ECOLOGY OF THE BURROWING AMPHIPOD
PONTOPOREIA AFFINIS IN LAKE MCIHIGAN

by
Wayne Paul Alley

Research supported by
Federal Water Pollution Control Administration Grant WP-00311

and
National Science Foundation Grant GA-1337

This report was also a dissertation in partial

fulfillment of the requirements for the degree

of Doctor of Philosophy in the Department of

Wildlife and Fisheries of the University of

Michigan, 1968

Special Report No. 36

of the

Great Lakes Research Division

The University of Michigan

Ann Arbor, Michigan

1968

TABLE OF CONTENTS

Page

LIST OF TABLES vi-vii

LIST OF FIGURES viii-xii

ABSTRACT xiii-xvi

INTRODUCTION . . ' . ■ 1

Research Plan . . 3

METHODS AND MATERIALS 9

Field Sampling Procedures 9

Long-Term Study Area 9

Short-Term Study Area 16

Laboratory Procedures 18

Long-Term Study Area 18

Short-Term Study Area 20

RESULTS 22

Life History and Reproduction , . . 22

Patterns of Spatial Distribution . 32

Short-Term Study Area 32

Long-Term Study Area ..-...- 51

Interspecific and Intraspecific Associations ........ 54

Environmental Relationships . 54

Environmental Parameters Treated Individually . . . 66

Depth 66

Time ■ , 66

Bottom Temperature 66

Sediment Type 70

Percent Carbon in the Sediment 70

Distance from Shore 77

Stepwise Multiple Regression Analysis of

Environmental Parameters 77

Seasonal Distribution 86

Patterns of Spatial Abundance . . . , 96

DISCUSSION 101

Reproduction 101

iii.

Page

DISCUSSION (continued) . . 101

Patterns of Spatial Distribution 102

Short-Term Study Area 102

Long-Term Study Area 106

Interspecific and Intraspecific Associations 108

Environmental Relationships 110

Environmental Parameters Treated Individually .... 110

Depth 110

Bottom Temperature ,.. 110

Substrate Relationship Ill

Stepwise Multiple Regression Analysis of the

Environmental Parameters 112

Seasonal Distribution 117

Patterns of Spatial Abundance 120

CONCLUSIONS . , . 123

LITERATURE CITED 127

LIST OF TABLES

Table Page

1 Location, depth, and sediment type of the

Lake Michigan benthos stations^ . . <> . „ • . . . • . . 7

2 Monthly occurrence of brooding (B) and spent (S)

female Pontoporeia affinis in Lake Michigan for 1965 . . 25

3 Monthly occurrence of brooding (B) and spent (S)

female Pontoporeia affinis in Lake Michigan for 1966 . .

4 The relationships that exist between Pontoporeia

affinis and the physical and biological environments

of Lake Michigan as determined by correlation

analysis •••.••••...•. ..««.. ........ 81

5 The proportion of the multiple correlation coefficient

and the coefficient of determination, expressed as a
percentage, that each statistically significant variable
contributes to the multiple regression equation of the
four depth zones in Lake Michigan . . . . . 83

6 The proportion of the multiple correlation coefficient

and the coefficient of deter mination, expressed as a
percentage, that each statistically significant variable
contributes to the multiple regression equation of all
depths combined in Lake Michigan.,. . . . . . . . . • . 84

7 The relationships that exist between Pontoporeia

affinis and the physical and biological environment
as determined by correlation analysis for the stations

of the C, D, and E transects . • . o « . . . • . • „ . . 87

8 The grand geometric mean number of Pontoporeia

affinis for the three sampling seasons . . . . . . . . . . 91

9 Spearman rank correlation coefficients indicating

the degree of association that occurred among the

patterns of seasonal abundance of Pontoporeia

affinis for the four depth zones and all depths

combined ......................... 94

VI

Table Page

10 The grand geometric mean number of Pontoporeia

af finis for the three sampling seasons of the seven

stations of the C transect .•••.....•. 95

11 Spearman rank correlation coefficients indicating

the degree of association that occurred among the

patterns of seasonal abundance of Pontoporeia

affinis for stations C-2 thru C-7 for the three

sampling seasons 95

12 Spearman rank correlation coefficients indicating

the degree of association that occurred between

the patterns of seasonal abundance of Pontoporeia

affinis for the stations of the C transect and their

respective depth zones, 96

Vll

LIST OF FIGURES

Figure Page

1 Index map of the Lake Michigan macrobenthos

stations of the long-term study area that were

sampled from 1964 to 1966 .............. 5

2 Location of the short-term study area . . » 8

3 Smith- Mclntyre dredge in opened position, being

readied for sampling ... ............. . 10

4 Ponar grab sampler in open position, showing

screened top, side plates, and tripping machanism . 11

5 Ponar grab sampler in closed position, being brought

on deck . . , . • 12

6 The elutriation- screening device used in separating

the macrobenthic organisms from the sediments . . 14

7 Hand- coring device used to collect benthos samples

at the short-term study area ............. 17

8 Design of the systematic and random sampling

procedures of the short-term study area ........ 19

9 Morphological features of Pontoporeia affinis .... 23

10 Size frequency histograms of Pontoporeia affinis

for selected depths of three sampling periods

late April-early May, late May-early June, and

early October 1965 .••.....•••....••• 28-31

11 Size frequency histograms of Pontoporeia affinis

collected at station C-l for the 1965 and 1966

sampling seasons ••..............••• 33-34

12 Size frequency histograms of Pontoporeia affinis

collected at station C-2 for the 1965 and 1966

sampling seasons . ....... . 35-36

VI 11

Figure Page

13 Size frequency histogiams of Pontoporeia af finis

collected at station C- 3 for the 1965 and 1966

sampling seasons . . « . .............. . 37-38

14 Size frequency histogiams of Pontoporeia af finis

collected at station C-4 for the 1965 and 1966

sampling seasons .......... . 39"-40

15 Size frequency histograms of Pontoporeia af finis

collected at station C-5 for the 1965 and 1966

sampling seasons 41-42

16 The variance-mean relationship of juvenile

Pontoporeia affinis *^2 miri in length per

quadrat, from the short-term study area ...... 44

17 The variance-mean relationship of juvenile

Pontoporeia affinis , ^Jl mm in length per

quadrat, from the short-term study area 45

18 The variance-mean relationship of the total

number of juvenile Pontoporeia affinis of the

short-term study area .....••••••...., 46

19 The occurrence frequency histogram of

Pontoporeia affinis, ^2 mm in length per
quadrat, of the short-term study area super-
imposed on a normal distribution curve with
the same mean and variance ............. 47

20 The occurrence frequency histogram of

Pontoporeia affinis, ^^7 mm in length per quadrat,

of the short-term study area superimposed on a

Poisson series with the same mean ......... 48

21 The occurrence frequency histogram of the total

number of juvenile Pontoporeia affinis of the
short-term study area superimposed on a normal
distribution curve with the same mean and variance . 49

IX

Figure Page

22 The variance-mean relationship of the replicated

samples of Pontoporeia affinis per square meter

of the long-term study area ............... 52

23 The variance-mean relationship of the replicated

samples of Pontoporeia affinis per square meter

of the long-term study area (square root transformation). 53

24 Association of large and small juvenile Pontoporeia

affinis of the short-term study . . . . . . . 55

25 Association of juvenile Pontoporeia affinis ^ 7 mm

in length and Oligochaeta of the short-term study

area .................... 56

26 Association of juvenile Pontoporeia affinis ^2 mm

in length and Oligochaeta of the short-term study

area ........... ..... 57

27 Association of juvenile Pontoporeia affinis ~ 7 ram

in length and Sphaeriidae of the short-term study

area ....................... 58

28 Association of Pontoporeia affinis and Oligochaeta

of the long-term study area ...... . ...... . 59

29 Association of Pontoporeia affinis and Sphaeriidae

of the long-term study area .............. 60

30 Association of Pontoporeia affinis and Chironomidae

of the long-term study area ............... 61

31 Average number of Pontoporeia affinis per square

meter pooled by 10 m depth increments ........ 65

32 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus depth

in meters . 67

33 Average number of Pontoporeia affinis, per square

meter pooled by monthly intervals . 68

x

Figure Page

34 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus day of

year . , ...•.••.•••••••. 69

35 Average number of Pontoporeia affinis per square

meter pooled by one degree centigrade increments • . 71

36 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus bottom
temperature in degrees centigrade 72

37 Average number of Pontoporeia affinis per square

meter pooled by sediment type 73

38 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus

sediment type . 74

39 Average number of Pontoporeia affinis per square

meter pooled by tenths of a percent organic

carbon in the sediment 75

40 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus

percent organic carbon in the sediment 76

41 Average number of Pontoporeia affinis per square

meter pooled by intervals of one mile versus

distance from shore .................. 78

42 Average number of Pontoporeia affinis per square

meter obtained at each station visit versus

distance from shore ..... ............ . 79

43 A comparison of the Pontoporeia affinis density of

the seven station located on the C transect, sampled

9 and 10 August 1966, and the coefficient of variation

for the combined sampling dates ............ 89

44 Patterns of seasonal abundance of Pontoporeia affinis

for the four depth zones and all depths combined ... 90

XI

Figure Page

45 Patterns of seasonal abundance of Pontoporeia

affinis for the seven stations of the C transect .... 93

46 The average density of Pontoporeia affinis for

the stations of the five transects ........... 97

47 The average density of Pontoporeia affinis for

the stations of the five transects with the effect

of depth removed 100

xu

ABSTRACT

The ecology of the burrowing amphipod

Pontoporeia affinis in Lake Michigan

by
Wayne P. Alley

The Great Lakes Research Division of the University of Michigan
collected macrobenthos samples and measured environmental parameters
at 35 stations in the southern two-thirds of Lake Michigan. The stations
of this three year long-term study were sampled monthly in triplicate
from August to November 1964, from April to November 1965* and from
March to November 1966. Environmental parameters: sediment type,
percent carbon in the sediment, water temperature, distance from shore,
and depth were chosen as variables likely to affect the abundance and
distribution of the macrobenthic biota,

A second study site, called the short-term study area, was located
off Mona Lake, Michigan. Eighty-eight samples were collected there
13-14 June 1967 by two divers using a hand coring device.

Physical and biological parameters collected at these two study
areas were utilized in studying the ecology of the burrowing amphipod
Pontoporeia affinis in Lake Michigan. The patterns of reproduction,
changes in seasonal abundance, patterns of spatial distribution, and the

xiii

interspecific and intraspecific associations of this amphipod were exam-
ined.

The brood chambers of adult female amphipods, collected in the
long-term study area, indicated that reproduction is completed by late
May- early June at depths less than 35 m and beyond 35 m the amphipod
population breeds intermittently throughout our sampling seasons. The
fact that adults die after reproduction simplifies the recognition of year
classes. Size frequency histograms, collected throughout the sampling
seasons, indicated that Pontoporeia living at a depth of 10 m mature in
one year, those inhabiting a depth zone of 20-35 m require two years to
mature, and those living at depths greater than 35 m possibly require 3
years to mature.

Juvenile Pontopbreia approximately 2 mm long conformed to a
normal distribution and juvenile amphipods approximately 7 mm long
were randomly distributed at the short-term study area. The two size
groups appeared to be positively associated with each other* Both size
groups were negatively associated with the oligochaetes, the large size
group was negatively associated with the sphaeriids, and no association
occurred between amphipods and the chironomids.

Total Pontoporeia of the long-term study area were associated in
a positive manner with the oligochaetes, sphaeriids, and chironomids.
This positive association is attributable to larger benthos samplers that
masked the microassociations of the benthic organisms.

xiv

A concentration of Pontoporeia exists around the 35 m depth contour,
which usually coincides with the junction of the profundal and sublittoral
zones. The low density of amphipods in the sublittoral is due to the
environmental extremes of this area, while the profundal has a more
uniform environment and fewer amphipods, which is here attributed to
a slower growth rate and less available food.

The lake was divided into four depth zones: 10-35 m, 3 6-65 m,
66-105 m, and greater than 105 m, and the effects of the environment on
the abundance of Pontoporeia were examined by stepwise multiple regres-
sion analysis for these depth zones. The multiple correlation coefficients
for these zones were: 0. 756 at 10-35 m, 0. 599 at 36-65 m, 0. 677 at
66-105 m, and 0.826 at the greater than 105 m depth zone. When all
depths were combined the multiple correlation coefficient was 0. 796.
These high correlation coefficients indicate that the environmental vari-
ables statistically contributed a considerable amount of information about
changes in amphipod abundance.

The average standing crop of Pontoporeia was examined for the
individual depth zones and all depths combined; and, in general, there
did not appear to be any major deviations among the average standing
crops for the three sampling seasons.

The patterns of seasonal amphipod abundance were examined at six
profundal stations of one transect, and these patterns appeared to be
quite similar, particularly for adjacent pairs of stations.

xv

The average standing crop of Pontoporeia was estimated for each
station of the five transects, and the effect of depth was removed statis-
tically from these. data*. The results indicated that stations located in
regions subjected to upwelling had a larger standing crop of amphipods
than other regions in the lake. Stations located in areas where bottom
sediments were variable had, on the average, a lower standing crop of
amphipods.

xvi

INTRODUCTION

The burrowing amphipod Pontoporeia af finis Lindstrfcim (1855), the
most abundant macrobenthic organism in Lake Michigan, was originally
described from lakes in northern Germany. Subsequently, Smith (1874)
describedJP. hoyi and P. filicornis from the Great Lakes, Ekman (1913)
described jP. weltneri from Lake VfeLttern, Sweden. Juday and Birge (1927)
sent specimens of JP. hoyi from Green Lake, Wisconsin, to G. O. Sars
who concluded that this species was synonymous with JP. af finis . Adamstone
(1928) decided that J\ filicornis was also identical with JP^. affinis.
SegerstrSle (1937, 1950) in a revision of the genus, showed that males of
J?* affi n is and P. welterni were identical and that JP_. kendalii , described
by Norton (1909) from Lake Champlain, Vt. , was also synonymous with
P. affinis. Bousfield (1958) indicated that a large, ecologically distinct
form of this amphipod found in the Ungava Bay region of Quebec may
prove to be distinct from the Lindstrbm form. However, Bousfield
concluded that the Ungava Bay amphipod should be designated Pontoporeia
affinis until more information on the taxonomy, distributional ecology,
and physiology of this organism becomes available.

