13.

LIST OF TABLES

River discharge in feet per second during
sampling program

Location of sample transects in miles below
plating plant outfall . . . . . . .

Maximum and minimum values for water chemistry
determinations made by Michigan Water Re-
sources Commission during period from 1953
to 1967 . . . . . . . . . . . . .

Average parts per million copper in each
sample series . . . . . . . . . .

Average parts per million nickel in each
sample series . . . . . . . . . . . . .

Average parts per million zinc in each
sample series . . . . . . .

Average parts per million hexavalent chromium
in each sample series

Average parts per million total chromium in
each sample series . . . . . . . . . .

Average parts per million cyanide in each
sample series . . . . . . .

Maximum amounts of each waste found in each
section of the river . . . . . . . . .

Numbers of invertebrates per square meter in
each sample series - pilot program . . .

Numbers of invertebrates per square meter in
each sample series - series 1 . . . . .

NUmbers of invertebrates per square meter in
each sample series - series 2 .

iv

Page

21

30

32

35

37

4O

43

45

50

56

57

58

Table

14.

15.

16.

17.

Numbers 0f invertebrates per square meter in
each sample series - series 3 . . . .

NUmbers of invertebrates per square meter in
each sample series - series 4 . .

Intercorrelation matrix for 11 families and
six metals

Factor loading matrix for principal com-
ponents analysis . . . . . . . . .

Page

59

6O

63

65

Figure

LIST OF FIGURES

Map of Red Cedar River showing location of
original study zones

First floor plan of metal plating plant oper-
ated by Utilex Manufacturing Company,
Fowlerville, Michigan .

Map of final study sections

Mean number of invertebrate families found
in each study section

Relation of mean insect-tubificid index to
distance downstream from the plating plant

Relation of observed to predicted insect-
tubificid index values

vi

Page

14

24

54

71

75

INTRODUCTION

This study was conducted in order to determine the
effect of metal plating wastes on the invertebrate popu—
lations in the Red Cedar River, a warm-water stream in the
southern lower penninsula of Michigan.

The Red Cedar River originates in Cedar Lake in
Livingston County and then flows for approximately 45 miles
in a northwesterly direction through Livingston and Ingham
counties and finally flows into the Grand River in the city
of Lansing. The river receives the water of 12 major tribu-
taries and drains an area of about 472 square miles (Figure
l). The upstream portion of the river drains marSh and
agricultural land used primarily for dairy farming and some
raising of small grains. The land along the lower reaches
of the river is more intensively farmed and the river is
bordered by several small communities. Even in areas of in-
tensive farming the fields seldom extend right up to the
river banks so the river itself is lined with trees through—
out the majority of its length.

The Red Cedar River has been the subject of several
studies in past years by graduate students in the Fisheries
and Wildlife Department of Michigan State University and by

the Michigan State water Resources Commission. For the

Figure 1. Map of Red Cedar River showing location of

original study zones.

3

H

 

a V 00!"
§ .g‘
'Ut‘VO 3
. . o
.3 J10,“ c '83‘ {NINTH}:
2 '30
C - zl3la
IO lvnos - 7
3;... as “00

LAKE

65'

 

 

O

 

A

 

g] _L L A
'

Kl LOMETERS

MILES

studies done by Fisheries and Wildlife graduate students,
the river was divided into five zones extending from Farm
Lane bridge on the campus of Michigan State University to
Van Buren Road bridge 30 miles upstream from campus and
about one mile above the town of Fowlerville (Figure 1).
This 30 mile section of the river drains an area of 355
square miles and the stream varies in width from about 20 to
80 feet. Average gradient is 2.4 feet per mile and dis-
charge at the campus varies seasonally from an all-time low
of 3cfs to a record high of 5510 cfs (King, 1964).

Past studies have been largely oriented toward
characterizing and comparing different zones of the river in
terms of chemistry and biology. In doing this, previous
students have noted that productivity is much lower in Zone
V, from Dietz Road bridge (10.32 miles below plating plant)
to Van Buren Road bridge (0.94 miles above plating plant),
than one would expect for a stream such as the Red Cedar
River.

King (1964) found live molluscs to be absent for
nine miles downstream from Fowlerville and found that aquatic
insects were decreased in numbers of individuals and numbers
of families in this area and that numbers of sludge worms
(Tubificidae) were increased. This he attributed to metal
plating wastes entering the river at the town of Fowlerville.

Linton and Ball (1965) found that the river supported

extremely low numbers of fish and that fish populations did

nOt recover until the lower stretches of the zone above
Dietz Road bridge. This was in spite of the fact that the
river in Zone V included the widest variety of major habitat
types of any of the study zones and the river was receiving
organic enrichment in this area, which might be expected to
increase the amount of fish food present.

Accumulated evidence from these past studies clearly
indicates that some factor is causing a detrimental effect
upon the aquatic life in Zone V. Further, the most likely
cause of this effect is a metal plating plant operated by
Utilex Manufacturing Company which discharges plating ef-
fluent into the river at Fowlerville. This plant has had a
long history of violations against established Michigan
State water quality criteria for plating effluents and has
been the cause of several fish kills in the past, the most
recent of which is known to have occurred in the spring of
1961. This kill in Zone V, which could be traced for six
miles downstream, was attributed to an accidental release of
ltoxic materials from the plating plant (Parker, 1961).

Although effects of the plating wastes were noted by
previous workers, the studies always encompassed such large
areas of the river that no really intensive sampling was ever
carried out in Zone V, so there was never any really adequate
characterization of the type and extent of effects of the
plating effluent. The plating effluent had an obvious detri-

mental effect upon the stream, but the exact nature of the

effect was not known nor was it known how far the effects
extended downstream from the effluent.

The present study was planned so as to concentrate
chemical and biological sampling in that area of the river
between Dietz Road bridge (10.3 miles below the plating
plant) and Van Buren Road bridge (0.94 miles above the
plating plant), formerly designated as Zone V, in order to
determine the nature and extent of plating plant pollution

and its effects.

DESCRIPTION OF STUDY AREA

That part of the Red Cedar River included in this
study flows through woods and farm land for a distance of
11.26 river miles from Van Buren Road bridge 0.94 miles
above Fowlerville to Dietz Road bridge, 10.32 miles below
Fowlerville.

Elevation decreases from 900 feet MSL at Van Buren
Road to 865 feet MSL at Dietz Road for a total drop of 35
feet in 11.26 miles. Mean gradient for this area of the
river is 3.11 feet per mile as compared to the mean of 2.4
feet per mile for the entire river.

Vannote (1961) used hydrological data from February,
1958 to February, 1959 to compute relative amounts of dis-
charge of the river and tributaries at various points in the
watershed. This information makes it possible to approxi—
mate rate of flow in the upper reaches of the river from dis-
charge records obtained by the United States Geological
Survey at the gaging station at Farm Lane bridge on campus
at East Lansing. According to Vannote, the flow at Van
Buren Road bridge constitutes 27.43 percent of the flow at
the campus and flow at Dietz Road bridge constitutes 50.04

percent of flow at the campus. These are approximations

since rainfall over the entire watershed is not uniform from
year to year.

High, low, and mean flow at the campus during each
of the five sample periods is available from USGS records.
Using these data at Farm Lane bridge and the conversion
factors available from Vannote's hydrological study of the
watershed, approximate high, low, and mean rates of dis-
charge were computed for the upper and lower ends of the
study area during each sampling series (Table 1).

The Red Cedar River has a widely varying rate of
flow and may increase or decrease greatly in just a few days.
This becomes especially important when pollutants are being
released into the river at varying rates.

Although surrounded by farm land, the river in the
study area is bordered by trees throughout most of its
length and seldom has open fields close to the banks. There
are very few people living immediately adjacent to the river
in this area except where it passes through the town of
Fowlerville.

In this area of the river there are three primary
sources of domestic and industrial pollution. The first,
and most important source, is the metal plating plant oper-
ated by Utilex Manufacturing Company in Fowlerville. The
effluent of this plant enters the river at Fowlerville, 0.94
miles below Van Buren Road bridge and 10.32 river miles

above Dietz Road bridge. The next pollutant is domestic

 

v.0m 0.0m m.0m
0.MH 0.NH m.mH
m.NH 0.HH 0.MH
m.mm 0.0H m.mm

0.0 0.0 0.0m

m.¢H 0.0H H.HN
0.5 0.0 0.0
0.0 0.0 H.h
0.NH H.m m.mH

0.0 N.N ¢.mH

0.Nm 0.00 0.55
5.5N 0.0m 0.Hm
0.0m 0.NN 0.0m
0.0V 0.mm 0.Hb
0.0a 0.0 0.00

00.

00.

00.

00.

00.

>02
000
054
Hub

05¢

5N

0H

#00 Hm 0

 

ammz sou roam
DCMHQ mcflumam

onmQ mmafle mm.0H

wmcwnm Umom numwn

 

 

amwz 30A swam
Human mcflumam

m>OQm meHE wm.o
cmom,cmnsm cm>

 

 

II

cmmz 30a swam
mcfimcmq ummm um
omnfinm mama Eumm

II
I

mqu

mom NH m
mas H m
can Hm H
can om uoaflm
mmaumm
onEmm

.Emumoum mcHHmEmm mcwusc ncouwm Hmm Doom Danna cfl omumnumww Ho>Hm .H magma

10

sewage from the Fowlerville sewage lagoons which discharge
into the river 0.65 miles below the plating plant effluent.
The last important source of domestic pollution in the study
area is Kalamink Creek which picks up domestic sewage from
the town of Webberville and discharges it into the river
about seven miles below the plating plant. The effects of
the domestic pollutants have been reported previously by
Vannote (1961) and King (1964) but no major previous study

has been concentrated on the effects of the plating wastes.

DESCRIPTION OF PLATING PLANT AND

ITS OPERATION

The primary source of industrial pollution on the
river is a metal plating plant operated by Utilex Manu-
facturing Company on the Middle Branch of the Red Cedar
River at Fowlerville, Michigan. This plant, which employed
187 people in NOvember, 1967, operates on a l6-hour day, six-
day week basis with an annual one-week shut-down around July
fourth. The two primary operations of the plant are zinc
die casting and decorative plating of plumbing and auto—
motive fixtures.

Pure zinc is purchased in ingot form and alloyed by
the company with 4 percent aluminum before die casting to
desired shapes. After casting, the parts undergo refining
operations such as trimming, machining, buffing, and
tumbling. Any washing or cooling operations here would be a
source of zinc contamination of wastewater. Die cooling
water is recycled with periodic addition of a phosphate base
compound. Floor drains in the die casting area of the plant
are presumed to discharge any spills or floor washings di-
rectly into the river (Caltrider, 1968).

After the fixtures have been cast and refined they

are transported to the plating area and placed on racks to

11

12

be carried through the plating operation. The plating se-
quence, diagramed in Figure 2, is as follows. The fixtures
are first cleaned in two sets of kerosene emulsion cleaners
to remove any dirt or residues from buffing or machining
operations. Wastes from this operation are expected to flow
into the emulsion waste pit and later be moved to an outside
waste pit. However, Caltrider (1968) noted that drippings
from the emulsion cleaner were falling into the metallic
waste holding tank because there was no drip pan present.
After the emulsion cleaner the fixtures go through two sets
of detergent cleaners with a clear—water rinse in between to
remove the kerosene emulsion.

After the second detergent cleaner come two electro
cleaners. In these cleaners the fixtures are made the anode
or cathode and an electrical current is set up through an
alkaline solution. The cleaning is brought about by the
chemical action of the alkali in conjunction with the
mechanical action of vigorous hydrogen gas evolution if the
work is the cathode (Richards, 1946). Cathode and anode may
be inter-changed by an electrical switch during the process
but with zinc castings it is generally preferred that the
zinc be the anode as any film produced during cleaning is
more easily removed in the subsequent acid dip (Metal Finish-
ing Guidebook Directory, 1960).

From the electro cleaner the fixtures pass through a

clear-water rinse and then into an acid pickle solution of

13

Figure 2. First floor plan of metal plating plant
operated by Utilex Manufacturing Company,
Fowlerville, Michigan. Adapted from

Caltrider, 1968.

14

   

 
  
 
 
   

l‘

  
  
   
 

p
2% «2504

 
 

ectre
Cleaner

 
 
  
  
  
  

r
Water
lln

   
  
  

ear
ter

ht

  

Irl
N

        
  
 

  
  
 
    
   
  

    

 
   
 

0 sea /
" r t
3.! c3 /
.- fl’i ’ /
E 'g Strike gall %
‘n : or '3': /
é .. H
2 :5 ° : /
‘ O- -...
1 s 3 2 /////
'- 3 ear ‘7’ O /
a u =
n 3‘ n
-; 2EIectro
‘ Cleaners

    
 

Detergent
C I e a ne r

    
 
    

\\\

     
  

\\\

   
   

  
 

C
a.
2
U \\\\\\QC\

15

0.7 percent sulfuric acid. This acid may produce some small
amount of etching to make the plating stick better but the
primary purpose is for neutralization of the alkali from the
electrocleaners (Metal Finishing Guidebook Directory, 1960).