SegerstrSle (1937) concluded that P. filicornis is identical with
the penultimate instar of JP. affinis and suggested that it may be a neotenic
form of JP. affinis . He designated P. filicornis as JP. affinis var,
brevicornis . It is interesting to note that while P. affinis var. typica

1

has been recorded from both European and North American lakes, P.
af finis var. brevicornis is found only in North America, which suggests
that the latter form developed in this continent since the Pleistocene,
SegerstrSle (personal communication) has found brevicornis in all the
St, Lawrence Great Lakes with the exception of Lake Ontario and feels
that it will also be found there with additional sampling.

SegerstrSle (1959) presents an excellent account of the distribution
°^ — a ^i n ^ s i n Eurasia, while Henson (1954) gives an extensive review
of the North American distribution. Pontoporeia affinis is found in the
brackish waters of arctic Russia, Siberia, and Alaska, and within the
Baltic and Caspian Seas, It inhabits the oligotrophic lakes of the once
glaciated sections of northern Europe and is also widely distributed in
Siberian lakes and the estuaries and rivers of arctic Eurasia, In North
America it has been reported in the larger oligotrophic lakes of Canada,
Ungava Bay, St, Lawrence estuary,, the'Great Lakes, Green Lake of
Wisconsin, the Finger Lakes of New York, and Lake Washington on the
west coast.

Various workers have provided estimates of the numerical impor-
tance of this amphipod in Lake Michigan,, Eggleton (1936, 1937) reported
as many as 8,400 individuals Imr dX one station and showed that this
amphipod comprised 64 percent of the total number of macrobenthic
invertebrates. An examination of his data indicates an average of approx-
imately 700 amphipods/m . Merna (I960) found the average standing

crop of JP. af finis in the lake to be approximately 1, 700 individuals Im 9
and that 70 percent of the total wet volume of the macrobenthos consisted
of Pontoporeia . Henson and Chandler (unpublished data), cited by Henson

(1966), found approximately 1,500 individuals /m in the Straits of

2

Mackinac Marzolf (1965) noted more than 14, 000 amphipods/m at one

station in the Grand Traverse Bay region. Powers et ah (1967) recorded
one station in the lake where JP, af finis exceeded 23, 000 individuals /m .

Powers and Alley (1967) showed that the average standing crop of this

2
amphipod exceeded more than 2, 700 individuals /m and that it comprised

64 percent of the total macrobenthos. In a comparative study, Robertson

and Alley (1966) showed that the average abundance of jP. affinis,

oligochaetes, and sphaeriids was significantly greater in 1964 than in

1931.

RESEARCH PLAN

With the present interest concerning the rates of eutrophication
and increase in pollution, and because this amphipod is the first macro-
benthic organism to be driven out by increased pollution, it is reasonable
to study it. The objective of this investigation was to gain information
on the ecology of the amphipod Pontoporeia affinis in Lake Michigan and
to evaluate its normal patterns of reproduction, normal changes in sea-
sonal abundance, and normal patterns of spatial distribution. When
these are known, pollution- caused deviations can be recognized.

The data presented here were collected during comprehensive

multidisciplinary studies of Lake Michigan carried out by the Great
Lakes Research Division of the University of Michigan, This program
represented the first coordinated long-term study of the physical- chem-
ical environment and of the biota of the lake. It provided an abundance
of data applicable to problems concerning the annual fluctuations of the
biota and to the interrelationships of these organisms with their
environment.

These data were collected in two different study areas. The first
study area, which is called the long-term study area, consisted of 35
stations located in the southern two-thirds of Lake Michigan (Fig. 1).
Thirty-three stations were located on five cross-lake transects. These
transects were assigned letter designations* and the stations of each
transect were numbered serially from east to west. Stations on transects
A, B, and C were positioned according to major surficial bottom sedi-
ment types as described by Ayers and Hough (1964), Since detailed
information on bottom types was not available for the northern basin.,
stations on the D and E transects were positioned according to major
bathymetric features. The remaining two stations of the long-term
study area were located off Muskegon and were designated C B -1 and
C'-2„ The locations, depths, and sediment types of these 35 stations
are presented in Table 1. Sediment-type nomenclature is explained in
a subsequent section under "METHODS AND MATERIALS. " Sampling
was carried out monthly from August to November 1964, from April to

MILWAUKEE

CHICAGO

RACINE

T r

10 20 30 40

FIG. 1. Index map of the Lake Michigan macrobenthos
stations of the long-term study area that were sampled
from 1964 to 1966.

November 1965, and from March to November 1966. Stations of the
B and D transects were not sampled after June 1966.

Patterns of reproduction of Pontoporeia were studied by examin-
ation of brooding and spent females and by analysis of the size frequency
of amphipod populations from various depth zones. The environmental
parameters, sediment type, percent organic carbon in the sediment,
water temperature, distance from shore, wave height, and depth were
chosen as environmental factors likely to affect the abundance and
distribution of Pontoporeia . These variables were measured and recorded
at each benthic sampling station.

The second study area, called the short-term study area, was
located near Mona Lake approximately 6, 000 feet from shore at a depth
of 56 feet (Fig. 2). Eighty-eight benthos samples were collected at two
adjacent 15 x 15 ft. areas 13-14 June 1967. These samples were used
to determine the small-scale patterns of Pontoporeia distribution and to
examine the microassociation of Pontoporeia and the other macrobenthic
organisms.

TABLE 1. Location, depth, and sediment type of the Lake Michigan
benthos stations.

Location

Depth

Station

N lat

W long

Meters

Sediment Type

A-l

42°06'30"

86°32 , 00"

18

Sand

A-2

42°06'00"

86 o 37'00"

35

Silty sand- sandy silt

A-3

42°05'30"

86°43'00"

70

Silt -clayey silt

A-4

42°03'30"

87°06'30"

74

Layered

A-5

41 o 57'00 M

87°18'30"

43

Silty sand; sand and gravel

A-6

41 o 52'00"

87°27'00"

18

Sand and gravel; gravel

B-1

42°24'00"

86°20'30"

19

Sand

B-2

42 o 24'00"

86°27 , 00"

47

Silt- clayey silt

B-3

42 o 24'00"

86°35'30"

68

Silt -clayey silt

B-4

42°23 , 30"

87°01'00"

129

Silt -clayey silt

B-5

42°22'30"

87°21'00"

108

Silt -clayey silt

B-6

42 o 22 , 30 H

87°30'00"

83

Layered

B-7

42 o 22'00"

87°40*00"

45

Silty sand- sandy silt

B-8

42 O 22'00 n

87°47'30"

11

Sand

C-1

42°49 , 40"

86°14'50"

20

Sand

C-2

42°49 , 40"

86°18'25"

50

Silt- clayey silt

C-3

42°49'10"

86°28'25"

77

Silt -clayey silt

C-4

42 o 48"50"

86 o 41'30"

108

Layered

C-5

42°49'Q0"

86°50'00"

157

Silt -clayey silt

C-6

42 O 47'40"

87 o 26'50"

99

Layered

C-7

42°47 , 30* 1

87°34 I 30"

55

Silty sand- sandy silt

C'-l

43°08»00"

86°23 ! 00"

38

Silty sand- sandy silt

C-2

43°12'00"

86°31'00"

93

Silty sand- sandy silt

D-l

43°57'00"

86°33'00 n

30

Sand

D-2

43 o 56'00"

86°39'30"

98

Layered

D-3

43°54 , 00"

86°51'30 n

170

Silt -clayey silt

D-4

43°48'00"

87°03 , 00"

131

Layered

D-5

43°38'40"

87 o 31'00"

119

Layered

D-6

43°44'00"

87°38'00"

30

Sand

E-l

44°37'30"

86°18'12"

44

Sand

E-2

44°37 , 00"

86°21'42"

197

Silt- clayey silt

E-3

44°34'00"

86 o 40'00"

271

Silt -clayey silt

E-4

44°30'18"

86°55 l 18"

216

Silt -clayey silt; layered

E-5

44 o 25'30"

87°10'18"

173

Silt -clayey silt

E-6

44°27'48"

87 26'25 n

33

Sand

N

A

LAKE
MICHIGAN

SHORT-TERM
STUDY AREA

FEET

10 20x10°

FIG. 2. Location of the short-term, study area.

METHODS AND MATERIALS

FIELD SAMPLING PROCEDURES
Long - Term Study Area

The macrobenthos of the long-term study area was sampled in
triplicate at each station, on a monthly basis, during three sampling
seasons, Data collected during the biological sampling from August
1964 to June 1966 are presented by Ayers and Chandler (1967), Benthic
samples were taken with a Smith -Mclntyre dredge (Mclntyre 1954)
(Fig. 3) until June 1965, and with a Ponar grab sampler thereafter
(Powers and Robertson 1967) (Figs. 4 and 5).

Powers and Robertson have shown that the Ponar and Smith-
Mclntyre are comparable in their sampling efficiencies. The Ponar has
proved to be a more desirable sampler because the Smith- Mclntyre is
large and unwieldy, mechanically complicated, and the jaws are subject
to premature closing. The powerful tripping springs of the Smith-
Mclntyre render it dangerous, and at least two men are required to
operate it.

Data collection was carried out by two Great Lakes Research
Division ships, the R/V MYSIS and the R/V INLAND SEAS. Each vessel
was equipped with radar, Raytheon fathometer, hydrographic and heavy
duty winches, and the complete complement of limnological, oceano-
graphic, and meteorological equipment necessary for large scale

10

r

CO

o

m

0)

u
bo

a
•i-i

o

to
o
ft

ft
o

l-H

a

!

s

CO

a

l-H

11

el

CO
0)

O

*d
eu

el

CD
<D

fH

U
CO

bO
Pi

o

CO

el
o

CO

O

Ph

Pi
CD

Oh

o

<D

r-H
Ph

0$
CO

Is

^ ft CO

. G

ii «*

ft g

• W)

^ s

o £■

ft

12

"*. <•<<- %; %.**'*.* % **> x * 4

^>* : ^tK*

.JK v - *■ . A ^*V«*»<*^

FIG. 5. Ponar grab sampler in closed position, being brought on deck.

13

investigations . The standard on- station procedure during the macro-
benthos sampling was: the ship was stopped on station; the depth, in
feet, was recorded from the fathometer and converted to meters; stand-
ard marine meteorological observations were taken; water transparency
was determined by a white secchi disk; the surface temperature was
measured with a thermistor; a bathythermograph cast was made; and
three bottom grab samples were taken.,

The sampler was opened and the sediment transferred into a
large washtub in an undisturbed condition.. The sediment texture was
classified as sand, sandy silt, silt, layered, and hard bottom after the
method described by Powers and Robertson (1968), Layered sediment
consisted of a layering of silt™ clayey silt over stiff plastic clay, and the
hard bottom sediments included gravel, pebbles, and glacial till.

The macrobenthic invertebrates were separated from the sediment
by means of an elutriation- screening device (Powers and Robertson
1965) (Fig. 6). A whole sample, including washings from the bottom
samplers, was transferred from the washtub to the hopper of the
separating device. The sediment was agitated by hosing with water
after which the hopper was tipped to pour the supernatant water with
suspended organisms into an attached cylindrical screen of 0. 5 mm
mesh. The fine sediments passed through the screen while the organ-
isms were collected in a pint mason jar that was attached to the screen
by a canvas sleeve. The hosing, agitation, and decanting were repeated

14

bo
u
o

o
u

a

0)
CO

T3
U

>

b0

<D
<D
U
O
CO

a

o
•ft

is .

•H CO
2 ^

H w

a)

SO B

a §

A'

15

until the macrobenthos was separated from most of the sediments.
The organisms were preserved with formalin buffered with calcium
carbonate.

Triplicate sediment samples for the determination of the organic
carbon content of the surficial sediment were taken at each station in
July and October 1965 and in April 1966. These samples were kept
frozen until analyzed in the laboratory. The sampling of the sediment
has been described by Powers and Robertson (1968). They found no
significant differences between July, October, and April samples so
the average value of each station was considered a constant during the
subsequent statistical analyses.

The bottom temperature, read to the nearest 0. 1°C, was deter-
mined from the bathythermograph (BT) data. Standard observations of
wave height were obtained from snyoptic meteorological information
collected by three car ferries that daily traverse Lake Michigan,, The
CITY OF GREENBAY crosses the lake from Frankfort, Mich. , to
Kewaunee, Wis, , on a course that approximates our E transect, while
the CITY OF MIDLAND crosses from Ludington, Mich. , to Sheboygan,*
Wis. , on a course near our D transecto The CITY OF MADISON crosses
from Muskegon, Mich. , to Milwaukee, Wis. Its route is somewhat
north of our C transect but close enough to give a fair estimate of
conditions.

16

Short-Term Study Area

The macrobenthic fauna of the short-term study area was collected
from two 15 x 15 ft areas of the lake bottom., designated study areas 1
and 2 These areas were sampled in detail by divers utilizing a hand
coring device patterned after that described by Fager et aL (1966)

(Fig* 7). The steel coring tube was 7 inches in length, had an inside

2

diameter of 2, 05 inches, and sampled a surface area of 0. 022 ft . The

butt-end of the corer was fitted with an adaptor made of rubber hose,,
machined so that a wide mouth pint mason jar could be inserted over it.

The 15 x 15 ft "sampling frame" used by R. F. Anderson , con-
structed of steel pipe and supported by 4 legs 3 ft* long, defined the
boundaries of the study area A grid network consisting of plastic line
with attached number tags was arranged on top of the frame to position
sample transects and to space the samples., The frame also provided
support for the divers, permitting sampling with a minimum disturb-
ance to the bottom sediment and organisms

The frame was assembled on the beach with flotation tubes attached
to each corner. It was then towed to the study area, lowered to the
bottom, and positioned by the diver So Two divers working together
were able to take samples rapidly* A wide mouth mason jar was
attached to the coring tube which was then inserted two inches into the

A study of fish predation on the macrobenthos of Lake
Michigan,, Paper presented at the 11th Conference on Great Lakes
Research, Milwaukee, Wis„ , 1968.