After the acid pickle and clear-water rinse the
fixtures go into the copper strike which is the first of the
plating solutions. This copper strike solution puts a thin
initial coat of copper on the castings to avoid blistering
and then they go to the bright copper strike solution where
a thicker coat of copper is applied by the electroplating
process. Copper is applied in a copper cyanide solution and
this is the source of the cyanide waste. After the bright
copper strike the fixtures go through three clear-water
rinses and this rinse water is sent to a special cyanide
treatment tank before release to the river.

Next comes another electro cleaner and clear-water
rinse and then another acid dip, this time into 2 percent
sulfuric acid. After another rinse, the fixtures are passed
through an electroplating solution of a nickel salt and a
layer of nickel is deposited.

After nickel plating and three more clear-water
rinses the fixtures are placed in the chromium plating vats
and a layer of bright chrome is applied to produce the
finished product. The most suitable chromium compound for

electro-deposition is chromic acid, Cr0 used in aqueous

3,
solution with small additions of other substances (Richards,

16

1946). In the best plating form the chromium is in the
.hexavalent state which is extremely toxic to living
organisms.‘

After the chromium plating the fixtures go through a
final series of rinses, shown in Figure 2, on the ground
floor and then are taken to the second floor for two more
clear-water rinses and a demineralized water rinse. The
finished product is then removed from the racks and the
racks are returned to the ground floor to be cleaned with
nitric acid before being loaded up again for another plating
sequence.

A similar operation is carried on in the barrel
plating section but barrel plating is done in smaller
batches for special pieces only.

Figure 2 illustrates the way the tanks for the en-
tire plating operation are set up on an iron grating over a
large holding tank so any spills or overflows in the oper— 9
ation go into the holding tank below. Small arrows show
where rinse waters overflow into the holding tank as a
normal part of the operation. Tanks holding stronger so-
lutions may beexpected to lose solutions from splashing,
drippings, or acCidents. .In addition, all acidic and alka-
line tanks are dumped into the holding tank approximately
once a week (Caltrider, 1968). Treatment of wastes seems to
be carried out in an incomplete manner and without suf-

ficient controls to assure adequate results.

17

Cyanide is removed in the three rinses after plating
in the copper cyanide solution and this rinse water is sent
to one of a pair of concrete tanks built for cyanide treat-
ment. The two tanks are filled on an alternate basis and the
full tank is treated by addition of chlorine and sodium hy-
droxide while the other tank is filling. After treatment,
the tank contents are tested for presence of cyanide and
then released to the settling lagoon. Some cyanide still
seems to get out but this is an improvement over the old
method where cyanide was treated by a continuous stream of
hypochlorite in one large flow-through tank.

Hexavalent chromium is treated by addition of sodium
bisulfite to the first rinse tank following the chrome
strike. Since hexavalent chromium is a yellow-orange color
and trivalent is green, one can make a crude analysis by eye
and this is the way it is done in the plant. Plant personnel
add the sodium bisulfite to reduce the chromium at a rate
such as to keep the rinse water "sufficiently" green by
visual inspection as it overflows to the holding tank below
(Caltrider, 1968).

Contents of the holding tank are pumped to an out-
side flow-through tank with two baffles so sludge can settle
and floating oil can be skimmed off the top manually. This
sludge and oil is dumped into a waste pit beside the tank
and the liquid waste goes on into the lagoon with the

treated cyanide wastes.

18

Metals such as copper, chromium, nickel, and zinc
are commonly precipitated by pH manipulation in other plating
plants but this is not the case here. The copper solution
may receive this adjustment during cyanide treatment but is
evidently not retained in the cyanide tank long enough for
precipitation, if the proper pH is reached. No pH adjustment
is made in an effort to precipitate the other metals.

The final phase of treatment occurs in the settling
lagoon where metals and any solids have some chance to settle
out. The settling pond is normally about 50 x 50 feet in
area by 4 feet deep (Robinson, 1964) but, on one survey I
witnessed, the lagoon was filled up to within six inches of
the surface. At this time the flow-through tank was also
nearly full. Settling in the lagoon obviously removes some
of the metallic waste but the effluent is still a turbid
blue-green in color. A 2 x 6" board skimmer is set up at
the outfall to hold back surface oils but this is not always
effective. Numerous stains show where the lagoon has over-
flowed so oils and metal wastes are carried into the river.
With all treatment facilities working normally the effluent
of the plant flows overland for about 75 feet from the lagoon
to the Red Cedar River where its milky, blue-green presence
may be traced visually for several hundred yards in the

stream.

METHODS
Sampling Program

The object of the study was to determine effects of
the metal plating wastes on the aquatic ecosystem, and spe—
cifically to detect any chfihges in numbers or types of in-
vertebrates. In order to determine the effect of the plating
wastes it was necessary to analyze the water for plating
wastes both above and below the plating plant and to take
enough bottom samples to adequately characterize the inverte-
brate communities in the stream, both above and below the
plating effluent.

Such a sampling program was planned to give a de-
tailed picture in those areas most affected by plating wastes
and in the area of recovery. Sampling could be less intense
in areas where recovery was more complete. (Since there was
no basis to judge howaar the zone of pollution reached and
where the recovery began a pilot study was conducted during
the summer of 1965 to provide a general idea of the inverte-
brate distributions in the river before the sampling program
was made final.

The pilot program was undertaken with the idea of

sampling only invertebrates so no chemical analyses were

19

20

included. Eighteen sample sites were selected with no at-
tempt at randomization. Instead, sample sites were picked

so as to be somewhat concentrated in the area of the plating
waste effluent and, when possible, were picked so as to be
easily accessible. Three of the sites were upstream from

the plating plant, the farthest being just below Van Buren
Road bridge 0.94 miles above the plating plant effluent. The
other 15 sites were scattered downstream from the plant, with
the last site 100 yards below Dietz Road bridge, 10.32 miles
below the plating plant (Table 2).

At each sample site four sample spots were laid out
on a transect across the river at a right angle to direction
of flow. One sample spot was located four feet from waters
edge on each side of the stream and the other two spots were
located one third of the distance from each side of the
stream. Each of the four spots on the transect was the site
of a bottom sample.

During the period from 30 June to 16 August, 1965,
bottom samples at each of the four spots on each transect
were taken with a Petersen dredge designed to sample 0.68
square feet of bottom. At the time of sampling, the dredge
contents were screened in a 30 mesh screen box to reduce the
amount of fine sediment and debris and the invertebrates and
larger debris were then placed in sample jars and preserved

in formalin for return to the laboratory.

 

mm.0H 0H

21

 

0m.0a 0m.0H 0m.0H 00.0H m0 ma.0H 5H
ma.0a 00.0 0H.0H . 0H.0H NH 00.0 0H
H0.0 0H.0 m0.0 00.0 Ha mm.0 ma
m0.0 m¢.5 05.0 H0.0 0H 5m.5 0H
0N.m H0.m 00.0 00.m 0 mm.5 ma
NH.¢ ma.¢ 50.0 0m.¢ 0 Hm.0H on 00.0 NH 00.0 NH
m0.m m0.m m0.m m0.m 00 0m.0 on 00.5 Ha 00.0 Ha
H5.m v5.m mm.m 00.0 5 mm.5 0» m0.m 0H m5.m 0H
00.0 50.0 m0.H v0.0 0 00.0 ou 00.0 0 05.~ 0
m0.0 00.0 m0.0 00.0 mm 00.0 On 00.m 0 05.H 0
m5.0 00.0 0m.a 50.0 m m0.m Op N5.m 5 0N.H 5
0m.0 50.0 mm.0 m¢.0 v H5.m 0» 0m.a 0 mm.0 0
wa.0 0H.0 0H.0 0H.0 mm mm.a ou 00.0 m 00.0 m
0H.0 0H.0 ~H.0 0H.0 m m0.0 on 50.0 0 NH.0 0
m0.0| m0.0| no.0: m0.0| H0 00.0 00 00.0 m 00.01 m
mm.0| m¢.0| 0H.0| 50.0: m 00.0 0» m¢.0l N 0v.0| N
00.0: 00.0: m0.0| 00.01 H 00.0: 00 00.0: H «0.0: H
0 mmwumm m moflumm m mmfluom H mmflnmm cowumum muHEHq mmum coflumoon cofiumum
amasmmm Emonum

mwflumm EOGCmH “masmwm mmflumm DOHHA

 

 

.Hammuso panda 0cflumam 3oamn meHE CH muuwmcmnu oHQEmm mo coHumuoq .N manna

22

In the laboratory the preserved samples were washed
and contents were sorted by hand. Invertebrates were
separated by family, counted and preserved in a solution of
70% alcohol in 2 ounce sample bottles. After counting, the
invertebrates were identified to family and weighed for wet
weight on a four-place Mettler balance. Using a method
which King and Ball (1964) found to give a very close ap—
proximation to live, wet weight, I removed all invertebrates
from the alcohol and soaked them in water for 30 minutes to
counteract dehydration from the alcohol. They were then
placed on small 30 mesh screens and spun for 30 seconds at
1,800 rpm on a small clinical centrifuge to remove excess
water before being weighed on the Mettler.

Results of the pilot sampling series, presented in
the results section, were then used to determine the
character of the invertebrate populations and were thus used
as a guide in laying out the final sample sites.

Based upon results of the pilot study, the river was
divided into 12 sections for the final sampling program, two
sections upstream from the plating plant and 10 sections downr
stream. The sections, designated with the help of county
maps and 1/5000 scale aerial photos, were not of equal size
but were chosen so that there were smaller sections, and thus
more samples, in the areas of gross pollution and initial

recovery (Figure 3 and Table l).

22

In the laboratory the preserved samples were washed
and contents were sorted by hand. Invertebrates were
separated by family, counted and preserved in a solution of
70% alcohol in 2 ounce sample bottles. After counting, the
invertebrates were identified to family and weighed for wet
weight on a four-place Mettler balance. Using a method
which King and Ball (1964) found to give a very close ap-
proximation to live, wet weight, I removed all invertebrates
from the alcohol and soaked them in water for 30 minutes to
counteract dehydration from the alcohol. They were then
placed on small 30 mesh screens and spun for 30 seconds at
1,800 rpm on a small clinical centrifuge to remove excess
water before being weighed on the Mettler.

Results of the pilot sampling series, presented in
the results section, were then used to determine the
character of the invertebrate populations and were thus used
as a guide in laying out the final sample sites.

Based upon results of the pilot study, the river was
divided into 12 sections for the final sampling program, two
sections upstream from the plating plant and 10 sections down-
stream. The sections, designated with the help of county
maps and 1/5000 scale aerial photos, were not of equal size
but were chosen so that there were smaller sections, and thus
more samples, in the areas of gross pollution and initial

recovery (Figure 3 and Table 1).

23

Figure 3.- Map of final study sections.

24

 

 

 

 

.
a E Red Cedar
- gé 5”
i Q 0 N "' Midd'. .fgnC“
2 ° "
«I
e
r.
.
.” Branch
0
0
" KC'CMim—V
Wolf CreOl‘ '
590 Creek A

Dog; Creek K

 

25

For each of four sample series conducted throughout
the summer of 1966, one sample transect was chosen at random
in each section of the river. This was done by measuring
river distance in each section in eights of an inch with a
map measurer on 1/5000 aerial photos. A table of random
numbers was then used to choose a number between zero and the
total number of eights of an inch in each section. This
number was then the number of eights of an inch measured on
the photo from the upstream boundary of a sample section to
the sample transect for that series. Thus, for each of the
four sample series there was one randomly-chosen sample
transect in each section of the stream (Table 1). Four
dredge samples were taken on each transect as described for
the pilot study.

In addition to the 12 randomly—selected sample
transects, there were also five other sample transects,
termed permanent sites. These were selected arbitrarily and
were sampled at the same place during each of the four
sample series.

During the four regular sampling series invertebrate
samples were taken at four spots on each transect and samples
were handled in the same manner as the samples taken for the
pilot study. But, during the four regular sampling series,
water samples were taken in addition to the bottom samples
and these water samples were returned to the laboratory for

chemical analysis for presence of plating wastes.

26

Duplicate water samples were taken in acid-washed
polyethylene bottles at each sample transect and additional
samples were taken directly from the plating plant effluent
at various times during each sample series. Analysis for
hexavalent chromium and cyanide was made within 24 hours.
Samples were then acidified with 1 ml nitric acid per 500 ml
of sample to prevent precipitation of metals and samples were
stored for analysis for copper, nickel, zinc, and total

chromium at a later date.

Chemical Analyses

Hexavalent chromium - Hexavalent chromium concen-
tration was determined in the water samples within 24 hours,
using the diphenylcarbazide colorimetric analysis given in
APHA (1965). Color development was measured by a Beckman DU

spectrophotometer.

Cyanide — Cyanide concentration was measured within
24 hours by the pyridine-pyrazolone colorimetric method de—
scribed in APHA (1965). Color development was measured by a
Beckman DU spectrophotometer.

Although colorimetric analyses were used for hexa-
valent chromium and cyanide, analyses for copper, nickel,
zinc, and total chromium were done on a Jarrel Ash atomic

absorption spectrophotometer.

Copper - Copper was measured directly from the water

sample with no prior treatment. Standards were made up in

27

concentrations such as to bracket the levels of copper found
in the unknowns and then standards and unknowns were run at
the same time and concentrations of unknowns were plotted
from a standard curve. Copper was determined by atomic ab—
sorption using a wavelength of 3247 millimicrons with a fuel

setting of 10 psi and an air setting of 15 psi.

Nickel - Nickel was also measured directly from un-
treated water samples. Standards were made up from NiSO4 to
bracket the concentrations of nickel in the river. Standards
and unknowns were run at the same time and a standard curve
was set up to determine concentration of unknowns. Samples
were run on atomic absorption at a wavelength of 2318.5

millimicrons with air and fuel pressures both set at 15 psi.

Zinc - Zinc was also measured from untreated water
samples and unknowns were determined by use of a standard
curve. Standards made from zinc wire were run at the same
time as the unknowns using atomic absorption at a wavelength

of 2136.5 millimicrons and a fuel and air setting of 15 psi.

Total Chromium - Total chromium was also measured
using the Jarrel Ash atomic absorption spectrophotometer but
samples were first digested with nitric and perchloric acids
by the following method. Twenty milliliters of nitric acid
was added to 50 m1 of sample and the combination was boiled
down to about 10 ml in a 250 ml Phillips beaker with a glass

bead. Then, five m1 of perchloric acid was added and this

28

combination was boiled down until the solution was clear and
dense white fumes were given off. This solution was then
cooled and diluted to 15 ml with deionized water.

Preliminary experiments indicated that some inter-
ference in the river water, presumably phosphates, caused
lower values for total chromium standards in river water
than in deionized water. But, addition of calcium carbonate
got rid of the interferences so all samples were diluted
after digestion with an equal part of 1000 ppm calcium carbon-
ate to produce a final sample with 500 ppm CaC03. This final
sample was run on the spectrophotometer using the flame
emission method at a wavelength of 4255 millimicrons and air
and fuel pressures of 15 psi. Standards made up from K2Cr207
were run at the same time after being put through the same
digestion and dilution procedures as the river water samples.
Values for the river samples were then obtained from a

standard curve.

RESULTS AND DISCUSSION

Water Analyses

Maximum concentration limits were set for copper,
nickel, zinc, hexavalent chromium and cyanide in the Utilex
Company effluent by an Order of Determination issued by the
Michigan Water Resources Commission on 28 January 1953.
Limits set were: 1.0 ppm copper, 1.0 ppm nickel, 0.5 ppm
zinc, 2.0 ppm hexavalent chromium and 0.6 ppm cyanide. No
limits were set for trivalent or total chromium. Since 1953
the effluent has been monitored 43 separate times by the
engineers of the Water Resources Commission and waste limits
have been exceeded a total of 126 times during these sample
series. These limits were not only exceeded in the effluent
but were often exceeded in the river itself, even after di-
lution. Table 3, listing maximum and minimun values for
chemicals found in the river by Water Resources Commission
engineers, does not show all violations but does give an idea
of concentration ranges in the river.

Since the company has such a long history of vio-
lations during Water Resources Commission surveys, it is no
surprise that allowable levels were exceeded during the present

sampling program. Results for each of the six elements

29

30

.COHHHHE H00 mpnmm Ga mum mm ummoxo madam> HH¢

.mcofiumcHEHmump mo Honasc pmufiefla 0Hm> m so 00me UGHN

 

mm.o mm.o 0.0 0.0 0.0 0.0 0.0 0.0 o.o «.0 0.0 mm.o m.e m.oa A.A o.m Amm.m0
nmom HmEmHo

o.o mm.o o.o mo.o 0.0 «0.0 H.o H.o o.o m.o 0.0 «.0 o.m v.0H 0.5 m.s Amm.m0
emom scum

0.0 -.o 0.0 No.0 0.0 e~.o H.o H.o o.o m.o 0.0 «.0 H.e H.0H 5.5 o.m Aan.m0

pmom comaonoflz

0.0 mm.0 up 0H.0 0.0 0m.0 0.0 H.0 0.0 0.0 0.0 mm.0 0.H 0.0 0.5 H.0 Amm.H0
0wom muomwuw

0.0 00.0 m0.0 N.0 0.0 0.0 0.0 N.0 0.0 0.0 0.0 H.H m.N 0.0H m.5 m.0 AON.00 cmom
Hm>flm cumuw

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0H 5.5 0.0 dewub 0>OQE

 

 

 

sou rmflm seq amflm sou amflm sou amflm sou roam sou amflm sou amflm sou amflm AHHmmuao
30HmA mmfimzv
26 no Hmuos one an Hz :0 .o.a mm a
I BE

  
 

.500H ou mm0H Eonm oownmm mafiusp :oflmmwseoo mmousomwm H

m
cmmasoflz an moms mcoaumcHEkuwn knumwsmno Hmum3 H00 mmsam> Edsflcas 00m 555 umz

flunm
S .m anmH

31

measured are here given separately, but it must be remembered
that concentrations of single elements may be very misleading
because of the possibility of additive and synergistic ef—

fects of the same elements in combination.

Copper - Results of analyses for copper concen-
trations in the water are listed in Table 4. Copper could
be detected in the river water for at least 10.3 miles down-
stream in all four sample series. If means of copper concen-
trations downstream are compared with means of all samples
upstream from the plating plant, we find that concentrations
downstream through section 11 are significantly higher by a
t-test at the 0.05 level. This would be for approximately
8.5 miles.

Just because an element such as copper can be de-
tected in the water in above-normal limits does not neces-
sarily mean it will have a harmful effect. Accumulated bio-
assay data do indicate harmful effects attributable to copper
in aquatic systems but toxicity depends to a great extent up-
on associated factors such as water temperature, hardness,
and dissolved oxygen. These associated factors are often
not defined in bioassay reports so much of the data from the
past is not comparable and cannot be used to set realistic
limits of pollution.

Some workers have found concentrations from 0.1 to
1.0 mg/l copper to be non-toxic to most fish while others

have found concentrations from 0.015 to 3.0 mg/l toxic to

32

 

 

 

Table 4. Average parts per million copper in each sample
series .
Series Series Series Series Mean of
Sample 1 2 3 4 all series
1 0.015 0.010 0.000 0 085 0.028
2 0.005 0.005 0 000 0.070 0.020
9-1 0.005 0.005 0 005 0.060 0.018
Utilex 0.870 1.250 1.345 1.530 --—
Utilex 2.220 1.950 1.750 1.460 ---
Utilex -—— 1.600 -_- _-_ -__
3 0.180 0.250 0 285 0.140 0.214*
p52 0.158 0.290 0 250 0.140 0.210*
4 0.110 0.320 0 245 0.140 0.204*
9—3 0.085 0.185 0.130 0.175 0.144*
5 0.090 0.275 0 090 0.140 0.149*
6 0.075 0.380 0.060 0.240 0.189*
7 0 075 0.165 0.067 0.185 0.123*
p—4 0.080 0.125 0.070 0.170 0.111*
8 0.087 0.105 0 075 0.190 0 114*
9 0.040 0.037 0.030 0 110 0 054*
10 0.037 0.035 0.022 0.115 0.052*
11 0.040 0.030 0.020 0.130 0.055*
12 0.045 0.012 0.010 0.100 0.042
9-5 0.022 0.015 0.010 0.100 0.037

 

*
Denotes those means found to be significantly

greater than the mean of all samples upstream from the
plating plant.

(Determined by t-test at 0.05 level.

33

many kinds of fish, crustacea, molluscs, insects, phytoé
plankton, and zooplankton, especially in soft water (McKee
and wolf, 1963). Doudoroff (1952) found that synergism was
an important factor to be considered. He found that fish
could survive for eight hours with 8.0 mg/l zinc alone or
0.2 mg/l of copper alone but most fish died in eight hours in
a combination of only 1.0 mg/l zinc and 0.025 mg/l copper.
The extent of the synergistic effect is also affected by
hardness of the water and other factors. And, even if past
bioassay data are considered comparable and are used, they
are measures of toxicity and even less is known about the
deleterious effects various elements or compounds may have
upon growth and reproduction.

As an approximation to be used in the interim until
better data can be obtained, Tarzwell (1967) has suggested
that 0.16 mg/l copper be set as the limit to be allowed in a
waste anytime-anyplace. This would be the limit for copper
alone and the limit would be lowered, as will be shown later,
in the presence of other toxic substances. Even if synergism
is ignored and 0.16 mg/l is taken as the maximum allowable in
the Red Cedar River, we can see from Table 4 that this limit
is exceeded for miles downstream from the plating plant.

And, Tarzwell's 10ndt.is very generous compared to the limit
of 0.02 mg/l set by the State of California for fish water

and aquatic life (McKee and WOlf, 1963).

34

The maximum allowable concentration of copper in the
Utilex Company effluent was set at 1.0 mg/l by the Order of
Determination issued by the Michigan water Resources Com-
mission on 28 January 1953. Since that time this limit has
been exceeded in 28 of the 43 sample series run on the ef-
fluent by Water Resources Commission engineers and it was
exceeded in eight of nine sample series run on the effluent

as part of the present study.

Nickel - Nickel concentrations detected during the
sampling program on the river are shown in Table 5. Nickel
could be detected for at least 10.3 miles below the plating
plant in all four series and mean concentration values were
shown to be significantly higher (t-test at the 0.05 level)
in every section downstream from the plating plant when com-
pared to the mean of all samples upstream from the plant.

Although toxicity of nickel varies according to the
compound in which it is contained, it does not seem to be
nearly so toxic as copper or zinc in the aquatic environment
(McKee and WOlf, 1963). But, here again, simple toxicity is
not the only measure of adverse effect and little is known
of the effect upon growth and reproduction of aquatic organ-
isms. It is in this conservative vein that the Mersey and
Severn River Boards in England have established the limit
concentration for nickel alone, or in combination with other
heavy metals, at 1.0 mg/l total in industrial wastes (Klein,

1957).

35

Table 5. Average parts per million nickel in each sample

 

 

series.
Series Series Series Series Mean of
Sample 1 2 3 4 all Series
1 0.00 0.00 0.00 0.00 0.00
2 0.00 0.00 0.00 0.00 0.00
P51 0.00 0.00 0.00 0.00 0.00
Utilex 6.70 3.00 3.35 1.40 ----
Utilex 3.80 8.80 2.10 3.50 ----
Utilex ---- 9.50 --—- -—-— ----
3 0.70 1.50 0.40 0.08 0.67
p-2 0.40 1.50 0.37 0.08 0.59
4 0.50 1.20 0.42 0.10 0.56*
p-3 0.30 1.00 0.35 0.10 0.44*
5 0.70 0.70 0.48 0.10 0.50*
6 0.20 0.85 0.28 0.19 0.38
7 0.10 1.40 0.33 0.25 0.52*
9+4 0.30 1.10 0.31 0.20 0.48
8 0.20 0.90 0.30 0.20 0.40*
9 0.10 0.70 0.29 0.18 0.32*
10 0.10 0.50 0.28 0.19 0.27
11 0.10 1.00 0.25 0.20 0.39*
12 0.10 0.10 0.25 0.19 0.16*
P-5 0.10 0.10 0.24 0.10 0.14*

*
Denotes those

means found to be significantly

sareater than the mean of all samples upstream from the
plating plant .

Determined by t—test at 0.05 level.

36

A limit of 1.0 mg/l nickel was set for the Utilex
Company effluent by the Michigan Water Resources Commission
Order of Determination issued on 28 January 1953. This
limit has been violated in 37 of 43 sample series conducted
by the Water Resources Commission since 1953 and was ex-
ceeded in all nine sample series taken at the effluent in
the present study. In addition, the limit set for the ef-
fluent was exceeded several times in the stream itself three

or four miles downstream from the plating plant.

.gipp — Zinc concentrations determined during the
sampling program are shown in Table 6. Zinc could be de-
tected for at least 10.3 miles downstream from the plating
plant in all four series. Mean concentrations downstream
compared to mean concentration in all samples upstream from
the effluent were found to be significantly greater (t-test
at the 0.05 level) for every station downstream.