17

&&$&&£ i ;7*< - "

FIG. 7. Hand-coring device used to collect benthos samples at the short-term
study area.

18

bottom. The corer was then tipped to one side, closed off by a metal
plate, and inverted, resulting in transfer of the contents to the jar. A
numbered lid was placed on the jar after the coring tube had been
removed. Divers were able to observe and report any organisms that
excaped when the corer was removed from the mason jar. Such losses
were rare. Formalin was added as a preservative after the samples
were returned to the surface.

In study area 1 a systematic sampling procedure was utilized.
Forty-five samples were taken at 1-ft intervals along 6 transects each
7 1/2 feet long that radiated in a spoke-fashion from the center of the
sampling frame (Fig. 8). Two samples were lost. Study area 2 was
situated adjacent to study area 1. Forty-five samples, representing
triplicates, were taken in 15 randomly chosen quadrats selected from

36 quadrats located within the boundaries of the 15 x 15 ft sampling

2

frame (Fig. 8) a The area of each quadrat was 6. 25 ft .

LABORATORY PROCEDURES
Long - Term Study Area

In the laboratory the organisms and residual sediment of a sample
were placed in a white porcelain pan. Each organism was counted and
placed in its respective taxonomic group. Pontoporeia affinis,
Oligochaeta, Sphaeriidae, and Chironomidae constituted the entire
sample. Occasionally additional organisms such as leeches, snails,

19

Study Area 1
Design of the Systematic Sampling Transects

Study Area 2

Design of the Random Sampling Area
(Asterisks Indicate Quadrats Sampled)

* * *

* *

* *

FIG. 8. Design of the systematic and random
sampling procedures of the short-term study
area.

20

roundworms, flatworms* mysids, ostracods, and bryozoans were found.
They were not included in the analysis because they were not represented
through the entire depth regime of the lake.

The telson- rostrum length of Pontoporeia from stations A-5, B-4,
B-5, B-6, B-8, E-2, and E-6 for the late April-early May, late May-
early June, and October sampling periods of 1965 was measured to the
nearest millimeter with an ocular micrometer inserted in a dissecting
microscope. The telson-rostrum length was also measured for the first
five stations of the C transect for the 1965 and 1966 sampling seasons.
At first each specimen was straightened with a pair of dissecting needles
during the measurement but it soon became apparent that the organisms
were telescopic and that this straightening procedure resulted in some
bias. It was found, by comparing straightened organisms with organisms
lying in their natural state, that with experience they could be measured
accurately while they remained in their natural position.

The brood chamber of each adult female from stations of the A,
B, C, and E transects for the 1965 sampling season and the late March-
early April, late April, and early June sampling periods of 1966 was
examined to determine whether it was brooding or spent. Brooding and
spent female amphipods were recorded for the first five stations of the
C transect for the entire 1966 sampling season.
Short - Term Study Area

The organisms were separated from the sediment using No. 5

21

bolting cloth (0,282 mm aperature size), counted and placed in major
taxonomic categories. The taxonomic groups considered were the
same as for the long-term study: Pontoporeia af finis, Oligochaeta,
Chironomidae, and Sphaeriidae. The individuals within the Pontoporeia
population were of two size groups, approximately 2 mm and 7 mm in
length, and each group was treated independently during subsequent
analysis.

RESULTS

LIFE HISTORY AND REPRODUCTION

The development of Pontoporeia affinis is direct through the
immature instars. Sexes become distinguishable in later instars on the
basis of obstegite formation in the females and increased length of the
antennal flagella of the male (Fig, 9).

The characteristics of the adult male are morphological modifi-
cations that are considered adaptions for a pelagic existence. The
elongation of the pleopods and the setation of the last uropods increase
its swimming capabilities. SegerstrSle (1950) felt that the adult male
seldom lives longer than one week* and apparently does not feed as its
alimentary canal is invariably empty and its mandibles are reduced in
size. The adult male ranges from 7 to 9 mm in length in Lake Michigan.

The thoracic region of the adult female becomes more robust in
appearance with the development of the large marsupial plates., or
obstegites, that form on the inner surface of the coxal plates of the
second to fifth pereiopods. The adult female ranges from 6 to 9 mm in
length in Lake Michigan Q Juday and Birge (1927) found the number of
eggs carried by each adult female ranged from 13-28 with a general
average of 20.

The female dies after producing its brood. The body begins to
degenerate during the course of the incubatory period., becoming trans-

22

23

FIG. 9. Morphological features of Pontoporeia affinis. A, male x9;
B, basal portion of first antenna of male; C, anterior end of female;
a2, second abdominal segment; a6, sixth abdominal segment; cp, coxal
plate; p3, third pereiopod; pi, third pleopod; t, telson; ul, first
uropod. (Modified from SegerstrSle; copied from Pennak)

24

lucent, while the gills lose their leaf-like form and become thick and
opaque. The young are slightly less than 2 mm long when they are re-
leased from the brood pouctu This reporductive cycle distinctly contrasts
with most amphipod species which produce successive broods at intervals
of a few weeks during the summer months.

During the daylight hours, in the shallower inshore regions, the
Pontoporeia population remains burrowed in the surficial sediments, but
at night some members of the population make excursions up into the
water column. However, C, F. Powers (personnal communication), in
observations made while participating in the 1967 STAR II submersible
dives in Lake Michigan, reported seeing large numbers of adult males
swimming in the water column in the deeper portions of the lake during
the daylight hours.

Wells (I960) and Marzolf (1965) reported nocturnal vertical
migrations of Pontoporeia in Lake Michigan., Marzolf sampled the
bottom sediments while examining the vertical migrations and found that
only the larger amphipods migrated while the majority remained in the
surficial sediments. SegerstrSle (1950) noted that adult males spend a
longer time above the bottom than adult females and immature organisms.
Mating is thought to occur during the nocturnal migrations of the adult
males and females.

Brooding and spent females were found in the lake as early as the
21 March and occurred in certain localities until November (Tables 2 and 3).

25

O >

■a

i

d

.tH

CO

m ;

• «-J

' 4->

CI

cj

^ u

• 1-1

<4H

r\ O

OJ

-d

«J

o

*H

d

CD

o ^

u

a

CO

"8

CD

(M P.

i CD
l> CO

4->

CO

d

CD

o

U

Pn

Ph

<L)

a

<d u

^ b£

r-H

«> w

^ 3

— ■» r-H

CO d

spent (
nt fern

bs

CM £

1 rj
2 ^

(U

TJ PM

CJ CQ

s ^

a*

d h

Q Jh

d d

m °

»-i ^

CM CM

w. PJ

CM i

&0 -rH

a tj

broodi
o broo

«1

<+H fi

vO ^

O 03

^H 1

£ ^

d *

* £

« CD

Ph j3

:urr
epr

< S

r~ tf

u h

r-H I

o

o O

>>

H „

rd lO

<■ — ,.

5. Mont
for 196:

* 2

a. <u

™ a

w a

d

•J -a

o

n -S

•r-t
H->

< ■■*

Ctf

H

s

CO

o o o o o

CO CO CO

oo^ooooooo^PQoooPQfflpq i <£
pq PQ PQ

qoooo iocoWoo^ooooooooooocooooo

CO

oo i co iooo icooffloooco^oooooooocoo

CO CO CO CO CO

ooPQfqooo^^^PQoooo^^ooPQfflooooPQo
W PQ pq pq pq

offlcoPQ ioopq«pqpqpq «o i i

i i i i i i t i i i

oofflopqoopqooto ipqoooooooo

oofflofiooffl iocooPOooPQcocoooWoooooo

cocooopqpqpqpQ » « to o pq co pq PQ pq i PQ t pq co W , ooffl

ooino^mooohooooofoir)

oor-ooc^-oin^t^-rHsococo

r— I r-H i— t (\J (\3 r-t

I I I I ! I 1 I 1 I I I I I I ! I I I I 1 I I I I I I

<!<'<< <<««Wfflffl««fflU-UUUUUUWWWWWH

26

ctf d

o

o

o

a)

CO

a)
ft

<D

CO

S

0)

CD Jh

MH <D

CO w

t* °

0) bJG

ft a

T> o

fl o

0$ U

PQ t>

bJO >

•i-i <L>

O JQ

O O

O co

ftj

u
o

o

o H

o

PQ u
<! a

H 2

>
o

CO

CM

ox

3

CM

OJC

2

o

CO

I
m

CM

«i3 ftJ

CM l

ft 0)

CD 4->

o
+->

co

i o o o PQ o

! i O O CO O O i i i e

i o PQ co o o

I o o o o o

o o

CO

ffl

oo in

r-H CO

i i t i i i it i i i i o PQ PQ PQ co o PQ ! i t t i i

CO CO CO

^aooo^m.offlpq^oofflpqooopqoffloPQoo
ffl PQ ffl

CO CO CO CO CO CO CO

° ° n n w m m ° ° ° « ° n n ° ° w ' B ! ' '■ 8 ' B

cococo cocococo coco

o^^oooc^c^ooc^oofOlnHoo^oo^c^lf1i^^H^i)cof f )
r^r^^r-Hr-n^vx>cMOoo^'^(Mu^t^oa^c>Lo^o^r^r-H[^-co

r-Ht-H ^h,_h H (\J (\] H

—I CM

I i i ( i i » i i i i i

I 1 I I I 1 ! I >

«J<«J<fflpqmfflfflmpqmuUUUUUUWWWHWW

27

The only winter benthic samples were obtained at stations C-l, C-2, and
C-3 on 27 January 1967, and no brooding or spent females were observed
in those samples.

It appeared that at depths less than 35 m, in Lake Michigan* reprod-
uction is completed by late May-early June. Beyond 35 m it seemed
that Pontoporeia bred intermittently throughout the sampling season. An
examination of the bottom temperature records indicated that this amphipod
did not breed at any depth when the bottom temperature exceeded 7 C.

The fact that adults die after reproduction simplifies the recog-
nition of year classes. The size frequency distributions for station B-8*
average depth 11 m, described a population of amphipods that matured
in one year (Fig. 10). The two major size groups of 2 May indicated
that the population was in the middle of the breeding season., The large
2 mm size group and small proportion of larger individuals on 2 June
showed that reproduction was in its terminal phase. The large 5-6 mm
size group of 15 October indicated that the cohort of the young of the
year has grown approximately 4 mm in six months .

At station E-6 where the average depth was 33 m 5 amphipods re-
quired two years to mature. Both brooding females and newly released
young as well as juveniles from the previous year were present in the
21 April sample. Recently released young, juveniles from the previous
yeaij and a few adults were present in the 2 June sample. The two size
groups of the 5 October sample represented the young of the year and the

28

Station B-8, average depth 11 a

1.0-
.8-

& .4-
.2-

(146)

■ i .i '

1

1.0

.8

.6
.4

1

(364)

2 3-4 5-6 7-8 >8mm
2 May 1965

i * . i * ; .i

1.0

.8

.6
.4
.2J

(242)

2 3-4 5-6 7-8 >
2 June 1965

2 3-4' 5-6 7-8 >8mm
15 Oct. 1965

Station E-6, average depth 33 »

1.0 -4

.8

|.6

.4-

.2

1

(373) 1.0.
.8 -J

.6
.4-!

I .2

J.

23-4 5-6 7-8 >8mm
21 Apr. 1965

2 3-4 5-6 7-8 >8ara
2 June 1965

•o_

(225)

• 8 1

.6 -

.4.

.2-

1 1

T-i-r-

1

2 >4 5-6 7-8 >8mm
5 Oct. 1965

FIG. 10. Size frequency histograms of Pontoporeia affinis for selected
depths of three sampling periods late April-early May, late May- early
June, and early October 1965. Number in parenthesis represents the
sample size.

29

Station A-5» average depth 43 »

1.0-j
.8

S.6H

•2-

L

(254) i'Qj
.8.

.6-
I .2.

±

(80)

1.0 4
.8

.6-
.4.
.2

2 3-4 5-6 7-8 >8mm
4 May 1965

2 3-4 5-6 7-8 >!
4 June 1965

(33*0

2 3-4 5-6 7-8 >8mm
5 Oct. 1965

Station B-6, average depth 83 m

1.0 A

I?

i.6-1

.4-

.2-

<138) 1.0.

.8-
.6.

.4-

.2-

— T— 1 ■ I

2 >4 5-67-8 >8mm
2May 1965

L

(212) 1.0
.8

.4j

1

.2.

2>4 5-6 7-8 >8mm
3 June 1965

(138)

2 3-4 5-6 7-8 >
14 Oet. 1965

FIG. 10. Continued.

30

Station B-5, average depth 108 m

1.0 H
.8
§ .6

6 .4-

.2-

(100) 1.0 i
.8

.6-1
.4

.2-1

(181)

2 5-V ' 5-6 ' 7-8' >8mm 2 3-^' 5-6'7-8'>8mm 2 '3-4' 5-6' 7-8 > 8mm

1.0 -

(<

«)

.8 -

.6-

.4-

.2-

1,1,

L rr J

-H-

2 May 1965

3 June 1965

15 Oct. 1965

Station B-4, average depth 129 m

1.0

.8-1

i .6

•2.

(124) 1.0 -
.8-

.6

.k-

.2-

(1^3)

1.0-
.8-

.6-
.4-

.2-

(99)

2 ■ J-b'SS^^^&m 2 , 3-V5-6 , 7-8 , >8mra

27 Apr. 1965 27 May I965

FIG. 10. Continued.

2' 3- V 5-6' 7-8 >
15 Oct. 1965

31

Station E-2, average depth 197 n

1.0 -

.8 .

§ .6

.2 A

(214) 1.0

.8

.6

.4-1

.2

L.

i ■ i

J.