_As in the case of copper and nickel, toxicity levels
of zinc as determined by bioassays are not well defined and
reports are often conflicting. Part of the conflict may be
due to differences in temperature or hardness of the bio-
assay waters since descriptions of such parameters are not
always included in older bioassay results. Jones (1938)
found that for mature sticklebacks the lethal limit for zinc
in water containing 1.0 mg/l calcium is only 0.3 mg/l but as
much as 2.0 mg/l zinc is not toxic in water with 50 mg/l

calcium. If this is true, toxicity of zinc would be reduced

37

 

 

 

Table 6. Average parts per million zinc in each sample
ser1es .
Series Series Series Series Mean of
Sample 1 2 3 4 all series
1 0.010 0.015 0.010 0 000 0.008
2 0.030 0.020 0.000 0.000 0.012
291 0.010 0.016 0 000 0.000 0.006
Utilex 1.100 1.310 4.100 0.550 ---
Utilex 0.980 4.300 0.370 0.985 ---
Utilex --- 0.855 --- —-- ---
3 0 160 0.190 0.555 0.086 0.247*
p-2 0.145 0.195 0.510 0.086 0.234*
4 0.215 0.305 0.355 0 080 0 238*
293 0.190 0.145 0 195 0.120 0.162*
5 0 200 0.180 0.185 0.133 0 174*
6 0.060 0.190 0.085 0 110 0.111*
7 0.060 0.235 0.080 0.135 0.127*
994 0.135 0.200 0.055 0.130 0 130*
8 0.090 0.125 0.065 0.110 0.097*
9 0.045 0.095 0.040 0.065 0.061*
10 0.035 0.040 0.025 0 070 0.042*
11 0.020 0.040 0.015 0.050 0 031*
12 0.025 0.030 0.013 0.030 0.024*
P95 0.040 0.020 0.020 0.040 0 030*

 

*

Denotes those means found to be significantly
greater than the mean of all samples upstream from the
Plating plant. Determined by t-test at 0.05 level.

38

in a river such as the Red Cedar where bicarbonate alka—
linity ranges from over 100 to nearly 300 mg/l, depending
upon river discharge rate at the time (Grzenda, 1960).

As previously mentioned in the copper section,
Doudoroff (1952) found a synergistic effect in soft water
between zinc and copper. The Water Research Board of
England (Anon., 1960) tested for synergistic effects of
copper and zinc and found a synergism present in soft water
(15 — 20 mg/l as CaC03) but absent in hard water (320 mg/l

as CaCO Since the Red Cedar River is well above the

3"
soft-water range, we would expect little synergistic effect,
if these findings are correct.

For fish, concentrations as high as 4.0 mg/l zinc
have been reported not to have harmed trout exposed for 48
hours. Other workers have found 4.0 mg/l toxic at exposures
of three days (McKee and WOlf, 1963). Toxicity among differ-
ent kinds of fishes as found by various individual workers is
so varied it is very difficult to put a definite value on
zinc needed to cause death. It is even more difficult, if
not impossible, to say what concentrations produce harmful
effects on fish short of death.

Effect of zinc on invertebrates is even less well-
known than the effect on fish but some work has been done to
provide at least an estimate. Cairns and Scheier (1958)

stated that 0.79 to 1.27 mg/l zinc was the 96 hour TLm for

the snail, Physa heterostropha, in soft water at 20 degrees

39

C. and 0.62 to 0.78 mg/l at 30 degrees C. In hard water the
results were 2.67 to 5.67 mg/l zinc at 20 degrees C. and 2.36
to 6.36 mg/l at 30 degrees C. Anderson (1950) states that in
Lake Erie water at 25 degrees C. Daphnia magna were immobil-
ized at very much less than 0.15 mg/l zinc chloride.

With so many varying, or even conflicting, reports
on toxicity it is difficult to set realistic limits for zinc
in wastes. The Mersey and Severn River Boards in England
have adopted 1.0 mg/l zinc, alone or in combination with all
heavy metals, as the limit concentration of zinc in wastes
(Klein, 1957). Tarzwell (1967) suggests that 1.06 mg/l zinc
be the maximum concentration allowed anytime-anyplace and
bases this on 1/20 of average 48 hour TLm measurements. This
figure is modified, as will be shown later, when other wastes
are present in the same water.

The limit set on zinc in the Utilex effluent by the
1953 Water Resources Commission Order of Determination is
0.5 mg/l. This limit has been exceeded in 25 of the past 43
sample series of the Water Resources Commission and was ex-
ceeded in eight of nine sample series in the present study.
In addition, the effluent limit was exceeded twice in the

river itself below the plating plant.

Chromium - Ammounts of hexavalent chromium found

 

during the sampling program in Utilex effluent and the river
are shown in Table 7. _At the time of the Water Resources

Commission Order of Determination for Utilex in 1953 it was

4O

 

 

Table 7. Average parts per million chromium6 in each sample
series.
Series Series Series Series Mean of
Sample 1 2 3 4 all series
Utilex 0.85 0.03 0.30 0.045
Utilex 0.78 0.40 0.30 0.125
Utilex --- 0.27 0.03 ---
3 0.015

 

Hexavalent chromium concentrations were below de—

tection levels at all stations not listed in the table.

41

believed that hexavalent chormium was toxic to aquatic forms
of life but that the trivalent form was much less harmful.
With this belief as a basis, a limit of 2.0 mg/l hexavalent
chromium was set for the effluent but no limits were set for
chromium in the trivalent form. In order to meet the re—
quirements, the plant operators have been adding sodium bi-
sulfite to the chromium rinse water to reduce chromium from
the hexavalent to the trivalent form before release to the
river. Although done in a rather crude manner this has evi-
dently been quite successful because hexavalent chromium
limits have not been exceeded in Water Resources Commission
surveys since June, 1964 and levels were well within limits
during all nine sample series of the effluent in the present
study. Only once could hexavalent chromium be detected in.
the river itself and this was just below the effluent and at
a very low concentration.

If it was really true that hexavalent chromium was
the only toxic form we would assume that chormium was of
little consequence in the effluent or in the river. But,
more recent studies have indicated that there is evidence to
support the conclusion that hexavalent chromium is no more
toxic toward fish than the trivalent form (MCKee and WOlf,
1963) and some evidence even indicates that trivalent may be
more toxic, although there is some uncertainty about this.
Jones (1941) found the trivalent form to be more toxic to a

flatworm, Polycelis nigra, but Anderson (1950) found

42

hexavalent to be most toxic to Daphnia magna in Lake Erie

 

water so there may be some variation in toxicity between the
two forms of chromium with various forms of aquatic life.

In light of this evidence it appears that we must consider
total chromium and not just one of the forms.

With chromium, as with the other metals discussed
‘before, it is extremely difficult to set realistic limits
for discharges based on present bioassay data so proposed
limits vary somewhat. The Severn and Mersey River Boards
in England have adopted standards to limit total concen-
tration of all heavy metals, alone or in combination, in—
cluding chromium, to 1.0 mg/l (Klein, 1957). McKee and WOlf
(1963) have proposed the same limits for chromium (either
form) to protect fish life in California but feel that a
concentration of 0.05 mg/l must not be exceeded if other
forms of aquatic life are to be protected. Fetterolf (1968)
has informally acknowledged that this limit of 0.05 mg/l
chromium appears justifiable at this time considering the
chronic toxic effects of chromium. This does not allow for
presence of other wastes at the same time, as will be dis-
cussed later.

Average concentrations of total chromium found
during the present study are listed in Table 8. The limit
of 1.0 mg/l was exceeded in four of nine sample series in
the effluent and a limit of 0.05 mg/l would have been ex-

ceeded in every area of the river in each series. In every

43

 

 

Table 8. Average parts per million total chromium in each
sample.
Series Series Series Series Mean of
Sample 1 2 3 4 all series
1 0.07 0.08 0.05 0.10 0.075
2 0.07 0.08 0.09 0.10 0.085
p—1 0.10 0.08 0.17 0.10 0.112
Utilex 0.17 0.70 0.75 1.55
Utilex 3.05 0.50 0.80 2.60
Utilex ---— 3.40 ---- ---—
3 0.35 0.29 0.28 0.21 0.282*
p—2 0.32 0.26 0.30 0.21 0.272*
4 0.28 0.28 0.25 0.15 0.240*
p—3 0.25 0.14 0.26 0.20 0.212*
5 0.24 0.20 0.31 0.15 0.225*
6 0.11 0.41 0.24 0.15 0 227*
7 0.11 0.15 0.20 0.17 0.157*
994 0.36 0.15 0.23 0.20 0.235*
8 0.14 0.13 0.22 0.18 0.167*
9 0.11 0.11 0.14 0.18 0.135*
10 0.10 0.09 0.29 0.17 0.162*
11 0.10 0.09 0.12 0.15 0.115*
12 0.14 0.09 0.15 0.16 0.135*
995 0.13 0.10 0.19 0.13 0.137*

 

*
.Denotes those means found to be significantly

greater than the mean of all samples upstream from the
plating plant. Determined by t-test at 0.05 level.

44

downstream section of the river the mean concentration of
chromium was significantly higher (by t-test at the 0.05
level) than the mean concentration at all stations upstream
from the effluent. In past surveys by the Water Resources
Commission, chemists have regularly detected chromium concen-
trations over 3.0 mg/l and concentrations over 6.0 mg/l were
not uncommon. This is not only far above what are proposed
as safe limits for aquatic life but is also above the limit
of 5.0 mg/l proposed by McKee and Welf (1963) as the safe
limit for chromium in water used for stock and wildlife

watering.

Cyanide - A limit of 0.6 mg/l cyanide was set for
Utilex effluent by the 1953 Water Resources Commission Order
of Determination. Through 1965 this limit was exceeded 29
times in 43 series of effluent samples taken by the Water
Resources Commission. One sample in 1965 had a cyanide
concentration of 14.0 mg/l and many samples had cyanide in
excess of 4.0 mg/l. Since installation of the new batch-
treatment system for neutralization of cyanide wastes, the
water Resources Commission has reported one violation (1.0
mg/l cyanide) but all other tests were within limits and all
effluent samples in the present study were within the limits
set in 1953. See Table 9.

The fact that better control of cyanide treatment
has enabled the company to meet the 1953 requirements is en-

couraging but much of the available bioassay data indicates

45

Table 9. _Average parts per million cyanide in each sample

 

 

 

series.
Series Series Series Series Mean of
Sample 1 2 3 4 all series
Utilex 0.05 0.05 0.05 0.02
Utilex 0.00 0.00 0.01 0.15
Utilex ———- 0.00 ---- ----
3 0.02 0.00 0.03 0.03 0.020*
p—z 0.00 0.00 0.01 0.04 0.013*
4 0.03 0.00 0.03 0.04 0.025*

p—3 0.00 0.00 0.01 0.01 0.005*

 

*
Denotes those means found to be significantly

greater than the mean of all samples upstream from the

plating plant. Determined by t-test at 0.05 level.

Cyanide concentrations were below detection levels
at all stations not listed in the table.

46

that this limit may be too high to assure complete pro—
tection, especially during periods of extreme low river flow
when the effluent may furnish a quarter or more of the total
discharge of the river at Fowlerville. These periods of low
flow were considered in the 1953 Order of Determination which
called for a five to one dilution factor for wastes in the
stream (Denniston, 1966).

Lethality of cyanide is largely associated with the
proportion of CN’ which is in the form of HCN so toxicity
increases with decrease in pH. Toxicity is also increased
with elevation of temperature and reduction of dissolved
oxygen. Nickel tends to complex with cyanide to reduce
toxicity but zinc complexes are exceedingly toxic. This
variability in toxicity makes it hard to set absolute limits
for cyanides in wastes and especially hard to set limits in
the Red Cedar River when temperature and dissolved oxygen
vary from day to day and there are variable amounts of both
nickel and zinc present with the cyanide. But, as an approxi-
mation, results of many bioassays indicate a lethal effect
to fish in a few days at concentrations from 0.05 to 0.1 mg/l
CN? (McKee and WOlf, 1963). These values are for lethal ef-
fect and we still do not know the effects of lower concen-
trations on growth and reproduction of fish and other aquatic
organisms.

After reviewing much of the research on cyanides,

the Aquatic Life Advisory Committee of the Ohio River Valley

47

Water Sanitation Commission (1960) recommended that free
cyanide in excess of 0.025 mg/l be considered unsafe in the
waters of the Ohio River. Fetterolf (1968), while realizing
the complexity of the problems in setting limits for cyanide,
has suggested in an interdepartmental memo that 0.01 mg/l
may be a realistic limit and bases this on 1/10 of the TLm
found by various workers. This does not seem unreasonably
conservative if we consider that highest concentrations in
the Red Cedar River would usually come during periods of low
river flow when the water temperature is highest and dis—
solved oxygen at its lowest.

Even with the new batch-treatment system in effect,
some cyanide is released to the river (Table 9). This limit
of 0.01 mg/l was exceeded in five of nine effluent samples
in the present study and was matched or exceeded as far as

0.65 miles downstream from the effluent source.