W?)

i.o H

.8

.4

.2J

(62)

2 3-4 5-6 7-8 >8nm
20 Apr. 1965

2 3-4 5-6 7-8 >8mm
24 May 1965

2 3-4 5-6 7-8 >
4 Oct. 1965

FIG. 10. Continued.

32

young of the previous year. At this station, growth was less than 2 mm
for the 6 month interval from April to October, which was about half
that observed at B-8.

Figure 10 also shows that stations deeper than 35 m have size
frequency histograms for Pontoporeia which are similar to each other.
In the 35-86 m depth zone a larger proportion of individuals fell in the
2 mm size interval than was the case in deeper regions of the lake. It
was difficult to determine how much time is required for the maturation
of Pontoporeia at depths greater than 35 m.

Monthly size frequency histograms were made for stations C-l,
C-2, C-3, C-4, and C-5 for April to November 1965 and from March
to November 1966 (Figs. 1 1 to 15). Pontoporeia found at station C-l,
average depth 20 m, also required two years to mature. The remaining
four C stations were positioned in waters deeper than 35 m. Stations
C-2 and C-3 lay in the 35 to 86 m depth zone and showed a higher pro-
portion of individuals in the 2 mm size class than did stations C-4 and
C-5 which were in deeper waters,

PATTERNS OF SPATIAL DISTRIBUTION
Short-Term Study Area

Patterns of spatial distribution in the short-term study area were
examined by two methods: by plotting the mean against the variance and
by examining the occurrence frequency of the number of individuals per
quadrato If the mean number of individuals per quadrat is independent

33

Station C-l, average depth 20 m

1.0 j
.8
§ .6J

£ .4-

.2-

(283)

2 '^'S-^-S'^mm
1 May 1965

1.0 .

(277)

.8_

.6-

.4_

.2^

.1.,... 1 .

. , ,i

1.0

.8.

.6
.4

.2-

2 '3-^5.6 ^-S'^mm
1 June 1965

(133)

1

2 >4 5-6 7-8 >8mm
10 Aug. 1965

1.0 -
.8 -

£.4 -I
.2

(85)

1.0
.8^
.6

.4 -J
.2

(400) 1.0-

.8-

.6 -

.4 -

.2

1

I . 1

2 3-4 5-6 7-8 > 8mm 2* 3-4 ' 5-6* 7-8 >8mm

7 Sept. 1965 10 Oct. 1965

(214)

2 3-4 5-6 7-8 >
6 Nov. 1965

FIG. 11. Size frequency histograms of Pontoporeia affinis collected
at station C-l for the 1965 and 1966 sampling seasons. Number in
parenthesis represents sample size.

34

vo

NO

T

o

00

NO

cm

. A vo vr\

vo vr\

r-j

**

esj

oo

~r

— r

CM

00

A

"vo *"*
J t •
■£ g

<^ ON

CM

CO

T-l

oo

vo

"T

CM

T
O

00 vo

T
CM

<9 *-N

A v\

^vo Os

<N»

21 s

V>

r^ cm

CM

I

00

nr

00

vo

CM

- r

vo

"nr

■nr

CM

00
A so

VO

I

CO
<^CM

CM

CO

•^CM
CM

0)

o
U

S
h

r

o

—) r

00 vo

CM

^oueubeoa

a cm

"3U

^ CM

CM

r
o

"T

00

VO

— r

-a-

CM

^ouenbeoji

a vo

"vo

i$

L.^^

CM

35

Station 0-2, average depth 50 m

1.0 -

.8 .

g.6 .

£.4 -

.2 -

(426) 1.0-
.8 .
.6,

.4-

.2-

(421) 1.0
.8
.6.

•4-
.2-

2^4 l 5-6 n 7-8>8mm 2 3-4 5-6 7-8 *8mm
29 Apr. 1965 1 June 1965

(430)

2 >4 5-6 7-8 >8mm
10 Aug. 1965

1.0 H

.8 -
g.6
£.4

.2

1

(290) 1.0
.8

.6
.4

2 3-4 5-6 7-8 >
7 Sept. 1965

.2 -

2 3-4 5

(204) 1.0 _j
.8

.6.
.4.
.2.

6 7-8 >8mm
10 Oct. 1965

(368)

2 3-4 5-6 7-8 >8mm
6 Nov. 1965

FIG. 12. Size frequency histograms of Pontoporeia affinis collected
at station C-2 for the 1965 and 1966 sampling seasons. Number in
parenthesis represents the sample size.

36

o

— r-

00

"T"
vO

-3-

CM

Op ^

/I VO H

3>&

Cvl

o

CO

— r

T
CM

- 00
-A vO

CM

On
CM

r
o

t

00

VO

— r

CM

OS

r

o

"T

CO

- r

vO

CM

A o

m ^O ^O

as*

vO <d
CM

o

JL CM

CM

o

nr

00

00

"T"
vO

vo

T"
3-

T

CM

— 00

2:

<s>

.it

VO

CM

00
A
U vO

CM

^ to

•t Ov

CM

§

O

O

CM

O

CM

O

—i r

00 vo

CM

&D\IOVi>OJ£

* VO00

vo
I

^CM

CM

o

T

00

"T"

vO

"T"

CM

iCouenbeoj

CO

-^ \o
<?&

^> &

CM

37

Station C-3» average depth 77 m

1.0 -|

.8
I.6J

£ .4.

.2-

(20)

11

2 T yJ* r 5^6 T 7-8>Qnm
17 Apr. 1965

1.0 -

(303)

.8 ,

.6_

.**_

.2-

J_

1 1

2 "3-4 5-6 7-8>8ram
1 June 1965

1.0.

.8.

.6.

A.
.2-1

(175)

1

2 3-4 5-6 7-8 >8mra
11 Aug. 1965

1.0 -J

.8.
§.6J

£.4-

.2-

(256) 1.0.
.8_
.6.

.4-
.2-

2 >4 5-6 7-8 ^8ram
7 Sept. 1965

(125) 1.0-

.8_
.6.

.4-
.2.

2 3-^ 5-6 V-8'>8mm
7 Oct. 1965

(197)

2 '>4 5-6 ' 7-8" >8mm
6 Nov. 1965

FIG. 13. Size frequency histogram of Pontoporeia affinis collected
at station C-3 for the 1965 and 1966 sampling seasons. Number in
parenthesis represents sample size.

38

00

o

nr

00

CM

f

VN

ST

- CM

r
o

""T"

00

nr

NO

nr

VO

u*l

.*

<M

CM

CM

0Q

r

o

00

—r

NO

CM

T

00

— r
vo

CM

„C"w

00

A

vn

8 s

.21

CM

T"
O

A vo

o

CM

Aw

CM

I

00

00

T

4r

"T"
vo

t

CM

T

CM

A \o

■•9*

VO

*f

CM

CO

0? 3v

VO

to

■AS

CM

CM

0)

PI
o

CO

r-H

£

r
o

00

"n r

vo ^

CM

vo

00

■s*

- CM

*~7

I

vo

—r

CM

iCouenbej^

CM

39

Station C-'i, average depth 108 m

1*0 _
.6-
.2-

(199) i«0.

.8-
.6_

.4-
.2-

i

2 ^ '5-6 7-8V8mm
1 June 1965

(326) 1.0
.8-
.6.

.4-
.2_

2' 3I45-6V-8 >8mm
11 Aug. 1965

(11*0

2 ' 3-V 5-6 7-8* ^8mm
7 Sept. 1965

I.Oh

.8-1
g .6-1

.2-

(5)

.0,

(1^5)

.8-

.6-

.4_

.2-

■ 1

... 1 . 1

2 3-4 5-6 7-8*8mra
10 Oct. 1965

2 3-4 5-6 7-8 ^8mm
6 Nov. 1965

FIG. 14. Size frequency histograms of Pontoporeia affinis collected
at station C-4 for the 1965 and 1966 sampling seasons. Number in
parenthesis represents the sample size.

40

CM

V

O

00

"T"

— r

CM

- Op NO - ^

j

4 >»-/ ^ ^

■'**

CM

r"
o

— r

00

NO

00

A \o

VO

I

CM

l

ON

CM

CO

00

— r

NO

"T"

■3"

— r

CM

00

A ^

rn ON

28*

M? (J)

A **

CM

00

VO

CM

■ 8

00

NO

I
VfN

CM

A®

NO
00

r
o

00

— r

VO

— r

— r

CM

00

A \o <^

"0?^ n5

d ^ W

VO
I

CM

A

CM

00

VO

- r

■a-

CM

00
A \o

o
u

o

CM

O 00 VO ^J*

— r

CM

00

?88

\0 ***

A

CM

CM

— r

00

VO

— P

-t

CM

jfoueribeja

00
-A vO

NO

I
.VTN

3

1" ■*

<M

41

Station C-5t average depth 157 m

1.0-

.8-

I -6-

k .4-

.2

(170)

1.0

.8.
.6

.2-1

(201)

1.0.
.8
.6
.4.
.2-

2 '3-4' 5-6' 7-8' > 8ran 2 i'3-4' 5-6 '7-8 '> 8mm

1 Kay 1966 1 June 1966

(159)

11 Aug. 1966

1.0 h

.8 -

§•6

.4-

.2 -

1

(125) 1.0-|

.8

.6-

.^

.2-

2 3-ii '5-6' 7-8'^ 8mm
7 Sept. 1966

(7*)

2 '3-i '5-6' 7-8-
10 Oct. 1966

1.0-

.8-
.6-

.2-

1

(110)

2 '3-4 '5-6' 7-8 '^8mm
6 Nov. 1966

FIG. 15. Size frequency histograms of Pontoporeia affinis collected
at station C-5 for the 1965 and 1966 sampling seasons. Number in
parenthesis represents sample size.

42

3

O

— r

00

— T
vo

T
CM

CO ^

,-A VO CM

CM

O

T
oo

VO

T

CM

00

IS 5

I

A<*

CM

£

r
o

— r
oo

vo

CM

00

Si

VO

CM

00

T"
o

—r

00

vo

CM

00

A \o

CM

(1)

o
U

r
o

— r

00

vo

— r

"T"
CM

00

• A vo -^

tfi

CM

- CM

O

— r

00

vo

CM

CO
-A VO

°?$

7 &

CM

If)

o

S

r
o

— r
oo

r
vo

T

•

jCouoTibej^;

CM

••as
.si

J CM
CM

T"
O

00

vo

CM

jfouenbej^

oo

"?$

_A^

CM

43

of its variance, then the population is normally distributed. If the mean
is proportional to the variance and the mean- variance regression has an
intercept of zero and a regression coefficient of 1, then the population
conforms to the Poisson distribution, which implies that the population
is randomly distributed within the sampling area.

The mean and variance of the 15 sets of triplicate samples from
the short-terra study area 2 were plotted against each other for the
small, large, and total Pontoporeia (Figs. 16-18). The resulting
correlation coefficients indicate that the mean is independent of the
variance for small and total Pontoporeia (r = . 226 and .415, d. f. = 13)
and that these two groups conform to the normal distribution. On the
other hand, both the mean- variance regression coefficient and the inter-
cept are approximately one for large Pontoporeia, suggesting a Poisson
distribution.

The occurrence frequency for the numbers of individuals per
quadrat was determined for each of the three categories from the com-
bined 88 samples of study areas 1 and 2. Small and total Pontoporeia
were treated in such a fashion that 3 to 7 individuals per interval were
combined and called 5, 8 to 12 were combined and designated 10, and so
forth.

An examination of the occurrence frequency histograms (Figs. 19-
21) indicated that the distribution of small and total Pontoporeia tended
toward the normal while large Pontoporeia followed the Poisson distri-

44

vD
CM
CM •

II

CO
U tH

o
o o

o
o

o

o

o

o

o

o

o

o

CM

O
O

o
o

vO

o
o

o
o

o
o

CO

o
o

CM

o
o

^pjpg nb/q^guai wj: uiui ^ sjutjjb

CO
•H

cd
cd

•H

a)
o

u
cd

a to
<d a
> a>

o

n 1
<u
,n cm

3*

(U

cd

M

(1)

3

m
cd

91 H

cd

O
ft

+■>

0) CO

0) u

>

'<-> J,

o

.rt CO

cd

cd cd

CD U

1 3
u

cd a)

• HI Q.

Cd r£

^ bo

<D a

A CD

a cvj

45

X

<f

<*

o

•

rH

+

m

CM rH

m <r

•

00 •

M-l

• +

•

XI

II II

CO

>> u

rH

\

\

\

I- -*

CM

- O

00

» vX>

\ •
\
\

\
\

° >•
oo

\

\

r"

"T"

CN

I—
O

v o

00

CN

^BjpHtib/q^Suai uf uiux ^ sxujjjb
HTaaoao^uo^ axT U3An f 3° ^aquin^ 9q3 jo aouBTJtBA

CO
•H

cd
•H

0)

o
oj
o
u

o

Pn

CD

cd
cd

00

a

cu

OJ
00

cd
cu

3

CO

• M

$^

.1-4

VH

*

cd

•i-i

cu

?H

o

•

ft

cd

o

CD

4£

fH

a

cd

o

&

£

0)

5*

rH

4->

• H

CO

a

cu

6

5

CD

«^>

<-H

i

O

+>

ft

CO

CO

£

CD

o

4->

4->

cd

t-H

B

0)

O

fc

U

«

<+h

cd

•»

CD

+j

a

cd
u

i

TJ

<v

cd

u

3

G

cr

cd

• H

u

M

CD

^

ft

>

r£

CD

4->

H

OJO
CD

i-H

•

P|

r^

•r-«

i-H

g

6

g

I-H

t>

h

}

46

o

00

o

o

CM

II

CO

o
o

o
o

•o
in

o

o

CO

o

CM

o

r-J

o
o

CO

o
o

o
o
vo

o
o

o
o

o
o

CO

o
o

CM

o

o

C
> 4J

cd

m cd

o 3
o j

u \

CU CO

!z; <-h

»+H

fH Cd

cd

4J cd

O -H

H CD

o
a

4J o

o o

CD
60

cd
n
CD

3

CD

0)

fH

o
a
o

o

CD

&

•>— >

m
O

u

CD

8

3

Gj
O

<D

m
O

a*

• iH
r3
CO

a

• 2 «;

cd <i>

2 *
5 £

cd 3

<D

CO

1

CD

a

a

cd

d o
cd w

rd CD
4->

00 °

o -9

47

10

z

00

d

» CM

IX X

U. CT U ODlil ZU>-

10

48

CO

c\i

ID

cm x5

00

iL^LJODliJZU)-

_£

4)

^ A

h

* s

-N

O
p.

CO CD

<

-CD U

<

O

o

■term
Line m

or

(Z

• TO

O

Q

o u

H O 4)

8 ^ ^

-in Q-

<

O

Z>

O

X

histog

of th<

es wit

DL

I-

n .r-4

o w

frequenc
er quadra
oisson se

5*

CD Ph a.

-|5

-oz^

e occurre:
m in lengt
Lposed on

6 a «

• v pu

O . d

<^ CO

CO

•w

,K G

OJ

C5 «H

CD

fe

d

0}

49

•

,o

to

0)

•

z

i-i

£>

o

juveni

ormal

CO <N

"oo

<

ri* .

a:

Q

M

^t ^ *-.
ix *:=

.o

N

.O

CD

-O

in

<

o
<

LU
CZ

O

Q_

o

H
Z

O

stogram of the total numbe:
tudy area superimposed on
and variance.

.o

GI-

ency '.
■term
j me a:

^r

LL

o

§•* a

J o m

£ rCj CO

_o

a:

CO

UJ

m
2

urrenc(
s of the
with th

O

3

<D
>

(M

Z

_l
<

. The c
reia affi

o
o

.O

h-

-H O

2

v-

o

1-

FIG.
Ponto

4->
CO

• H

t

» < ' 1 ' 1

i i

i i J

o

00(O^C\lOeD(D^C\l

u_ cc lj O z> lu z (J >

50

bution. The theoretical curve for the normal distribution was calculated
for small and total Pontoporeia utilizing the equation: m= (N i/ S/2-rr. )
e -(X-X) / b ^ Tlie theoretical curve for the Poisson distribution was
calculated for large Pontoporeia by: m=Ne~ X (l, X, X /2 ! , X 3 /3!,...
X J /j!). In the above equations N represents the number of observations,

X the average number of amphipods per category, i the size of the class

2
interval, S the variance, S the standard deviation^, e the base of the

natural logarithm, and m the expected number of individuals in an interval.

The theoretical curves were superimposed on their respective histograms.

The chi square test for goodness of fit comparing the theoretical
values with the observed values was calculated for each amphipod cate-
gory. Because the probability of occurrence is exceedingly slight at the
extreme ends of these curves and the 88 samples were not enough to
represent adequately the distal areas of these curves, it was necessary
to pool the extreme observed and expected values. The terminal inter-
vals of the observed and expected values for each tail were combined
until their sum approximated 3 organisms for the combined intervals .
The following intervals were combined: 0-10 and 75-85 for small amphi-
pods, 6-8 for large amphipods, and 0-10 and 75-85 for total amphipods.

In computing the normal distribution curves, the mean and standard
deviation were estimated, thus the degrees of freedom for these curves
are k - 3 where k represents the number of intervals. One degree of
freedom is lost in estimating the chi square distribution. The degrees

51

of freedom for the small and total Pontoporeia are 11. Only the mean
was estimated in calculating the Pois son distribution so the degrees of
freedom for this curye are k - 2. The degrees of freedom for large
Pontoporeia are 5. In every case the chi square test was not significant,
indicating that the observed histograms did not deviate significantly from
the expected distribution.
Long- Term Study Area

The mean and variance of the replicated samples taken at the 35
stations of the long-term study area were plotted against each other for
total Pontoporeia per square meter (r = .417, d. f. = 575) (Fig. 22).
The correlation coefficient computed for this plot implied a significant
interaction between the mean and variance. The square root transfor-
mation was calculated for the data from every station stop. The mean
and variance of these transformed data were plotted against each other
(r = . 076, d. f. = 573) (Fig. 23) and the mean appeared to be independ-
ent of the variance. The cube root and logarithmic transformations
were also calculated for these data but the resulting correlation coeffi-
cients of the mean-variance relationship were significant. These results
indicated that the square root transformation best removed the inter-
action that existed between the mean and variance.

On the other hand., the cube root transformation best removed the
interaction that existed between the means and variances of the oligo-
chaetes, sphaeriids, and chironomids. Transformed data have been

52

o

rH
X

*- °

rH

M-l

9

m
m

o -

o o

o
o

o
o

o
o

o
o

o
o

o o
o o

o
o

o o

o
o

o
o

o o o

o

o o O

o o

o o o

o o

o o o

© o o o o

o o

o o o

o o o o

o

o o o

o o o

o o

o o

o o

o

rH

X

r-
o

LO

o

— T-

o

en

— r-
m

— T"

o

CM

— r

m

CM

CM

o

00

— r

m
en

o

o o

o o

o o
o

o o

o o

o o

o o

o o

o o

o o
o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o

o o
o
o

o

LO

CM

d

•i-i

CD

M

m

O

•

P.

CM

CM

*

O

m

O

CM

CM

CQ

|

CD
r-l
ft

CQ

a

m

•H

•

a

0$

rH

CQ

crj

"0
CD

aJ

4^

• r-l

<tf

a>

u

O

u

• iH

•

o

i— 1 •

m

0* Ctf

rH

CD CD

4->

U U

m

o

PL,

0) *

CM
rH

*H

MH

V CQ

r-f

onship
-term

o

2

o

£

£ bJO

rH

oj a

a>

r-*

r3

<D rH

4->

u u

m

a rd

•

aJ +>

r^

fl

CD <+h

a)

s °

s

1 fH

CD CD

O

s a

m

r-l CD

•1

m

J ?

CM

. 22. T
nis per s

o

2

>*

xu/stuxjjB -Btsjodo^uod; jo joquin^ aq; jo aoumi^A

53

. o

CO

•

rH

u

• rH

o o

+2

%

o

o

o

en

rH

2

U
O

m
to

o

I

■si

o

o

o

. o

fi

CO U

7i +»

©

o

o
o

o
o

o

CM

rH

s

s °

o

•

o

o

CO <D

+

T3

o

o o
o

o
o

o
o

■s

1

4-> P

II

CO

o

o

o

o
o

1-1

a cq

u

m

o

o

o

it

i

• rH w
rH

o

o
o o

o

o

o

o
o
o

A at

rH rH

o

o

o

o

- o

>*•

„> *

o

o

o

o
o

o o

o

o

o o

o
o
o

o
o
o

©

c^

CM
C9

o +■>

o

o

o

o
o o

o
o

o

eI

U CQ
05 M

o

o

o o
o o

o
o

o
o

<M

o

o

o

o

m

s *>

o o

o

o

o o

o

©

"V

o

o

o

o

o

o

6

•»-* Wft

o

o

o

o
o

o
o

©
o

1

4-> bJO
at rt

o

o

O

o

- d

o

<D rH

o

o o

o

o

v£>

pi

rH «

o
o

o

o
o

o
o

o
o

o

o

o

o o
o
o

o

o

o
o
o

o

o

o

o
o
o
o
o
o

1

o

e-mean
er of the

o

o

o

o

o

o

o

o

1

Lane
met

o

o o

o

o

Jo

o

o

o

o

o

o

s

CO Lj

o

o

o

o

o

o

o

z

o

o
o

o
o

" s

0>

o

o

o
o

o
o

3

o

o

©

o

o

H

o o

o
o

o

rH

• 8

o

o

- m

8

or> A

o

o
o

o

o
o

rH

o

S3

FIG. 2
affinis

•

■"" I" " ■■

1 1 I • 1

1

1

•

d

d

o o o o o

d

d

o

;

o

o

o o o o o

o

o

o

o

oo

vo <r cm o oo

vO

<r

CM

CM

rH rH rH rH rH

(uo^vuuojsireax ^ooy e«c*nbs)

z m/*

?

piTJJPi

uteaodo^uo^ jo aoqiun^ eift jo eotrop**A

54

used in the statistical treatment of the data collected at the stations of
the long-term study area.

INTERSPECIFIC AND INTRASPECIFIC ASSOCIATIONS

The associations between Pontoporeia and the other macrobenthic
groups of the short-term study area were determined by the correlation
coefficients* The two size categories of this amphipod were treated
independently of each other in the association analysis. Figure 24 indi-
cates a significant association between small and large Pontoporeia;
while figures 25 and 26 indicate that both small and large Pontoporeia
exhibit a significant inverse relationship with the oligochaetes. Figure
27 shows a significant inverse relationship between large amphipods and
the Sphaeriidae. Correlation analysis did not show significant associ-
ations between small amphipods and sphaeriids nor did it reveal signif-
icant relationship between this amphipod and the Chironomidae.

The associations between total Pontoporeia and the other macro-
benthic groups of the long-term study area were also determined by
correlation coefficients. Figures 28-30 indicate that total Pontoporeia
were positively associated with oligochaetes, sphaerids, and chiromonids.

ENVIRONMENTAL RELATIONSHIPS

For this investigation it was assumed that Pontoporeia responds
to changes in its environment and that these changes are reflected by
increases or decreases in abundance. A low density indicates response

55

•00

II

Ctf

O
CI
O

*

o

On

o o o

Pi

o o o o

© o

o o o o o o o o

ai
bo

o o o o o

o o o o o o o

oooo o o o o

o o

— r

OOP

R

o

00

o

o

o

o

o

CM

o

d
o

'ft

4->

u
o

CO

a

fa

4->
CO

a

0)

o

CO
<D

56

r

00

CM

li

U 00

o
o
o
o

o
o

"T

T

o
o

o

o
o
o

o
o

X

o
o

o

o

o

o
o

o
o

o
o

o
o

o

o

k«

r 3

Si

o
o

I

- 00

T

VO

O

CM

VO tr\ -3* c^ c\2

CO

•I-i

•H
J*
0J

CO

a

0)

u
o

o
o

&

o

CO
CD

m
o

sxuTJj^ Bfeaoctoq.uoj axT^^AUf jo .leqiuTVH

£ cu

CD (l)

•m o

m bo

O 3

a O

tJ «

.2 *

o £

co bfl

co PJ

in
cm g

57

VO

II

o o o o

o O

o o o o o o

ooo o o o o o o o

oooo ooo oo oo

o ooo o o o o o o

o
o

o o

r

o

OD

"T"
o

— r
o

VO

o
T

o

3

— T

VO

CM

o

CM

q.Bap^t!t)/qqJii9T[ unf max g a/

1

?

^

CO
*H

a

o

T-4

1

at

J

d

o

• H

o

0)

tkO

fc

•

oo

a

P.

o

<M

*

u

O

o
0i

11

•
VO

juvenile
term stu

m *

■■3-

at ion o
e short

•
CM

. Associ
aeta of th

•
O

26
och

O

•i-i

h O

sprfjj* «feaodo^uod axpu&Aiif jo jeqnmpi

58

'■5

ii

VO
fcCO

-VO

3

0--3-

o

T"

o
nr

o
nr

o

o

O

"T"

U

o

CM

O-v*

00

NO

XTN.

c-\

CM

i

<D
U

t

Ii

1

V* o
o

0)
O "D

SI

if
.1

59

Average Number of Pontoporeia affinis/m^

o

*A

CO

T>

CM
|_

o

o

o

o

o

o

o
o

o

o

o

o

c

o

o
o

o

o
o

o

o
o

6

o
o

o

o

o

o

o

o

o

o

o

o

o
o

o

o

o
o

o

o
o
o

o

o

o

o
o

o

o

o
o

o

o

o
c

o

o
o

o

o
o
o

o
o
o

o
o

o

o

o
o

o

o
o

o

o
o

o

c

c

o

o

o
o

o

o
o

c

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

c

o

o

o

o

o

o
o

o

o

o
o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

.'•>

o

o

o

o

o

o

o
o
o
o

o

o

o
o

o

o
c

o
o
o
o
o
o

o

o

o
o
c

o
o

o
o
o

o
o

o
o

o

o
o
o
o

o

o

o
o

o

c

o

o
o
o
o
o

o
o

o

o
o
o

o

o

o
o

o

o
o

o
o
o
o

o

o
o

o

o
o
o

o

o

o
o

o

o3

<D

U

as

>?

nrt

2

P

CO

*»-N

fi

•

o

§

0)

p

<M

•H

?

E

o

H

o

<D

P

o

o

cd
P

•
*A

o

0)
o3

o

O
tiO
•H
H
O

u>

(uox^tuaojsuiMj, q.ooH ea^nbg)
jysxuTjjv ^xeaodoq.uoj jo aeqiiiriN bSvabav

-P

<D

o

o

O

u

<D

8P

l

*d

<H

05

cd
•H
<D
U
O

O
P

I

o
o

IS

O
O
CO

CO
CVJ

H

60

Average Nmber of Pontoporeia af finis / m 2

VT\

O

Tl CM
X CM

L

o
o

I

CM

o
o

o
o

o c

■a

o
o

o o

o
o

o
o
o
o
o
c

o
o

o

o
o

o
o
o

o
o

o
o

o
o
o

o
o

o o

o
o

o
o
o

o

o

o

o

o
o
o

o

o
o
o
o
o

o

o

o o

o

o

o
o

o
o

o

o
o
o
o
o

o
o
o

•o

CM

VA

o o

o
o

O - *A

o
o

o
o o

o

o

I

•

o

1

•

o
o

( uoxq^uuio js ireax

i

•
o

3.003 8«renbg)

§

•H

1

Eh

-P
O
O

CM

8

t

<X>
O

u

qj

U

U

b0

o

I— I

U

■a

CO

a

Jh
O
ft

o

ti

o

a*

m
O

a
o

a
o

CO
CO

<

a*

CM

s

u

4->
CO

/stvltjjv vxejLO<ioyxo < 3i jo aeqttmu oSvjlqav

61

Average Number of Pontoporeia affinis /m^

o

CM
CM

o
o

-JL

CN

o

o

3" •

o

3" <H

o

• •

o

+ Tf

o

o

o

It ON

C\J

o

o

fc ^

o

o
o

o

o
o

o

o
o

o
o
o
o

o

o

o
o

o

o
o

o

o
o
o
o

o
c
o

o
o

o

o
o

o
o
o
o

o

o
o

o
o

o

o

o
o

o
o
o

o

o
o
o

o
o

o

o
o

c
o
o

o
o
o

o
o

o

o

o

o
o

o

o
o

o
o
o

o

o

o

o
o

o

o
o

o

o

c

o

o
o

o

o
o
o
o

o

o

o

o
o
o
o

o

o
o

o
o

c
o

o
o

c
o
o
o

o
o

o

o

o

o

o
o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

c

o

o

o

o

o

o

o

o

o

o

o
o

o
o

o
o

o

- Os

OOO06OOO0O

VO

>cr\

o
in

o
o

o

(uoxq.*uuojs trea£ + ooh ejnrobg)
JH/stxixjjb •efajtocto+uoj jo aeqimift eS^aeAy

y-N

<D

c

&

O

■P

•H

*

O

El

0)

o
Vn

3

•H

Eh

&
O

8

-P

o

•H

O

O

«

Tj

0)

§

8

CQ
•H

•H

cd

o

•H

H

<D

*3

^

i

O

o

H

•P

K

o

o i

«H

<H i

O

o

G 05

O <D

•H fc

-P a3
a3

£5

O 2

0)

W 43
09 W

<D

O (D

tC\-P

• fctD
t5 £

H O

pq

H

62

to the environment in a negative fashion while a high density represents
a positive environmental response. Data collected from 35 stations of
the long-term study area were used in examining these environmental
relationships.