Wastes in combination - It has already been stated

 

that individual wastes were found in the effluent and in the
river in concentrations greater than those allowed by the
Michigan Water Resources Commission and much greater than
those concentrations suggested as safe for aquatic life by
various workers in the fields of toxicology and water pol-
lution. But, this is only part of the story. Combination
of the various toxic wastes produces a more lethal environ—

ment due to simple additive effect and a much more lethal

48

environment if synergism, such as that between copper and
zinc or zinc and cyanide, is considered.

The effects of combination were recognized in England
when the Mersey and Severn River Boards adopted a limit of
1.0 mg/l copper or zinc or nickel, etc., with a maximum
total concentration of 1.0 mg/l (Klein, 1957). This is much
more restrictive than the system of setting limits for indi-
vidual wastes with no combination limit but it still does not
emphasize the fact that 0.5 mg/l cyanide and 0.5 mg/l zinc
may be much more toxic than 1.0 mg/l of zinc alone.

Tarzwell (1967) in his Interim Report on Water
Quality also recognized the presence of additive or syner-
gistic effects and recommended use of the following formula

to set limits for combinations:

Ca Cb .gp
La+Lb +Lnsl
Ca, Cb ... Cn are the measured concentrations of

toxic materials in the water.

La, Lb ... Ln are the respective permissible

concentration limits derived for each indi-

vidual material.
The formula can be used in two different ways. Concen—
trations and allowable limits may be given as 24—hour aver-
ages or as maximum amounts allowed "anytime-anyplace." In
setting the permissible limits (usually 1/20 of 48-hour TLm)

it is assumed that limits are chosen with an adequate safety

49

factor to allow for synergism., This may not prove to be
true in all cases but it appears that there has been no ade—
quate data to propose a better system.

Since concentration of metals from the plating plant
effluent is so variable in the Red Cedar River, and because
a "slug” of effluent could cause great harm in a very short
time, the "anytime-anyplace" limits are the most practical
ones to use in this case.

Maximum amounts of wastes found in each section of
the river are given in Table 10. If these amounts are used
in the formula proposed by Tarzwell, with the maximum allow-
able limits suggested by most recent studies, the resultant
sums exceed the allowable in every section of the river down-
stream from the plating plant effluent. In section three,
just below the plant, the sum is 14.84 and in section 12,
nearly 10 miles downstream, the sum is 4.7. Both are many
times the allowable sum of 1.0.

The formula proposed by Tarzwell is based on past
bioassay data and was prOposed with the realization that this
might not be perfect but is the best available at this time.
Another method will be proposed later in this thesis and com—

parisons will be made at that time.

50

Table 10. Maximum amounts of each waste found in each
section of the river.

 

J t
__‘ r

 

 

L _
*— ‘_:

 

 

 

 

 

Miles to Plating wastes in parts per million
Section section center Cu Ni Zn Cr6 Cr CN
Total
1 -0.69 0.08 -- 0.01 —- 0.10 —-
2 -0.22 0.07 —- 0.03 —— 0.10 --
Effluent 0.00 2.22 9.50 4.10 0.85 3.05 0.15
3 0.13 0.29 1.50 0.56 0.02 0.35 0.04
4 0.45 0.32 1.20 0.36 -- 0.28 0.04
5 1.10 0.28 1.00 0.20 -— 0.31 0.01
6 2.13 0.38 0.85 0.19 -- 0.41 --
7 3.27 0.19 1.40 0.24 —- 0.36 --
8 4.36 0.19 0.90 0.13 -- 0.22 --
9 5.41 0.11 0.70 0.10 —- 0.18 --
10 6.74 0.11 0.50 0.07 -— 0.29 --
11 8.57 0.13 1.00 0.05 -- 0.15 --
12 9.95 0.10 0.24 0.04 -— 0.19 --
Maximum single limits
for wastes anytime-
anyplace 0.16 1.00 1.06 0.05 0.01
Tarzwellts formula for Cu Ni Zn _Cr CN
combination“ 6.16 + "II-0'6 “hi—706 + 6.05 + 0 01 " $1
2:22.225‘1322 £232: 1“ —— + —— + —— + —— + 33:44...

 

With maximums found in 0.10 + 0.24 + 0.04 + 0.19 + 0.0Q_ 4 70
section 12 of Red Cedar 0.16 1.00 1.06 0.05 0.01_ °

 

 

51

Invertebrate Sampling

Traditional methods of analysis - In the case of the
Red Cedar River, as in most instances of pollution, the pri-
mary point to be considered is the effect on the biological
community. A combination of chemical determinations and
past bioassay data would enable us to make a rough pre-
diction of effects in the river but it would be almost im-
possible to assess the effect of the infinite combinations
of physical factors throughout a year or more. Because of
the variability in concentration and types of pollutants,
Hynes (1963) and previous workers have considered that the
relatively immobile invertebrates are the best indicators of
pollutional effects. They require much more time to repopu-
late an area than do fish and they give a good indication of
the least favorable as well as average yearly conditions in a
body of water.

Although there is wide-spread agreement as to the
validity of using invertebrates as indicators of pollution,
there is some disagreement as to how the data should be
handled in order to best detect pollution effects. In this
study the invertebrates were identified to family and ef-
fects of pollution were assessed by some of the frequently
used methods such as indicator types, numbers of families
present and community types. In addition, a more easily
used index of pollution is proposed which may be compared to

more time—tested methods.

52

A basic theory in ecology is that a natural, un—
polluted environment will contain more niches, and thus sup-
port more types of animals, than will an environment which
has been made less favorable due to pollution or other arti-
ficial changes. Thus, in a clean, unpolluted stream, con-
ditions would be favorable for many families of inverte-
brates. But, if the stream becomes polluted, the conditions
are such that only part of the original inhabitants are able
to survive. With fewer families present those remaining
have less competition so are often found in much greater
numbers than would be present in a clean environment.

This is exactly the situation found in the Red Cedar
River (Figure 4). Upstream from the plating plant effluent
the river averaged about 14 families per dredge. Just below
the effluent the average number of families dropped to two
and family numbers increased gradually downstream as plating
wastes became less concentrated.

In addition to the change in numbers of families
present in various areas of the river, there is also a de-
finite pattern shown in the recovery of families downstream
from the source of pollution. This is demonstrated in
Tables 11-15 which show the numbers of individuals in each
family upstream and downstream from the plating plant out—
fall. In the upstream sections many families are present
and none have excessively high numbers of individuals. Just

below the plant there is usually only one family present,

53

Figure 4. Mean number of invertebrate families found in

each study section.

54

20-1-

18*-

16v

4

ae_.een set

2
I

P D ’ p
1 ‘ d ‘

o a 6 4

ne...Ee.._ «O seas—:2 sees

an

10

Dlstanee Downstream In Miles

Tables 11 -

15.

55

Numbers of invertebrates per square
meter in each sample series.

T Pollution tolerant

F Facultative

I = Intolerant to pollution
Classification as to tolerance to
pollution is based upon past ex-
perience of personnel of Michigan
Water Resources Commission
(Personal communication with Ron

Willson, 1968).

56

00

0d
ON

000
00

verve-0v

NH

0H

00

NH

05
nv

55¢
00

NH

Nd
Nd

Nd

0H0
H5

§'§.0'OQ

no 00
nv
NH

C‘fi'.

O
‘
¢
'0
H

N50H NOH 50
0a 0 0a N0 0 ON

mn 0N Nd 0 Na

00 00 0 NH
05nN man 000 00 v v 50

000 down 00¢ HNNw 00¢.0H anm.¢m HHN.0H

 

00.0H
0H.0H
5d

00.0a
00.0
0H

5n.0d
00.0
0H

06.NH
50.5
#6

005N

000

000
ON

NH
505
00H

no
00

000H

00

mm
00H
00

mm
«5

00

0N
ON

mevv

0H0
00H

HH

Hrduanuurqu ~1H

E4CuII.hmhmhhhhhhhhhhhfifihhhHhHo-tHFBPNhHHhhh

namumfl 0000000
oocauuaou coausaaom

 

 

Influenced;
envahena>a>
oeoawsabuueuum
eeoaauea
unvauceoauae:
oauasseaoao:
enpwasuAu9620
escapauaz
eepanmuhm
eac~a>zm
occasacaonesa
omnwcoau0<
omenuoan
oupanuocoaa
uncsaauaouosa
nuoasnae<
«0002:0004
eevaaesoeauco
oevacwu>o
acquaueav>m
eav«:0uumacoou
ompusdm
onvwcomoQOuauou
encasedoand
onuauousmz
empaxauoo
006.400:
eapaaulazmm
unpsdmam
empaoaau>0
owcanuxamouOSx
unpaaaasan
omodflaraouuss
066025556
wapaaouxm
eopacmnme
oceanoaaoaos
06600200028

 

 

 

noun! one: agosws< uo euonasz
00.0 ma.0 mn.v 00.4 05.N no.N 00.0
v0.0 N0.n n5.N 05.N v5.d 0N.« n0.0
NH dd OH 0 0 5 0
usoam mcuuaam radon aucnueun use escaunum

00.0
0N.O

0H.o
NH.0
0

00.01
00.01

H5.01
$0.01

0¢.a1
N0.01

muouoEost
00A“:
sunfisz COAunum

umsmsd ma
1 0:30 On
mmna
255000 0.0150

57

 

M "NH HH

H
h
h
H
h
H
h
H
h
h
h
h
h
h
h
h
h
h
h
h
h
a

 

eaoauelesau
eeoaueaa>a>
eevaxauoo
o-oausnaooaanm
neonanaa<
oaoaaanaocasa
ennuaoooQOuauoO
eeoaahsozuaem
eeowzo>enouvsn
eeauuasoueo~oa
oauaasu
eao«a%ao«uauum
eooacau>o
consensuaa
unassuuam
eeawaassam
«massages
oooaaanaouosn
eaoacOAumaceou
oooaooaaocoe
oaossnam
eeuauauanaa

acquo oaeocoxma

 

 

ms
o
a
e
on o v
o e
on
v
saw an «N o v
a v
a o c
a a
v as v o
am one
«a v
o o e 0 mu 0
o v «a e v a
can mm m5 mu «m as ma «a «a 55« sea
as am as a o a e
«44 Boss mnaa own own ommm m5 «om amen one» one.un mam.o~ macs
wave: can u eueldat no euenlfll
m~.ma m~.na mm.oH mv.m vn.5 o«.m mm.¢ m~.v a..« mo.« «5.0 n~.o n~.o
oH.oH ¢~.o Ho.w oo.m mm.v no.n mo.~ em.~ 50.6 no.0 ne.o e«.o ¢«.o
ma Ha ca 0 a an 5 u a nu e «a a

uaeam unwuanm laden eoaaueqa as: eaouuauu

ueAESI Goduaum

>058 m - 0:58 «N
moms

a mmummm

58

0d
00¢

0N
«fin

and

ON
0N

O¢N
NH

00
man

Nn

0d

8

N”

no
Na

50
NOON

O5OH

.iN
H

i¢¢O¢N0
v4

:2
:02..

ON
«0

'
fi'iO‘

nv O¢N 0 0 ON
0
«0H 0N 00H
Nd ON 5v 0 ON 50

v
v
v
«0 v 0 5*
N
0

 

00.0d
0H.OH
Nd

0

«10.00 000.00 500.Nv «0d.an NN0.00 #555

0N0

NH
500
500

#0«

ON

«1v

0N
00H

005
HO0H

«0

«N
NON
nHN

HHH

H
Est.ulhhuhluh.hthh.hlnhvdhahunk-Hrun.hrdmnuh.HiahihthhohluhnHlufith

 

eanuox Mddmde

ouaaueHOu sofiusaaom

 

unequaEOuaern
encaaconaaauodo
eepanuocnam
eevwwcoonuaom
unusuaau>a>
empaaunaoam
060035084
oaoaauosuoo
envious:
encaaauAunEuq
0000456020500
eapauuoanmm
cavauwnaouoxrm
eaoano>e00upam
eauaueeonam
enauaumuuhm
onaaaazaocaaa
eeuuael0n>uzu
eepwunem
confiscaumt
caoasaunouvss
eavacau>o
oepaueam
0664:0600
onoaeHm
vacuueUOLQoq
uneasdnmouo>c
0500:6909
unpauuwo
0660000050
eepaazaaonaq
onesaaasmm
eouaxwuoo
«apauaaau
unpacODOQODMuoO
onp«:o«umncoou
ompdassm
oncaaoaaacws
oocssausm
unpauauanaa

maouo vusocoxda

 

 

naunan< no acne:-
¢~.~ no.“ 06.6
mn.a no.0 no.6

00.0
50.0

0d.0
00.0

0N.O
0H.O

u

00.0 00.N

N0.n ne.«

0 vb 5 0 0 00 a N0
caldm mawuaam radon evacueaa can anewuauu

00.01
00.01
«A

«0.01
0H.01
N

00.HI
00.01
H

aueueao~ux
nodal
noneaa newuaum

unsms< o
1 uns0a¢ «
000a
N muumum

59

ma

wMN

mp
00

«OHN
ma
ONWH

mmN
mvNH

0mm

mm

NH

ON
ownh
Hm
mm
mm

‘5"

mm

own
OOH
Nun

OH
hv

mm

an

ON

an

mm

0N
vN

hONn

v
NnOa
omNd

ON
NH

mm
mNN

mm
ON
«0
mm

Odd
as

mmv

«

0
mnm
NH
0500

NH

mON

ON
mnOa

an
0d

Nn
Hm
vN

hm

mOfl

NH
m¢OH
0*

H¢w¢
Hm
NH

«mm
mmn.wa

QVQ'Q'fl

 

hm.ma
Om.OH
mm

mm.ma

vm.m
NH

#h.¢d mm.dH

0a.m
HH

nv.h
OH

uncut mcauuam onom oocauuan can «noduaum

 

ufiuwt onus m no

¢0N
hm
ma #
v
Hmwa HON

Hm ¢

awn mm

mm.n mn.d
hN.N v0.0

0

(DV‘CDCDQQ‘Q
V‘N

v

¢
Hahn Nnm

v

¢ 0d

ma 0
NOH NOH

Hmv.na ham.NN mmm.mN N¢n.mn nmm.mNa n¢m.aaa nOa.nm va.aOa

(av

‘90 fild‘

0H

dw¢

HONH

Had
mama
ON
meN

ma
«N

i‘!’

Ova

Na
Na

ma
mm

ah

Nan
NH

nm¢
coma
mva

an

v
HHH

mm

HHH

an
hmv H

v0

055

m¢h

ma
ondm H
«mm
¢Nm
hon

thfinflah‘uNil-(uthis.HhhhhhhhhhHh-hhE-‘HhhHmE-‘hhHHHIthHHh

omufiasodunuxg
oac_n>ca
oncudcomouaw:
o¢©l~>uc4
omu.-u>omwz
«mnfiuouanmm
«m04ummm
oqunu>aaouaux
ouoddaaue
vouanuohmHm
omvuacozmanmoHo
umoaaoudced
monuaflnawncaaq
oucauauuqm
man~=o-mmzm
annuaheonu>mm
mmvauuwmcmm
wmnqcuamauuv>$
macaunfiu>r
mnnacmaoo
wmnaququmA
unuuwumcmHm
unvacflu>0
mmcaamam
wmvwaamvnomum
umnacmnmfi
oavaouaau
onflguuwCOuoz
mmoaum I Jumgwm
unnacamm
uuvchanmd
wmvuxauou
omuaausa0uo>x
mavacomoQOuauwu
vandaaaawnflq
macad>50£0>mm
wmcacmmnncmh
unflaawhom>ucu
mandamu>m
anomanmnd
mmnacaw
wmnucodummcmou
wmnuoauqnne

 

manumz uHcmmuO
mucouwaou nodusaaom

 

macaw DabchzmH

 

nadfiacc no ummamz

mO.H 05.0
00.0 hv.O

mm

m¢.an mumumEoHux
Nm.ou moaaz
H umn532 codumum

Hansoumom 5N

u nonEwummm .4
mwma

n mmHmmm

60

m5

0N

50

HmH

N0

0v0H

00¢

HvON

V"?

00

NOH

N0

000
m0

«00H

005

00vN

05v
0N

00

0NH

00H

50 NH

000 050v
NOH ON

00N 00H

000.vH 00H0

¢¢N
N0
55NH
000

ON

NmN
HHO.vH

VQ'Q' 000‘
N

00
00H
NH ON

v NH
000 000 000 O¢N 000
000 00H 0¢ v

0 ¢ #N v

00 0 «N 0*

q:
Q)
0""

000
5000

mON
QQO.NH

0N0 H0 H0
0000

NH NH
00H.00 0v0.0m H0O.N0 N00.0Q N00.00

 

50.0H
O0.0H
0m

00.0H
0H.OH
NH

om.¢fi
ao.m
.H

\

0H.HH 0¢.0
00.0 «N.0
OH 0

00.0
NH.«
0

'0

N0

NvH

005

m0
0H

mm
«a

ON
6H0

100

500
v

050H
OHH

0H
mNN

00

NH
0H

#'

0N

NH

NOHN

ON

000

mHH
mwN
HmH

 

 

HH HHH HHHH

H
thhh‘uhfuhhfiahhhhhhhfi BhHhhhHhhhhHHhHHHhIoF-‘E‘h

oaouuaaa>u>
oooun>cm
occuucouaumo:
OQUHHOUACE<
onoucuuxu
oovuawsn-ce.q
onuHHCEdsmu
oaoauooa
cachouousm
caudaaua0uc>m
oovucouumd
omvHHaoauneaq
omuuHaaae
ououHNEOHUouum
ouoflaaagoz
oavwwheonoxnm
oaano>ua0uo>=
omcHHHaEam
ocoHuoEOunoHom
oovHuOUOuqu
onvaauucoam
uuvwwuoucam
confinmeoo
occuwconmannoHo
unawaaaaunfia
omuaxuuou
moomdnae<
oachaflaaz
oouchunm<
oscucomOHOuMuou
uuvwcmnae
ououdmam
omcfioaaau
wuvuuoucOuoz
omuuvomdncwh
omoaawnm0u0>z
omUHOm~u>o
wmcdcamw
ooUHHMuxm
mandaam
omcdcoHuwmcoou
wnvnoauanafi

 

nHouoz UHcmmuo

oocmuOHOu :oHuDHHom

anuo owmdcoxma

 

N0

 
 
 

uouox cum: nHuEHqc u noses!

0H.0 50.0 10.0 HN.H 0O.H 00.0 0N.O
H5.0 ON.N 05.0 00.0 00.0 iH.O
wm 5 0 0 00 v Na 0

uanflm mcfiumdm aoflom uncouofia can acowuaum

00.0!
00.0!
Hm

00.0:
00.0I
N

50.HI
00.0!

nuouoaodax

anHZ

uoflfisz coHuaum

uuflEo>oz m
I Honouoo H0
00mH
¢ munmmm

61

Tubificidae, and the number of individuals may exceed 100,000
per square meter. As the effluent is diluted out farther
downstream families return in relation to their ability to
withstand the plating wastes.

The Michigan Water Resources Commission from past ex-
perience has labeled some families as intolerant to plating
wastes (Personal Communication with Ron Willson, 1968) and
these families are indicated in the tables. In nearly all
instances these are the families which are absent for the
greatest distance below the plating plant. In some cases
when intolerant families are represented by only one or two
individuals just below the plant, their presence is probably
due to drift and not to the fact that they normally live in
that area.

The five tables represent five different sample
series taken at different times of the year so the numbers
of families present vary but the pattern is the same in
every one whether samples were taken in spring with few
families present or in late summer when many more families

were found.

Correlation analysis - If the drastic changes between
stations upstream and downstream from the plating plant are
really due to plating wastes, there should be a statistical
correlation between wastes and numbers of families and indi-

viduals in the river. Analysis for such correlations was run

62

on the CDC 3600 computer using the 11 most prominent families
above the plant and the six metals in the effluent. Data for
the analysis was taken from the three upstream stations and
only three stations downstream to avoid interference by ef-
fluent from the Fowlerville sewage lagoons. _As shown in
Table 16, there was a positive correlation between Tubifi-
cidae and all plating wastes except hexavalent chromium and
there were negative correlations between wastes and all other
invertebrates and between Tubificidae and other inverte-
brates. All animals except Tubificidae are inhibited by the
plating wastes and it almost appears that the Tubificidae re-
ceive a beneficial effect. This is probably due to lack of
predation and competition since they are the only survivors
in the area and not due to any benefit from the wastes.

A similar correlation analysis was run using all
families present in that section of the river. _Results,
which are on file and available at Institute of water Re-
search, Michigan State University, are essentially the same
and so were not included here due to unwieldy size of the
intercorrelation matrix. It would be possible to use the
correlations to set up lists of relatively intolerant fami-
lies including Baetidae, Tabanidae, Amphipoda, Glassiphon-
iidae and a short list of more tolerant families composed
primarily of Tubificidae so we could describe "clean" or
"polluted" water communities. But, although used extensively

in the past, there seems little to be gained from this. As

63

 

 

oooo.~ 000m.o 0nm~.0u 0H5n.0 00H0.0n 0000.0 00N~.0u 0HNH.01 H50H.0| 0000.0: HOHH.0- H00~.0u H00H.0I 000H.0| H0nn.ou ~50~.o| 0~00.0 5H
20
oooo.H 0n0~.o 5505.0 0005.0 0000.0 5H~n.on 0000.0: 0050.0: 5000.0: 0000.0: 0H0~.0| 500H.0n H0dn.0u 0HON.0u 0000.0: 0Hn0.0 0H
anuok
no
oooo.H 0000.0 0000.0 0000.0 0050.0: 0000.0: 0H00.0| 0000.0: 0000.0: 0000.0- 0000.0: HH50.0| 0~50.0u 00~H.0n 0000.0: 0H
ouU
oooo.H 0000.0 0000.0 0000.0: 05nH.ou 505H.00 H000.0s 5H0H.0I «000.0- 00~H.0| 0000.0- 0000.0: 0000.0: 5050.0 0H
cu
0000.H 0005.0 0000.0: 0nuH.0u 000H.0| 0500.01 050H.0n 0000.0- 00HH.0I 0~0H.0a ~5H~.0| 05Hn.0- 000H.0 ma
“2
000.H 0500.0: 0«5H.0n 00Hn.0u «000.0: 000H.0- 0H0~.0u HO0H.0| 0500.0: 0500.0: 0000.0: 0000.0 NH
:0
0000.H 0000.0: 0000.0 0050.0 0000.0 H00~.0 HH00.00 0000.0 0000.0 5HH0.0 ~0H~.ou HH
coca
nuuomzmm
0000.H 0505.0 00H0.0| 0000.0 000H.0 0000.0 0000.0 Hnno.0 000H.0 anH.ou 0H
onowwconm
uuumoHo
0000.H Hnwo.0 NOH0.0 0NNH.0 5050.0 0500.0: 0000.0 nnHH.0 0n0~.0u 000
up
-0cae<
0000.H 00H0.0 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0: 0
. uupfisHm
0000.H ~5~H.0 0000.0 00H0.0u H0~0.ou 0HH~.0 5050.0- 5
ooUHcoHum
uncoou
0000.H H000.0 0000.0 0000.0 nHOH.o 00¢H.ou 0
man
oHcman
0000.H 00H0.0| 0000.0: 0H50.o 050H.o: 0
wan
nauwam
0000.H 0000.0 0000.0 000H.on 0
006
uHqum
0000.H 0050.0 00H~.oa m
omchomom
-oumumo
ooo.H 000~.ou N
0000000
Iflncme
oooo.H H
omufluHu
naps?
5H 0H 0H 0H 0H 0H HH 0H 0 0 5 0 0 0 n w H
20 Hmuoe 0uu :0 «z :0 0000 oaoHHconn upon ooUHiHu oaowcoHu0 000 vac can 0000:0000 omofiooa «meanwu
Ho -33ng $.qu his... -338 .332. -323 .33m .338 -659 -32.

 

 

 

.nHuuoE 0 00¢ nauHHeau HH how xHuqu coHuuHouuouuuucH

.cH UHQMH

64

Hynes (1963) has pointed out, these ideal communities seldom
occur as predicted due to the variability in nature so they
can be misleading. And, in this case, simple family lists as

shown in Tables 11-15 accomplish the same purpose.

Principal components analysis - Although the corre-
lations between invertebrates and wastes are quite strong
(Table 16), the effect upon different families is so varied
that we might suspect that environmental factors other than
plating wastes are also involved. In fact, variability in
nature is already so high that we would never expect the
plating effluent to be anything but one of many factors af-
fecting the stream. For this reason data on the same 11
families and six metals for three stations above and three
below the plating plant were used for a principal components
analyses.