The data were analyzed by three methods. In the first an environ-
mental factor, such as depth, was divided into increments and a geometric
mean calculated for the combined amphipod counts of each intervaL The
geometric mean of each interval was accompanied by an upper and lower
limit of one standard deviation unit* In the second method a correlation
coefficient was computed by comparing the average density of Pontoporeia
with an environmental parameter. Stepwise multiple linear regression
was used as the third method of investigating these relationships. The
effects of environmental factors on changes of amphipod density are
considered simultaneously in this method of analysis.

The geometric mean is usually calculated by (X^ X2»X3 o o o X-) ^
where X\ represents a sample of amphipod counts » Counts in this
analysis were converted into logarithms with the geometric mean being
determined from the antilog of iogX^+ logX^ + logX3 + . . logX;)/j
The upper and lower limits of this mean were determined by first com-
puting the standard deviation of the log converted data, adding this
standard deviation to and subtracting it from the long mean, and the
antilogs of these two values finally constituting the upper and lower
limit So

63

The geometric mean was used because it is a more conservative
estimate of the theoretical mean and is not affected by extreme values
to the same extent as the arithmetic mean, The geometric mean also
gives smaller counts relatively more importance than they enjoyed in
the computation of the arithmetic mean. The upper and lower limits
were calculated in the previously described manner because this method
eliminates the possibility of negative confidence intervals*

The correlation coefficient was employed as the second method of
analysis because it is a useful tool for measuring the degree of a rela-
tionship. The computer program written to compute the correlation
coefficient was also designed to present the relationship graphically.
The computer automatically rounded off the results in the construction
of the graph so these values are printed in regularly spaced columns
and rows. It is impossible to interpret the relative importance of these
data points because each point may actually represent more than one
value. These graphs merely show the trends of the relationships while
the correlation coefficients indicate the degree of the relationship.

The third method, that of multiple linear regression, is a useful
statistical tool in the study of biological systems because it has the ability
of mathematically removing the effects of confounding environmental
factors. Stepwise multiple regression analysis has the additional feature
of generating the multiple regression equation, variable by variable, in
the order of their relative importance.

64

The general multiple regression equation is written Y = a ■+ bjXi+
^2X2 + . . . fr-jX.: where Xj represents an independent variable such as
depth, bj the regression coefficient^ a the intercept, and Y the depend-
ent variable. The dependent variable in this study is the density of
Pontopor eia /m while the independent variables include surface and
bottom temperature, depth, depth squared, sediment type, day of the
year, percent carbon in the sediment, distance from shore, and the
densities of oligochaetes, sphaeriids, and chironomids.

An initial examination of the depth distribution of Pontoporeia
(Fig, 31) indicated this was a curvilinear relationship which could pos-
sibly be satisfied by a second order polynomial equation. Depth was
squared and this value was treated as the additional variable that would
possibly satisfy this curvilinear relationship. In another study involving
the stations of the C, D, and E transects an additional parameter, wave
height, was added to the previously discussed independent variables.

In the first section of the following analysis, a two dimensional
presentation is made showing the effects of environmental parameters on
the abundance of Pontoporeia . This is accomplished by comparing the
geometric mean number of Pontoporeia with the environmental factors
and also by examining the correlation coefficients of these relationships.
Stepwise multiple regression analysis was employed in the second part
of this section to show how- the environmental factors, acting together,
affect the quantity of Pontoporeia.

65

Geometric Mean Number of Pontoporeia affinis /m 2

o

*r\

x CM
CM

O

•'

o

CM

O

o
o

o

(uox^uuojstrcaj; q.ooa ea^nbg)
gW/spuxjji fxeooao^uog jo •leqimjjj treejj oxaq.emoe£)

>^

•

,Q

1

270

d)

p.

u

<D
+■>
(1)

a

•

<D

o

U

_tH

rt

CM

3
cr

CO

u

CO

• r4

a

• r-l

m

<-h

•

CO

rt

o

b

a)

-vr\

<D

• iH

r-i

-P

<L>

.2

^

a

•H

o

£

o

P-

Ph

<D

Q

o

•

CD

-o

rQ

ON

Average num
th increments.

O

IG, 31.
m dep

Ui

r-t

66

Environmental Parameters Treated Individually

Depth : The samples of Pontoporeia were pooled into 10 m depth
increments,, The shallowest depth interval included all observations
between 9 and 15 m and was designated the 10 m increment. The next
depth interval, including all observations between 16 and 25 m, was the
20 m increment, and so forth. An inspection of Figures 31 and 32 shows
that the maximum density of Pontoporeia occurred around the 30 m On
the average, there were fewer individuals at the 10 and 20 m than at
30 m and the confidence intervals show considerably more variability at
these two depths than at any of the others. There is a sharp decline of
amphipods from 30 to 40 m, then a leveling off from 40 to 50 m with
another sharp decline from 50 to 60 m. They gradually decrease from.
60 to 230 m, beyond which the density is fairly uniform. The correla-
tion coefficient showed this same relationship.

Time: All observations of the three sampling seasons were com-
bined into monthly intervals for Pontoporeia and no sighnificant monthly
variation in abundance was found (Fig* 33)„ Figure 34 likewise indicates
no significant correlation between average number of Pontoporeia per
station stop and day of the year.

Bottom Temperature: The samples of Pontoporeia were pooled
with respect to bottom temperature increments consisting of 1°C inter-
vals. The 1 interval included all observations between 0. 6 and 1.5 C
while the 2 increment contained all observations between 1. 6 and 2. 5 C

67

o

o

o

o

o

o
o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

m

o\ •

o

LO «4-J

o

o

o

r tJ

o

o

o
o

o

o

h m

o

o

o

o

r^

o

o

o

m m

o

o

o

o

o
o

o

o
o

o
o

o
o
o

o

o
o

o
o
o
o
o
o
o
o
o
o
o
o
o

o
o

o
o
o

o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o

o
o
o

o
o
o

o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o

o

o

o

o
o

o
o

o
o
o
o

o
o

o
o

o

o
o

o
o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o o

o

o
o

o
o

o
o

o
o

o
o

o

o

o

o

o

o

o

Q

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o o

o

o

o
o

o

o

o

o o

o

r

I

1

1

1

1

o

00

o
m

o

CM

o

o

o

CO

o

CO

o

•

<u

o

4->

r^.

rt

CM

0)

•i-i

•

4->

o

rQ

<r

o

CM

U

CD
CD

O

s

i-4

CM

CD
U

$

O

CO

00

CO

u

rH

u

CD

0)

A

CO

•rH

#

c

• r-i

O

•H

^

m

m

rH

JC

ctf

u

a,

p

CD

O
CM

o
p.

•

rH

O

CO

4->

u

G

CD

4->

On

CD

#

m B

O

° a

C^

u •*

S *

rO ^

3 <d

o

3 '•**

v£>

rage ]
ersus

.

o> >

o

>

CO

32. A

on visit

•

• »fi

o

O

CO

(uot 3Bimo j sug ax ^oo^ aaBnbs)

sIutjjb Bxaaocloauod: jo saaquin^ aqi jo tosh

68

Geometric Mean Number of Pontoporeia af finis / m^

o •
x <1

o

— i-

-O

- CO

•g

10

§

<> I

o

O
O

O

(uoxq.«uwojstreax V>oa ejnmbg)
w/Fprpjj? ^feaodoq.uoj jo aeqiimfj treejj o*caq.einoex)

T>

0)

r-l

o

o

Oh

M

<D

«4->

<D

a

CD

fH

«J

2

CT

(0

*H

0)

A

&

(0

<D

PI

*H

• H

<D

tf

-P

• r-i

<IH

<D

o

fH

!

O
Ph
O

*
o

ft

V+H

O

u

V

A

a

^ •

a «»

r-l

o> rt

W) >

rt *

U V

<tf 4J

> .5

.3

CO tJ

co S

o

o-B

*r* >s

fri

•o

69

Average Number of Pontoporeia affini s /w

o •
L

o
o

o3

o
o

CV2 <H

o •

• nd

II ON

CM

o
o

o
o

o

o

o
o

o
o

o

o o

o
o

o
o

o

o

o
o

o
o
o

o

o

o o

o
o

c
o
o
o
o

o

o

o o

o o o o

o
o

o
o
o
o

o
o

o

o

o o o

o
o

o
o

o

o

o
o

o o

o
o
o

o L

o

o

o

•

TJ

o

<D

CM

a

C\

u

V
<D

•

U

o

<S

•3-

2

CM

cr

CO

u

0)

CO

u

•rt

Q)

a

^
H

•i-i

<H

aJ

•

o

aJ

o

•r-i

\0

>>

CD

•

T-i

ctf

u

u

Q

o

rt

Ph

(!)

>n

4->

PJ

m

o

O

Pk

>N

o *

n 2

•

o

<D 3

^Q CO

00

Average num
ation visit vei

+■>

• CO

00 rC

O

. (tf

<u

o

h 13

(uoxq.Biauojsu^ai q.oog e^renbg)
^nysxuxjj? iBX8aoclocj.uo c j jo aeqiuxiji eS^aeAy

70

and so forth (Fig. 35). The extremely high value found at 16°C repre-
sents 3 observations from only one station stop. Although a positive
correlation at the 1 percent level was obtained, the scatter diagram
reveals that this is a very weak association (Fig. 36).

Sediment Type : In the statistical analysis, the sediment categories
wei^e ranked from one to five, with sand receiving a rank of 1, sandy
silt a rank of 2, silt 3, layered 4, and hard bottom 5. Figure 37 indicates
that Pontoporeia was most abundant in the sandy silt and was next most
abundant in the sand sediments. Abundance was about the same in silt
and layered sediments. It is not surprising that Pontoporeia , a burrow-
ing amphipod, was least abundant in the hard bottom samples. Figure 38
shows the variability in amphipod density was much greater for the sand
sediments, which probably reflects the fact that these sediments are
found in the shallow inshore regions of the lake that are subjected to
greater environmental extremes.

Percent Organic Carbon in the Sediment: The samples of
Pontoporeia were pooled with respect to tenths of percent organic carbon
in the sediment (Fig. 39) There is a definite increase in average abun-
dance of Pontoporeia from 0. 04 to 0. 2 percent organic carbon. A general
decline in density takes place from 0. 2 to 1.6 percent with a sharp in-
crease occurring at 1.7 percent. Density gradually decreased from
1. 7 to 2. 5 percent, followed by an increase at 2. 7 percent, and a general
decrease when organic carbon exceeded 2. 7 percent. Figure 40 shows

71

Geometric Mean Number of Pcntoporeia affinis /ar

o

o
o

o
10

(uoTcpmiojstnwx q.ooy ejretibg)
w/sxuxjj« Bfeaodo^uOci jo aaqiun])! uree^ opx^aiuoer)

•H
-P

§

<D

bO

C
•H

I

CD
Eh

8

O

03

CD

o

rJO

TJ
CD

rH
O
O

CD

<d

a 1
to

CD

a,

CO

CD

O
P-
O

O

u

CD

?3

CD

B

CD

a

CD

CD

>

<!

?h

W)
CD

in

CO CD
CD

O bo

1-4 (D

72

Average Number of Pontoporeia affinis/m 2

(^

v>

o

v>

o

•

•

•

T-4

cvi

o

CM

X

CM

rH

+
II

•a

CM

o o c o
o

Q
O

O
O

o
o
o

o
o

o
o
o

o
o

o o

o
o

o
o

o o
o

o

o o

o o

o o

o o

o c

o
o o

o
o

o
o

o
o
o
c
o
o

o
o
o

-00

SO

--3-

CM

o

o

o o

o

00

\o

CM

o

o
o

o

(uoxq^iuJtojsuB^j, q.ooa ejretibg)

O

.3

d

4->

^

O

u

a>

<D

T*

4->

(TJ

0)

0)

u

*§

a

bO

• r-4

ttO

<D

■s

•H

$H

a)

•P

a*

o

c

3

fi

cr

to

O

CO

<D

CO

*

^i

<D

a)

bo

to

d)
-d

a

<L>

CD

TJ

U

|

aJ

2

+>

•iH

a)

Ctf

0)

14

fc

*H

<u

1

en

O
P.

o

6

3

§

ft

a

O

m

O

o
o

KM

wQ

rQ

to

a

3

CO

2

u

a

CD

<u

>

b©

4->

rt

• H

(D
>

to

>

<!

o
•f-i

#

4->

vO

(tf

CO

4-»
CO

a

o

HH

tf

h

<D

? w/sxuxjj« BTeaodoqxiod jo asquin^ ©S^^eAy

73

Geometric Mean Number of Pontoporeia af finis / m

o

CM
CM

O

o

II II fi II II

<•-• CM C^ *A

- *n

CM

o

o
o

o

(uoxyswiOjsvcejLi q.oo# e«renbg)
~xa/s*fUTJJ* Bxe.todoq.uoj J° «*©qronN ueejj oxa^etuoer)

u

U

&

CO

u

CD

p.