The principal components analysis is a statistical
method initially used in psychology to separate non-
correlated factors having an effect on behavior (Seal, 1964).
In this case the method was used to separate non-correlated
factors each of which had an effect upon the total varia-
bility of invertebrates in the three upstream and three down-
stream sample stations. The columns labeled P1 .... Pl7
(Table 17) are vectors of factors which ;ffect each family

or metal by the relative amounts shown in the columns. At

the bottom of each column is given the relative amount of

65

 

 

 

00.00 00.05 00.00 05.00 05.00 0.00 accouom
o>~00~«.LL10
0.0 0.0 0.0 0.00 0.00 0.00 acvuuum
c202.0 0000.0 0500.0 0000.0 0000.0 0000.0 0050.0 0000.0 0000.0 0000.0 0005.0 0000.0 0000.0 0000.0 5005.0 0000.0 0000.0 wucmaun>
00 0.0.1.950
o>uumqmm
2000.0 0000.0: 0000.0 0000.0: 0050.0: 5500.0 0000.0: 5000.0 0000.0 0000.0 0050.0 0000.0 0000.0: 0000.0 0000.0: 0000.0 0500.0 20
0000.0: 0000.0 0500.0: 0000.0 0000.0 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: 0000.0: 0000.0 0500.0 0000.0 0050.0 0000.0 «much
00
0H00.c: 0000.0 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: 5000.0 0550.0 0050.0 0000.0: 0000.0: 0000.0 0050.0 0000.0 0uu
0000.0 0000.0: 0000.0 0000.0: 0000.0: 0000.0 0000.0 0050.0: 5000.0: 0500.0: 0500.0 0000.0: 0000.0 0000.0 0000.0 0000.0 0005.0 :0
anoc.0 0000.0: 0000.0 0000.0: 0000.0: 0500.0: 0000.0: 0000.0 0000.0 0000.0: 0500.0 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0 02
Hmfic.c: 0000.0 0050.0: 0000.0: 0050.0: 0000.0: 5000.0: 0000.0: 0000.0: 0000.0: 0000.0: 0000.0: 0000.0 0050.0 0000.0 0000.0 0000.0 30
0000.0 0000.0 0000.0: 0000.0 5000.0: 000000: 0000.0 0000.0 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0 0500.0 0000.0: 0000.0: mama
:Huwm: 0
0000.0 5500.0 0000.0 0000.0: 0000.0 0000.0: 0000.0 5000.0 0000.0 0000.0 0000.0: 0000.0: 0000.0 0000.0 0000.0 0055.0 0000.0: unvaacon0
. :wmmOHU
0500.0: 0000.0 0000.0 0050.0 0000.0: 0000.0 5500.0: 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0000.0 0055.0 0000.0: mm00
:0; 50
0000.0: 0000.0 0500.0 0050.0: 0050.0 0000.0 0000.0: 0000.0: 0550.0 0000.0 0050.0: 0000.0 0000.0: 5000.0 0500.0 5000.0: 0000.0: 0000600
0000.0 5000.0: 5000.0: 0000.0 0000.0: 0000.0 0000.0: 5000.0: 0000.0: 0000.0 5000.0 0000.0: 0000.0 0000.0 0000.0 5000.0 0000.0: omvacoHum
:mcmou
0000.0: 0000.0: 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0: 0000.0: 0050.0: 0050.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: mmo
:acmnme
0000.0: 0000.0: 0000.0: 0000.0: 5500.0 0000.0: 0000.0 0000.0 0500.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0: umvflummm
0000.0 0000.0: 50Ho.o: 0000.0 0000.0: 0000.0: 0050.0 0000.0: 0000.0 0000.0 0000.0: 0500.0 5005.0 5000.0 0050.0 0500.0: 0000.0: vandaMAm
0000.0 0000.0: 0000.0: 0000.0 5000.0: 0000.0: 0000.0 5000.0 0000.0 0000.0 0000.0: 0500.0: 0000.0: 0000.0 0000.0 0000.0: 0000.0: 000000000
:Oumumu
0000.0: 0500.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0050.0: 5000.0 0000.0 0050.0: 0500.0 0000.0 5000.0 0000.0: 0000.0: 0000000
:flccoe
0000.0: 0000.0: 0000.0 0000.0 0000.0 0000.0: 0000.0 5000.0: 0000.0: 0000.0 0000.0 0000.0: 5000.0 0000.0 0000.0: 0000.0 0000.0 0000000
:0n19

500 000 000 000 000 000 000 000 00 00 50 00 00 00 00 00 00
.00000000 nucoconeou 000000000 000 xwuume 0:00000 uOuUNL .50 00908

65

 

 

 

00.00 00.05 00.00 05.00 05.00 0.00 ucoouvm
o>0000vLLau
0.0 0.0 0.0 0.00 0.00 0.00 accouvm
2:00.0 0000.0 0300.0 0000.0 0000.0 0000.0 0050.0 0000.0 0000.0 0000.0 0005.0 0000.0 0000.0 0000.0 5005.0 0000.0 0000.0 muco0um>
00 quOE¢
0>0um0om
0000.0 0000.0: 0000.0 0000.0: 0050.0: 5500.0 0000.0: 5000.0 0000.0 0000.0 0050.0 0000.0 0000.0: 0000.0 0000.0: 0000.0 0500.0 20
000r.0: 0000.0 0500.0: 0000.0 0000.0 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: 0000.0: 0000.0 0500.0 0000.0 0050.0 0000.0 0muoe
no
0000.0: 0000.0 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: 5000.0 0550.0 0050.0 0000.0: 0000.0: 0000.0 0050.0 0000.0 000
0000.0 0000.0: 0000.0 0000.0: 0000.0: 0000.0 0000.0 0050.0: 5000.0: 0500.0: 0500.0 0000.0: 0000.0 0000.0 0000.0 0000.0 0005.0 :0
0r00.0 0000.0: 0000.0 0000.0: 0000.0: 0500.0: 0000.0: 0000.0 0000.0 0000.0: 0500.0 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0 02
000a.0: 0000.0 0050.0: 0000.0: 0050.0: 0000.0: 5000.0: 0000.0: 0000.0: 0000.0: 0000.0: 0000.0: 0000.0 0050.0 0000.0 0000.0 0000.0 30
0000.0 0000.0 0000.0: 0000.0 5000.0: 000000: 0000.0 0000.0 0000.0 0000.0: 0005.0 0000.0 0000.0 0000.0 0500.0 0000.0: 0000.0: me0
:0ummc 0
0000.0 5500.0 0000.0 0000.0: 0000.0 0000.0: 0000.0 5000.0 0000.0 0000.0 0000.0: 0000.0: 0000.0 0000.0 0000.0 0055.0 0000.0: 00000coca
. :0nn00u
0500.0: 0000.0 0000.0 0050.0 0000.0: 0000.0 5500.0: 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0000.0 0055.0 0000.0: amen
:0: E<
0000.0: 0000.0 0500.0 0050.0: 0050.0 0000.0 0000.0: 0000.0: 0550.0 0000.0 0050.0: 0000.0 0000.0: 5000.0 0500.0 5000.0: 0000.0: 0000E0m
0000.0 5000.0: 5000.0: 0000.0 0000.0: 0000.0 0000.0: 5000.0: 0000.0: 0000.0 5000.0 0000.0: 0000.0 0000.0 0000.0 5000.0 0000.0: wm00c00u0
:mcmou
0000.0: 0000.0: 0000.0: 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0: 0000.0: 0050.0: 0050.0 0000.0: 0000.0 0000.0 0000.0: 0000.0: mmv
:0cmnmh
0000.0: 0000.0: 0000.0: 0000.0: 5500.0 0000.0: 0000.0 0000.0 0500.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0000.0 0000.0 0000.0: oa00ummm
0000.0 0000.0: 5000.0: 0000.0 0000.0: 0000.0: 0050.0 0000.0: 0000.0 0000.0 0000.0: 0500.0 5005.0 5000.0 0050.0 0500.0: 0000.0: 00000m0m
0000.0 0000.0: 0000.0: 0000.0 5000.0: 0000.0: 0000.0 5000.0 0000.0 0000.0 0000.0: 0500.0: 0000.0: 0000.0 0000.0 0000.0: 0000.0: 000000000
:Oumumu
0000.0: 0500.0: 0000.0 0000.0: 0000.0 0000.0 0000.0: 0000.0 0050.0: 5000.0 0000.0 0050.0: 0500.0 0000.0 5000.0 0000.0: 0000.0: 00000mm
:0ccwe
0000.0: 0000.0: 0000.0 0000.0 0000.0 0000.0: 0000.0 5000.0: 0000.0: 0000.0 0000.0 0000.0: 5000.0 0000.0 0000.0: 0000.0 0000.0 mmv0o00
:0nab

50m 00m 0000 000 00.0 000 00a0 000 00 0.00 5.“0 0A0 00 00 0A0 0A0 00
.a0az0ocu nucocoaaoo 000000000 new X0uuae 0:00m00 Houomm .50 w0nme

66

variance attributed to each factor and from this is figured
the percent of variance attributed to each factor.

An important characteristic of principal component
analysis is that it separates out the factors, shows the ef-
fect of each factor on each metal and each family, and indi-
cates how much of total variance is attributed to each
factor, but it does not identify the factors. Identification
of the factors is based upon all available evidence and is
dependent to a large degree upon the amount of evidence and
the experience of the researcher. So, a weakness of the
method is that it never conclusively proves the effect of any
one identifiable factor in the environment. The main ad-
vantage is that, with non—correlated factors, it enables one
to at least recognize that these factors are present and to
hypothesize as to their nature.

In this case the primary aim is to see how much of
the total variance can be attributed to the plating effluent.
Factor Pl which accounts for 30% of variance is almost cer-
tainly due to the effluent. Metals and tubificids all in-
crease and all other forms decrease in relation to this
factor. Other methods of analysis, such as the family lists,
have already shown that this is the effect of the plating
effluent on the stream. When we consider the great natural
variability in a stream, 30 percent is a large amount to be
caused by one factor, especially when we know that the ef-

fect of this factor is detrimental.

67

Other major factors such as P2 and P3 may be related
to season, temperature, stream discharge or any number of en—

vironmental forces not even suspected. Factor P seems es-

3
pecially likely to be season or discharge (or this may be
one factor) since most invertebrates increase as summer pro-
gresses and metal concentrations increase with decrease in
stream discharge. It is probably impossible to identify all
forces in the environment but it is still important to be
able to recognize that these forces are present and to see
that, while the plating plant effluent is not the only

factor contributing to overall variance in the stream, it is

the major factor in this area.

Index of pollution - Family occurrence lists such as
shown in Tables 11-15 and lists of indicator species have
been used extensively in the past as measures of pollution
and a combination of these indicators has been accepted by
most biologists as quite reliable. However, these methods
require a great deal of time and effort and much of the work
must be done by persons with good training in taxonomy. Even
when the time and effort are spent, the results are often
meaningless to anyone but an aquatic biologist. Some biolo—
gists such as Hynes (1963) and probably most engineers claim
that we need simpler methods of quantifying pollutional ef-
fects to make the results more useable by industry and under-

standable by the lay public.

68

The fact that Tubificidae can thrive in concen—
trations of plating wastes which eliminate all other forms
of invertebrates suggests that some relationship between
tubificids and other forms might be an indicator of pollution,
at least from the plating plant. King and Ball (1964) recog-
nized this relationship and proposed a ratio of aquatic in-
sect weight to tubificid worm weight as an index of domestic
and plating plant pollution in the Red Cedar River.

This ratio worked quite well for King but it had one
major disadvantage. If worms were absent the index reached
infinity and this made it impossible to calculate means,
variance, etc. Substitution of King‘s index into a simple
mathematical formula solves this problem while retaining the
basic idea of the index.

The basic idea of the formula is that if a number t

goes from 0 to infinity,

 

 

. 1
then x — l + t goes from O to l.
. . Insects . . .
Since the index TubifiCids goes from O to infinity,

it is substituted for t.

 

So the index x =.__;L____ = TubifiCids
1 _lflégggg TubifiCids + Insects
TubifiCids

and goes from O to 1.

If insects are absent the index will be 1 and if tubificids

are absent the index will be 0. An index of zero would seldom

69

occur since we would almost always have at least a few
tubificids in a well-balanced clean-water system.

The revised formula was used to calculate an index
value for all sections of the river based on all samples in
each section. The result (Figure 5) shows much the same
pattern as a simple list of families but the curve is more
even. Mean index values at the lower end of the study area
had nearly returned to the levels found upstream from the
plating plant, after having almost reached maximum values of
1.0 just below the plating effluent. This is much the same
pattern found in the chemical analyses for metal concen-
trations. Concentrations 10 miles downstream were almost

down to the levels found above the plating plant.

Multiple regression analysis -’If the index is
really a measure of the "health" of the river, and if this
"health" is influenced primarily by the plating plant ef-
fluent, there should be a measurable relationship between in-
dex and metals so we could predict the effect of one upon the
other. To determine this relationship, metals and index
values for all stations in all four regular sample series
below the plating plant effluent were used to compute a
multiple regression of index on the plating plant metals.

The resultant regression equation is: Index =
0.15 + 0.47 Cu + 0.05 Ni + 0.42 Zn + 9.10 Cr6 + 1.46 CrT +
10.03 CNr- The equation is significant at less than 0.0005.

A comparison of regression and error sums of squares

70

Figure 5. Relation of mean insect-tubificid index
to distance downstream from the plating
plant. Calculated using both regular

and pilot series data.

71

1.01!

.74.

.69

P
1

...c m. 3 2.

sateen so; O:_u> love. can!