CD

o
p.
o

*

o

Oh

J3

« "5
2? «
2 £

%%

< 2

•JT

CD
r-H

o

PL

O

74

Average Number of Pontoporeia affinis /m 2

o •

L.

O

o

CM

O O O

O O O

- vr\

vr\ •

• •

I! c*

oooooooo

*-• CM C^^J- tA

ooobooooob

ooooooooooooo

OOOOOOOOOOOOOOOOO -tH

en -P

CM

o

o
o

o

(uofq^uuojsireai ^ooy; aaimbs)

rt

"d

CD

a

• H

ftf

4^

^Q

O

fc

0)

4->

V

B

0)

fH

rt

3

a 1

CO

»H

0)

Ph

CO

•iH

a

• H

MH

*H

Ctf

aJ

•i-i

a>

•

fH

<D

o

ft

PL

J>N

+>

3

3

(1)

H.

S

m .h

n3

. <»>

*J CO

<D

rD W

6 *

2 CO

3 H

fi a>

a> >

ojd 4-»

aJ "r*

< §

• i-i

• tJ

oo <2

a o

H at

b

<tf

75

Geometric Mean Ntuaber of Pontoporeia affinls/ m 2

° t

38

o

o

CM

•8

•H

10

g

d

3

i

i i

11 r= «h

C^k §

+3

§

O

<D
A,

O

O
O

o

(uofq.«iiuojsinMi q.ooa ejrenbs)
g^ysprfjji B-feaodoq-UOj jo aequmti u«en opiqemoeo

rH

0<

fH
0)

CD
U

CO

)H

<U

0<

t>

CO

S

• H

a>

PI
• H

g

m

•iH

m

T*

d

0)

a*

CO

CD

O

4-»

d

a

o

• r-l

3

a

&

<+H

aJ

u

u

(1)

.1-1

,d

P!

a

^

fc

a

d)

"ti

W)

U

m

(U

u

a)

fH

>

cu

0}

O O

76

Average Number of Pontoporeia affinis/m 2

<a
o

CM
CM

O

o

CM

11 c>

CM

o o o o o

o o o
o o o

o o o o

o o o
o o o o o

o o o © o
o © © o

© o o

o o o
o o © o o o

o o o

o © © ©

o o o o o

o o o © ©

o

o o o o
© o o
o o o o

© o o
© o © © © o
oo© ooo©©oo©
oooooooo© ©oooo©©

oooo oooopo

o

- r

o

o

T

o

r\

CM

o

VA

(uoxq.«uu:ojstreaj[ ^003 ejrenbs)

d

•d

V

a

$

4->

.a

^i

CD

4-> •

CD «g

8 aj

*1

Tl

a 1 w

«

O

0) +*

JZ

Oi rt

•P

• r-4

C

CO

Pi

•H

a

S

O

aJ

a)

•P

u

d

ctf

•T-t

O

O

O

• H

u

«

•e

a)

c

PH

bJO

O
U

CD

cm

m a

u

. u

U Q.

,D CO

S

3 n

H

<D >

W) +;

vera
visi

< s

.1-1

. 4->

O W

^ CO

►H nt

h

a)

gttt/sfufjj^ i : feaoaoqSoJ jo aeqiunfj eSsaeAy

77

the wide range of variability that exists in the average number of
Pontoporeia found in the sediments containing little organic carbon.

Distance from Shore : Samples of Pontoporeia were combined into
groups that were the same distance from shore (Fig. 41), Stations two
miles from shore represented the most inshore sampling area while a
station 38 miles from shore was the greatest sampling distance. There
appears to be a general inverse relationship between amphipod abundance
and distance from shore. The greatest variability in amphipod abundance
occurred at the more inshore stations (Fig. 42)„
Stepwise Multiple Regression Analysis of Envir onmental Parameters

The lake was first treated as a single system in the examination
of the interrelationships of Pontoporeia and its environment. It was
next divided into four depth zones: 10-35 m, 36-65 m, 66-105 m and
greater than 105 m. Stations in the 10-35 m zone were considered the
sublittoral and those deeper than 35 m, profundal.

First the simple correlation coefficients were calculated for the
four depth zones and all depths combined, then the stepwise multiple
regression equation was computed for these data. A 5 percent level of
significance was chosen for the F test that controls the likelihood of
committing the error of entering a variable into the prediction equation
when it is insignificant. This same level of significance was also used
for the F test that controls the error of removing a variable from the
equation when it is significanto A multiple correlation coefficient and a

78

Geometric Mean Number of Pontoporeia affinis/ m 2

(^

\f\

O >A

o

•

• •

tH

CM

O CM

X

cm

i

t-4

f |

p0

■a

•

<u

o

-**

ft

1

u

I

Q)

1

4->

J

0)

a

1

<D

1

?h

•
o

3

a 1

CO

"8

L

CQ

Oh

CQ

•

c

^v

0)

•r-l

^

•H

1 f>4

iH

Ci

o

1

' ^-.

£

• r-l

,3

CO

O

N.

rt

a

W

^V

• H

o

c

>v

<D

(D

Fh

o

| ^> .

•

J*

*H

<+H

•H

' — ■** "

o

O

o

•8

"J

-CM

£

ft

s

£

/

g

ts

3

n

o

4->

J

(0
• r-l

<U

• ^y •

«H "O

1 rt*"| '

O to

§

S 3

\

•

« ,3

S*-"^^

b0 8

w ^ I 3* M

d ^

» ^^-t

► °

1 J' •

< "S

1 ^*s^

o

>v

• W

* ** ^

-h r3

* £

• u

1 1

1 J ■""■ J

£ 5

•
O

• o
o *n

SI

>T\

o

T-*

T-1

(uoxq.wLiojsuBJij, V>o# ea-enbg)

J3tt/spLrfJJ« Bf©aodoq.uo c i jo aeqimu u^ejj opi^euioeg

79

Average Number of Pontoporeia affinis/ m^

o

X

•
CM

o

•

o

•

CM

1 .

o

o o o o
o o o

• •

1 TJ

U -3-

o o

o o o o

o o
o o
o o o

o o
o c o

o o o o

o
o

o

o
o

o o

o o
o o

o o o o o

o o
o o

o

o

o

o o
o o

o o o o
o

o

o

o

o

o

c
o

o o
o o

o
o

o

o o

o

o

o
o o

o o o
o o

o

o

o o

o

o

o o

o
o o

o

o

o
o

o o
o

o
o

o
o

o o
o

o o o

o o o o o

o

o o

1

1

1

-2

O

q

CM

o

o
o

o

Tl

0)

a

• H

«J

^

&

U

o

+*

CD

a

CD

U

rt

2

10

cr

^

CO

H

n

u

CD

G

a.

•H

CO

•

.1-1

CD

<D

• r-1

M

o

^H

O

S

CO

1

at

• 1-4

u

O

§

P.
O

*

CD
a

•P

o

ai

CO

n

ft

O

+->

CO
•r-i

n 2

u 5

^ u

8 a)

2 >

a ^

<D (Q

bX) .h

aJ >

< %

4->

• CO

. aJ

O

<D

fc s

80

coefficient of determination, in addition to the multiple regression
equation, were also calculated for the depth zones.

The multiple correlation coefficient measures the closeness with
which the multiple regression equation fits the observed points, L e. the
multiple correlation coefficient measures the combined effect of the
environmental parameters on the abundance of this amphipod« The coef-
ficient of determination, -the square of the multiple correlation coeffi-
cient, shows how much of the variability in the abundance of these organ-
isms is explained by the environmental factor s.

Table 4 shows the correlation coefficient of Pontoporeia and the
environmental factors for the four depth zones and all depths combined.
Day of year is significant only at the 10-35 m zone Depth and depth
squared show a significant positive relationship for the 10-35 m zone
while the remaining depth zones show a significant negative relationship.
When the lake is treated as a single system, there is a very significant
negative relationship between depth and amphipod abundance.

There does not appear to be a significant relationship between
amphipods and surface or bottom temperature for any of the four depth
zones. However, if the lake is treated as a unit, there is a slightly
significant positive relationship for both water temperatures.

There is no significant relationship in the sediment types, for 10-
35 m, a slightly significant negative relationship from 36-105 m, a
slightly positive relationship at depths greater than 105 m and a very

81

r d

a

cxJ

r-l

U

.|H

03

CO

>s

•r-i

CO

>s
r-H

(!)

rt

r d

r-i

•+->

cti

T>

a

a

rt

• r-l
4->

ro

d

• r-«

i-H

CI

(U

•|H

u

MH

at

u
o

u

•iH

^

0)

rQ

o

nj

PL

o

«rH

a

o

rH

11.

CD

4->

a

CD

<D

T3

CI)

4->

CO

CD

+->

bJD

ro

•r-i

•|H

r£

X

O

(\)

•r-l

4->

s

CD

rt

(/J
Oh

J

• r-i

o

CO

fl

CO

O

• r-l

3

+-»

0)

a

cu

a

r-l

<d

.r-l
>

H

Pi

CD

,

r-H

Tt*

Ctf

O

W

• i-(

J

OJU

n

ID

i— i

<J

o
••h

IH

^

CD H

Q o

u

m
o

in
o

r-H

PI
o

2 a

<D ^

Q si

00

CO
I

o

•r-l

*

-X-

•X-

■*

tt

X

-5f

x

-X-

-X-

*

•X-

-X-

*

*

X

*

-X-

*

4C-

I>

LO

CO

CM

l>

CO

m

^

o

^

co

CM

t-H

O

O

l>

m

o

m

o

^

^

O

vO

LO

r-H

(M

^

^

^

m

fs.

^

I

J

1

'

;

i

i

-X-

•X-

*

->£

4C-

-X-

#

-X-

»)f

-X-

X

•5(-

■»■

o

o

CO

vO

i-H

00

CO

r^

m

m

v£)

r-H

vO

vD

r-i

m

00

^

o

o

o

O

r-H
t

1

i

O

o

CM

m

o

vO

vO

CO

-X-

■X-

-X

-)f

■&

■3f

-X-

*

#

«5(-

-X-

-X-

X

■3f

-)f

o

sO

r-H

t^

^

00

CM

m

r-H

I>

o

r-H

o

O

vO

r-H

m

in

o

CO

00

o

H
i

m

o
1

O

CM

i

CM

CM

i

-X-

m

■5^

CM

&

*

-Xr

-X-

•5?-

-X-

o

o

m

00

i-H

r-

r-H

vO

o

vO

vD

I— J

o

o

CO

rH

CM

m

O

CO

in

CM

r-H

i-H

r-H
1

o

O

CM

!

o

I

CM

m

CM

*

-X-

-X-

-X-

*

-X-

-x-

*

*

-X-

•5C-

X

-X-

#

.st* .

t^

o

vO

o

r-

CO

CO

m

o

vO

r-H

LO

m

m

r-H

o

CM

O

CO

m

t^-

(M

LO

m

o

CD

u
2

o

Jh

i-H

1

CM

CO

+->
CD

a

•rH

CD

CO

CD

CO

CD
U
O

CO

m

CM

4->

•rH

r&

ctf

ccJ

CO

u

u

fl

rH

0)
Jh

0)

Oi

a

CD

CD

U

CD

CD
■+->

rO

u

a

o
w

MH

CD*

CD

•-d

•iH

CD

mh
O

3

cr

CO

CD

a

o

4->

a

CD

a

CD
U
U

CD
U
CI

u

.r-J

• iH

rH

CD

a

o
d
o

>N

&

<+H

4J
4-J

• rH

CO

bJO

•r-H

rt

,£

Jh

•rH

a

CD

CD

£

-

CD

CD

•iH

r-H

Ph

r^

Q

Q

Q

W

PQ

W

0.

Q

o

o

u

r-H

r— i

cm

CO

^

in

vD

c^-

00

o

r-H

r-H

X

X

XI

X

X

X

X

X

X

X

X

On
CM

CO

CM

O

in
o

CM

^

^

m

r-H

4->

+->

«J

aJ

+^

+->

a

fl

aj

rt

a

o

•rH

.iH

MH

MH

•fH

•rH

fl

fl

bJO

bJO

• r-i

•r-i

CO

CO

-)C-

-X-

-x»

82

significant inverse relationship when all stations were combined.

Oligochaetes, sphaeriids, and chironomids showed a significant
positive association with Pontoporeia for the four depth zones and when
all depths were combined.

Tables 5 and 6 show the multiple regression equations for the four
depth zones and all depths combined. These equations contain only the
environmental factors that contributed a "significant 1 ' amount of inform-
ation about changes in the abundance of Pontoporeia . In addition to the
regression equation, a table was constructed to show the relative pro-
portion that each significant variable contributes to the multiple corre-
lation coefficient and the coefficient of determination.

Normally, when multiple regression is utilized in the analysis of
biological systems, considerable "noise" is caused by the incomplete
measurement of all possible variables or the inability to measure the
variables with precision. This excessive noise results in a small
multiple correlation coefficient. The large multiple correlation coeffi-
cients of Tables 5 and 6 indicated that the physical and biological factors
used in this study contributed considerable information about changes in
Pontoporeia abundance.

Table 5 shows that the abundance of sphaeriids statistically con-
tributed the most in explaining the fluctuation of amphipod density at the
10-35 m zone. The other significant parameters, in their order of
importance, were depth, depth squared, surface temperature, sediment

83

TABLE 5. The proportion of the multiple correlation coefficient (R) and
the coefficient of determination, expressed as a percentage, (R^) that
each statistically significant variable contributes to the multiple regres-
sion equation of the four depth zones in Lake Michigan. The dependent
variable of the multiple regression equation (Y) is represented by the
numbers of Pontoporeia affinis /m^

Variable

R

R 2

10-

•35 m

X 10

Sphaeriidae

.559

31.3

x 2

Depth

.084

10. 1

x s

Depth squared

. 039

5. 1

x 4

Surface temperature

. 031

4. 5

x 6

Sediment type

. 028

3.9

X 8

Distance from shore

.015
. 756

2.2
57. 1

Y =

. 756 + 7. 927X 2 - . 122X

3+1.