.11-

10

Distance Downstream In Miles

72

indicates that 57% of the index variability in the river be-
low the plating plant can be explained by regression of the
index on the plating plant wastes (Li, 1964). Regression

coefficients and standard errors of coefficients are as

 

 

follows:
Variable Regression Coefficient Standard Error
Constant 0.1571 0.0842
Cu 0.4763 0.5239
Ni 0.5266 0.1029
Zn 0.4290 0.4231
Cr6 9.1750 16.5505
CrT 1.4679 0.4919
CN 10.0363 3.1528

From the regression coefficients it would appear
that total chromium and cyanide have the most effect upon
the index. Hexavalent chromium also shows a very great ef-
fect but the standard error is very high, due to the fact
that Cr6 was absent from all but one of the river samples.
It is not surprising that cyanide has a pronounced effect
since the toxicity of cyanide is well-known. But, it is
interesting to note that chromium is so important since it
was thought to be relatively harmless in the trivalent state
and was not even limited by the 1953 Order of Determination
of the‘Water Resources Commission. Complete data and

analysis are available on file at the Institute of Water Re-

search, Michigan State University.

73

If the regression of index on metals is really worth-
while we should be able to predict the index from metal
concentrations. This was done and a plot of observed and
predicted indices is shown in Figure 6. The multiple corre—
lation coefficient is 0.7593 which is quite good for a highly
variable biological system. At the same time, we should be
able to determine a desirable index level and use the re-
gression equation to set waste limits to ensure the main-
tainance of such an index level.

As a measure, the index may not be meaningful by it-
self, but it can be tied in with other things to help de-
termine a desirable index. Fish production for recreation
is a primary concern in this area and fishing has long been
deteriorating on the Red Cedar River. Linton and Ball (1965)
have indicated that fish populations in the river are drasti—
cally reduced for about eight or nine miles below the plating
plant. If this is the case it appears that we cannot have an
acceptable recreation fishery in the river in an area where
the insect-tubificid index is higher than about 0.4.

If 0.4 is taken as the maximum allowable index we
see that plating wastes are at the maximum allowable limits
about nine miles below the plant. This point in the stream
corresponds to section 11 and mean metal concentrations in
mg/l found there are as follows: Cu 0.05, Ni 0.39, Zn 0.03,
Cr6 -0.0, CrT 0.12, CN’ -0.0. Use of these mean concen-
trations in the regression formula results in a predicted

index of 0.39.

Figure 6.

74

Relation of observed to predicted insect-
tubificid index values. .Indices are based
upon data from the four regular sample
series in which both invertebrates and

water samples were collected.

75

Observed -———
Predicted --- - --

 

 

 

J

D i 2 3 4 5 6 7 5 9 10

Distance In Miles to Center Of Section

76

If these same values are used in the formula pro-
posed by Tarzwell (1967) for maximum allowable wastes the

following results are obtained:

Cu Ni Zn CrT
0.056 + 0.39 + 0.03 + 0.12 = 3.07
0.16 1.00 1.06 0.05

From this, we would asSume that wastes are still too high be-
cause the results of Tarzwell's formula are not supposed to
exceed 1.0. It appears that his formula may be a bit more
conservative than necessary but he may not be far off when

we consider long-term effects on the population. Linton
(1967) claims rock bass populations are declining in the
river so, if this is due to plating waste, an index value
closer to 0.3 may really be needed. If this is true we might
have to consider concentrations in zone 12 as too high so
maximum allowable amounts would be closer to the amounts pro-
posed by Tarzwell's formula and a great deal lower than were
proposed by the Water Resources Commission. To maintain an
index below the 0.3 range with the metals present in the

same proportions found in section 11, maximums would have to
be about half the average amounts found in section 11. These
would be 0.026 Cu, 0.20 Ni, 0.015 Zn, 0.06 CrT expressed as
mg/l. The resultant index would be 0.27. If these are used
:in Tarzwell's formula the result is 1.52 so still supposedly

too high.

77

The fact that Tarzwell's formula is high is due pri-
marily to the very low limit (0.05 mg/l) which is used for
total chromium as suggested by Fetterolf (1968). This limit
may be lower than necessary but we cannot be sure at this
time.

Based upon invertebrate populations in the Red Cedar
River and fish populations in the corresponding areas, I be-
lieve maintainance of an insect—tubificid index not much
greater than 0.3 would ensure use of the river for warm-water
sports fisheries. To maintain this index and general state
of "health" of the river the following are proposed as maxi-
mum concentrations of plating wastes in the river: 0.056

mg/l Cu, 0.39 mg/l Ni, 0.03 mg/l Zn, 0.00 mg/l 0:6

T

, 0.12
mg/l Cr and 0.00 mg/l CNr. This should assure a maximum
index value of 0.39 below the effluent and an index of 0.3 or
better could be expected within less than a half-mile dis—
tance downstream. If these limits are maintained in the up-
stream sections near the plant outfall, we could expect to

be able to maintain sports fish populations there (excluding

any other pollution or habitat problems) and could expect

the river further downstream to be even better.

SUMMARY

Metal plating wastes discharged by Utilex Manu-
facturing Company into the Red Cedar River at Fowlerville,
Michigan are causing a decrease in productivity of fish and
bottom organisms in the river for many miles below the ef-
fluent. In the present study, water was analyzed for plating
wastes above and for 10.3 miles below the plating waste out—
fall and standing crop estimates were made of invertebrates
in the same areas in order to determine the effect of the
plating wastes on numbers and types of invertebrates in the
stream.

Nickel, zinc and total chromium were found in sig-
nificantly (by t-test at 0.05 level) higher concentrations
for 10.3 miles downstream as compared to concentrations up-
stream from the plating plant. Copper was increased sig-
nificantly for over eight miles. Cyanide and hexavalent
chromium could not be traced nearly so far but they were
present in the river for as much as one-half mile downstream
from the effluent. All six waste elements were found in the
river in quantities which have been found to be detrimental
to aquatic organisms.

Families of invertebrates in the riVer were reduced

from an average of about 14 families upstream from the

78

79

plating plant to two right below the effluent. Families of
invertebrates begin to appear again downstream from the ef-
fluent according to their tolerance for plating wastes and
an average of 12-14 families is attained again in the river
about seven miles below the waste effluent.

Use of correlation analysis has shown that all fami-
lies of invertebrates except Tubificidae are adversely af-
fected by the plating wastes. Tubificids seem relatively
insensitive and increase in numbers in higher concentrations
of wastes, presumably because of lack of competition and
predation. Use of principal components analysis has shown
that over 30 percent of variability among invertebrates at
three sample stations upstream and three stations downstream
from the plating plant can be attributed to the plating
wastes.

A revised version of an insect-tubificid ratio (King
and Ball, 1964) is proposed as an easier method of assessing
pollution in a river such as the Red Cedar. The weight ratio

Tubificids
Tubificids + Insects

averaged about 0.25 upstream from the
plant. Just below the effluent the ratio exceeded 0.99 and
it slowly declined downstream to a value of about 0.3 ten
miles downstream from the effluent.

Multiple regression analysis was used to determine

the relationship between insect-tubificid index and plating

waste concentrations in the river. The following regression

80

equation resulted: Index = 0.15 + 0.47 Cu + 0.05 Ni +

0.42 Zn + 9.10 Cr6 + 1.46 Cr Total + 10.03 CN'. Regression
coefficients indicate that total chromium and cyanide have
most effect upon the index in the Red Cedar River. Cr6
seems very important also but is subject to a very large
standard error.

Since Linton (1967) found fish production greatly
reduced in areas which had an index value much greater than
0.3, it is suggested that this index be used as an indi-
cation of general "health" of the river and that an index
value of 0.3 to 0.4 be considered the maximum allowable in
the river. With 0.3 to 0.4 as the maximum allowable index
value the regression equation was used to arrive at a recom-
mendation of maximum amounts of wastes to be allowed in the
Red Cedar River. Maximum wastes recommended as allowable
are: 0.056 mg/l Cu, 0.39 mg/l Ni, 0.03 mg/l Zn, 0.00 mg/l
Cr6, 0.12 mg/l Cr Total, 0.00 mg/l CNF. If these limits are
not exceeded, an index below 0.3 to 0.4 can be expected to
be maintained below the plating plant and downstream stretches

of the river should improve in numbers and types of inverte—

brates and presumably in fish production.

LITERATURE CITED

American Public Health Association. 1965. Standard methods
for the examination of water and wastewater in—
cluding sediments and sludges. 12th edition. Boyd
Printing Co., Albany, N.Y., 1965. 769 p.

Anon., "Aquatic Life Water Quality Criteria, Third Progress
Report." Aquatic Life Advisory Committee of the
Ohio River Valley Water Sanitation Commission
(ORSANCO). Jour. of the Water Pollution Control
Fed., 32:65, 1960.

Anon., 1960. Report of the Water Pollution Research Board,
with the Report of the Director of the Water Pol-
lution Research Lab. for the Year 1959. Dept. of
Scientific and Ind. Res., H. M. Stationery Office,
London, as quoted in McKee and wolf. 1963. Water
quality criteria. 2nd edition. The Resources
Agency of California State Water Quality Control
Board. Publication #3—A.

Cairns, J. Jr. and Scheier, A. 1958. The effects of
temperature and hardness of water upon the toxicity
of zinc to the pond snail, Physa heterostropha (Say).
Netulae Naturae of the Academy of Natural Sciences
of Phila.: 308.

Caltrider, R. 1968. Michigan Water Resources Commission.
Report of Industrial Survey, Utilex Manufacturing
Company, Fowlerville, Michigan. Nevember 28, 29 and
30, 1967.

Denniston, W.E. 1966. .Report on survey of Utilex Manuf. Co.
effluent of NOV. 9, 10, 22, 23, 24, 1965. Michigan
State Water Resources Commission, Lansing, Michigan.

Doudoroff, P. 1952. Some recent developments in the study
of toxic industrial wastes. Proc. 4th Annual Pacific
N.W. Ind. waste Conf., State College (Pullman, Wash.).

Fetterolf, Carlos. 1968. Michigan Water Resources Commis-
sion. Limits for toxic wastes in Saline River, Mich.
Memorandum to W. G. Turney, Region Engineer. State
of Michigan. February 1, 1968.

81

82

Grzenda, Alfred R. 1960. Primary production, energetics,
and nutrient utilization in a warm—water stream.
Thesis for Ph.D., Michigan State University. 99 p-

Hynes, H.B.N. 1963. The biology of polluted waters. Liver-
pool University Press. Liverpool, England. 202 p.

Jones, J.R.E. 1938. The relative toxicity of the salts of
lead, zinc, and copper to the stickleback. Jour.
Exp. Biol. 15:394.

Jones, J.R.E. 1941. A study of the relative toxicity of
anions, with Polygelis nigra (Planarian) as test
animal. Jour. Exp. Biol. 18:170.

 

King, Darrell L. 1964. An ecological and pollution-related
study of a warm—water stream. Thesis for Degree.of
Ph.D., Michigan State University, 154 p.

King, Darrell L. and Robert C. Ball. 1964. A quantitative
' biological measure of stream pollution. Jour. water
Pollution Control Federation. May 1964.

Klein, L. 1957. AspeCts of river pollution. Butterworth
Scientific Publications, London and Academic Press,
Inc., New York.

Li, Jerome C. R. 1964. Statistical Inference, Volume II,
The multiple regression and its ramifications.
Edwards Bros. Inc., Ann Arbor, Michigan. 575 p.

Linton, Kenneth J. and Robert C. Ball. 1965. A study of
the fish populations in a warm-water stream. Quart.
Bull. MSU Agric. Exp. Sta., 48:255—285.

McKee, Jack E. and Harold W. Wolf. 1963. Water quality cri-
teria. 2nd edition. The Resources Agency of Cali-
fornia State Water Quality Control Board. Publi-
cation #3-A-

Metal Finishing Guidebook Directory. 1960. 28th edition.
Metals and Plastics Publications, Inc. westwood,
New Jersey. 754 p.

Parker, Robert W. 1961. Staff investigation of a fish
mortality in the Red Cedar River, Fowlerville, Mich.,
May 15, 1961. Michigan Water Resources Commission
Report.

83

Richards, Edward S. 1946. Chromium plating, a textbook for
the practical electro-plater, the production engineer
and the student. Charles Griffin and Co., Ltd.,

London. 131 p.

Robinson, John R. 1964. Michigan Water Resources Commis-
sion. Wastewater Survey Report, Utilex Corporation,
Fowlerville, Michigan, June 10, 11, and 12, 1964.

Seal, Hilary L. 1964. Multivariate statistical analysis
for biologists. Spottiswoode, Ballantyne and Co.,
Ltd. London and Colchester, England. 207 p.

Tarzwell, C.M. 1967. Interim report on water quality for
fish, other aquatic life, and wildlife. National
Technical Advisory Committee to the Federal Water
Pollution Control Administration on Water Quality
Criteria. Federal Water Pollution Control Ad-
ministration, U.S. Dept. of Interior, Wash. D.C.

Vannote, Robin L. 1961. Chemical and hydrological investi-
gations of the Red Cedar River watershed. Thesis
for degree of M.S., Michigan State University.

126 p.

"7111111117111111is