485X 4 -

7. 106X 6 - 4. 074X 8 + 3.431Xjq

36-

■65 m.

X 10

Sphaeriidae

.556

30. 9

X 6

Sediment type

. 043
.599

5.
35.9

Y =

48. 104 - 3.804X 6 + 2. 757X 1Q

66-

• 105 m

L

x 10

Sphaeriidae

. 586

34.4

X 3

Depth squared

. 033

5.2

X ll

Chironomidae

. 034

3.0

x 9

Oligochaeta

. 024

3.6

677 45.6

Y = 39- 592 - . OO9X3 + 1. 032X„ + 1. 766X 1Q + . 926X U

> 105 m

X 2 Depth . 769 59. 1

X9 Oliogochaeta . 037 5. 9

Xiq Sphaeriidae . 020 3. 3

.826 68.3

Y = 49. 216 - . 133X 2 + 1. 371X 9 + 1. 732X 1Q

84

TABLE 6. The proportion of the multiple correlation coefficient (R)
and the coefficient of determination, expressed as a percentage, (R )
that each statistically significant variable contributes to the multiple
regression equation of all depths combined in Lake Michigan. The
dependent variable of the multiple regression equation (Y) is
represented by the numbers of Pontoporeia affinis/ m .

Variable R R 2

.744 53.3

.018 2.8

.008 1.2

.007 1.2

.007 1.0

.005 .7

.005 .8

.002 .4

X 10

Sphaeriidae

x 3

Depth squared

x ll

Chironomidae

X 2

Depth

x 7

Distance from shore

X l

Day of year

X 6

Sediment type

X 5

Bottom temperature

,796 "63.4

Y = 12.333 + ,049X 1 + . 360X 2 - 12. 174X 3 + . 779X 5 - 4. 193X 6

- 3.476X ? + 2.874X + 2.355X U

type, and distance from shore. Approximately 57 percent of the
variability in amphipod abundance was attributable to these six variables

Only two parameters, sphaeriids and sediment type, were statis-
tically significant in explaining the deviations in amphipod density of the
36-65 m depth zone. The sphaeriids accounted for 31 percent of the
changes in Pontoporeia abundance while the sediment type contributed
only 5 percent.

Within the 66-105 m depth zone the abundance of sphaeriids sta-
tistically represented the most important parameter, followed by depth

85

squared, chironomids, and oligochaetes in that order of importance.
These four variables accounted for approximately 46 percent of the
fluctuation of Pontoporeia abundance.

At depths greater than 105 m, depth became the most important
factor in determining the variability in amphipod abundance followed by
ologochaetes and sphaeriids, in that order. These three parameters
accounted for 68 percent of the fluctuations in amphipod density.

When the lake was treated as a single unit, additional parameters
became important in determining the fluctuations of Pontoporeia density*
The sphaeriids were the most important, followed by depth squared,
chironomids, depth, distance from shore, day of the year, sediment
type, and bottom temperature. These eight variables accounted for
approximately 63 percent of the changes in the amphipod counts. How-
ever, the last seven of these eight parameters contributed only eight
percent of the coefficient of determination.

When the lake was subdivided into four depth zones, three environ-
mental parameters, day of year, percent carbon in the sediment, and
bottom temperature had no influencing effects on the amphipod popula-
tions. When the lake was treated as a single system, surface temper-
ature, percent carbon in the sediments and oligochaetes did not signif-
icantly affect amphipod abundance. Multiple regression analysis indi-
cated that the density of these organisms is not influenced by the percent
carbon in the sediment.

86

It appeared that turbulence caused by waves might contribute
information about changes of Pontoporeia abundance in the lake. In-
formation collected at the stations of the C, D, and E transects and wave
heights measured by the personnel of the three car ferries were utilized
in the subsequent statistical analysis. Wave heights were coded from
to 5 with a rank of one representing a wave one meter high, a rank of
two a wave two meters high, and so forth. The largest wave height
noted for a period three days prior to and including the sampling day was
recorded for the stations of each transect* The lake was treated as a
single system and then divided into two depth zones: 20-50 m and less
than 105 m. The shallowest station of the three transects was 20 m.

Table 7 shows the simple correlation coefficients of amphipod
abundance versus the environmental and biological parameters. There
does not appear to be any significant relationship between Pontoporeia
and wave height for the 20-50 m zone., less that 105 m, and all depths
combined. The multiple regression equations for these data also indi-
cated that wave height is not associated with Pontoporeia abundance .
Unfortunately, the turbulence caused by waves would achieve its grestest
effects at depths less than 20 m, which would indicate that these results
are inconclusive.

SEASONAL DISTRIBUTION

In the investigation of the seasonal abundance patterns of Pontoporeia ,
the fundamental question concerning the reliability of data collected at

87

TABLE 7. The relationships that exist between Pontoporeia affinis and
the physical and biological environment as determined by correlation
analysis for the stations of the C, D, and E transects.

Variable

20-50m

< 105 m

All
Depths

x l

Wave height

.038

.019

. 017

X 2

Day of year

. 167

. 023

-.021

x 3

Depth

-. 125

-. 617**

-.823**

x 4

Depth squared

'-. 146

-. 614**

-.748**

X 5

Surface temperature

. 134

. 013

.080

X 6

Bottom temperature

. 181

. 342**

,476**

x 7

Sediment type

.049

-.483**

-. 610**

x 8

Percent carbon in sediment

-.087

-.336**

-. 673**

x 9

Distance from shore

-. 146

-. 622**

-. 618**

X 10

Oligochaeta

.258*

. 472**

.701**

X ll

Sphaeriidae

, 320**

. 630**

.829**

X 12

Chironomidae

. 333**

. 373**

. 586**

* Significant at 5%
** Significant at 1%

70 d.f.

136 d.f.

267 d.f.

sequential time intervals was examined by sampling the stations of the C
transect on two successive days, 9 and 10 August 1966 (Fig, 43). It was
felt that the density of Pontoporeia at these stations should not radically
change within a 24 hour period, and any significant deviation in amphipod
density between the two collecting dates would be the result of experi-

88

mental error or inefficient sampling techniques,. The Wilcoxon sign
rank test indicated there was no significant difference in the abundance
of amphipods between the stations of the two sampling dates.

The result of the rank test justified the combining of the two data
sets for each station to obtain an estimate of sample variation. The co-
efficient of variation, which measures the relative variation of amphipod
abundance, was calculated for these combined data (Fig. 43), There is
no particular pattern to the relative variation with respect to depth or
position along the transect. The average relative variation for the sev-
en combined stations was 9. 3 percent.

The amphipod counts for all 35 stations of the long-term study
area were pooled for each month of the three sampling seasons (Fig. 44).
A Kruskal-Wallis sign rank test indicated there was no significant dif-
ference between the average standing crop of the three sampling seasons.
Table 8 shows the grand geometric mean number of amphipods of all
depths combined for each of the sampling seasons and the results of the
Kruskal-Wallis test. The average number of amphipods for 1964 is 53
while the means for 1965 and 1966 are 55 and 54 respectively.

The Pontoporeia collected at the 35 stations were also combined
into four depth zones: 10-35 m, 36-65 m, 66-105 m, and greater than
105 m for the monthly sampling periods (Fig. 44). A Kruskal-Wallis
test was calculated to determine if the average density of amphipods
was significantly different for the four depth zones. A significant differ-

89

Coefficient of Variation

CO VO ^ CM O CO -VO

r-i tH t>4 rH tH

-J -J 1 1— 1 1 L.

o

•H

-P

*.

1

>

*

\o \o

•P

T-f *H

s

<*S

£

18 18

*

6) &

s

33

1

o

o
o

o

a

t3

8? G

^ s

<u

6

density of the s
10 August 1966,
ing dates.

VO

6

■— i

6

CO

o «

+>

a*

o

CD

01

§

U

p-

u

co O

H

■s

o

+r o>

o

Ph

o ■&

C\

" CD +■>

6

<U co u

* S o

CM

6

•mparison of
d on the C tr
6f variation

r-i

9 a) 4-i

6

G. 43. A c
ations locat
e coefficien

CO 4^

(uox^uuojsurej:! q.oon ejmbg)

Jt/BjufjJ* *\&xodo%uQ& jo jeqwnji OTe W opiq.ercoer)

90

GEOMETRIC MEAN NUMBER OF PONTOPOREIA AFFINiS /M 2

(NOIlVHdOdSNVdi lOOd BdVflOS)

z

CO

O

d

• r-l

■

m

(/>

<+H

at

<

rt

• H

~5

to

CO

<7)

■■5

o

2

Ph

o

•y

<

d

t>

2

o

CD

d

Li.

~3

° s

?( o

Q

8 °

Z

aJ to

O

3 &

i/>

<

->

in

CD

onal
d all

->

0)

co pj

2

<

(0 w

2

*+-■ d

L_

CO N

- )

Q

Z

3*

O

CD

o>

.

i/)

« ^

<

G. 4
r the

o

2 JA|/SINtddV VIHdOdOlNOd dO d38HHN NV3IAI DldlBHOBO

91

ence in abundance was found among the four depth zones; the density
of numbers was greatest in the 36-65 m zone followed, in order of de-
creasing abundance, by the 10-35 m, 65-105 m, and greater than 105 m
zone.

Pooled amphipod counts of the four depth zones were again treated
with the Kruskal-Wallis rank test to determine if the standing crop of
amphipods were significantly different for the three sampling periods.
This test showed no apparent difference in the standing crop for the
10-35 m, 36-65 m, and greater than 105 m depth zones. The grand
geometric mean number of Pontoporeia for the four depth zones of the
three sampling seasons and the results of the Kruskal-Wallis test are
presented in Table 8.

A reexamination of Figure 44 and Table 8 shows that although
there was no significant seasonal variation in the abundance of amphipods
at the 10-35 m zone there was considerable variation within comparable
sampling periods. In 1965 there was a general increase in the popula-

TABLE 8. The grand geometric mean number of Pontoporeia affinis
for the three sampling seasons. H represents the results of the
Kruskal-Wallis test.

Depth Zones 1964 1965 1966 H

10-35 m

62

55

67

2.78

36-65 m

78

75

75

.52

66-105 m

50

59

58

7. 76*

>105 m

35

40

36

3.81

All stations combined

53

55

54

.55

* Significant at 5%

92

tion from April to August followed by a slight decline from August to
November. The 1966 population decreased in numbers from March to
April, increased from April to August, and then decreased sharply
from August to November,

A Spearman rank correlation coefficient was computed for the
four depth zones and for all depths combined to determine if any asso-
ciation patterns of Pontoporeia abundance existed among the three
sampling seasons (Table 9). The results indicated no correlation in
the seasonal patterns of abundance for any of the four depth zones •
Only the 10-35 and 36-65 m zones showed any association when all the
stations were combined.

Figure 45 shows the patterns of seasonal abundance of Pontoporeia
for the seven stations of the C transect. The Kruskal-Wallis test
indicated that only at station C-3 was the average standing crop sign-
ificantly different for the three sampling seasons; there appeared to be
no seasonal differences at the other C stations . The grand mean number
of amphipods for these 7 stations for the three sampling seasons and
the results of the Kruskal-Wallis test are presented in Table 1CL

Figure 45 indicates that the patterns of amphipod abundance for
1965 and 1966 are quite similar for stations C-Z, C-3, C-6 S and C-7,
and changes in amphipod counts of stations C-4 and C-5 were similar
in many respects. The seasonal changes at station. C-l, however,
differ strongly from the other six stations.

93

GEOMETRIC MEAN NUMBER OF PONTOPOREIA AFRN1S / M*

U

UJ

c/>

z
<

u

W

I

li-

O

I/)

z
o

<

ro in
O c\I

o
d

in

CVJ
L_

q

d

in

n t ^ S
i i i i

u u u o

z
o

to
<

~> CD

CD

-, 0)

<

Ll
— )
Q

Z

8

O
O

o

in

o
o

o
in

O *

. <D

O

O

0\

O

ti

o

m
O

<D
U
Pi
ai

o

bunc
ect.

<

CU (Q

r-H Pi

-> in

CD

a) rt

0)

o +->

~> v-

s^

2

S «

W rS

<

m *»

<+H

z

~ o

CO

U_

PJ w

H G

— >

V

4-> 4->

aJ ai
Oh 12

. PS

>

<D
• CO

o

S3

in

(NOIlVhidOdSNVdl iOOd 3d.vnos)

,H/$lNlddV VI3U0d01N0d dO dBQlAinN NV3IAI 01813^039

94

A Spearman rank correlation test was calculated for stations C-2
through C-7 in order to determine if there was a significant association
in the patterns of Pontoporeia abundance (Table 11). The results show
that these stations have similar patterns of seasonal variation, partic-
ularly when adjacent stations are considerexlo

The seasonal patterns of amphipod abundance for the seven
stations of the C transect were next compared with the abundance pat-
terns of their respective depth zones by the Spearman rank correlation
(Table 12). Although only stations C-2 and C-4 showed any significant
association with their respective depth zones, all the correlation co-
efficients were positive, which would indicate that, in general, the
seasonal fluctuations of Pontoporeia of the seven C stations were sim-
ilar to the fluctuations of their respective depth zones.

TABLE 9» Spearman rank correlation coefficients indicating the degree
of association that occurred among the patterns of seasonal abundance
of Pontoporeia affinis for the four depth zones and all depths combined*

10-35 m

36-65 m

66-105 m

>105 m

36-65 m
66-105 m
> 105 m

085

204

.400

188

421

429

All stations
combined

529

459-

119

.415

* Significant at 5%

95

TABLE 10. The grand geometric mean number of Pontoporeia affinis
for the three sampling seasons of the seven stations of the C transect.
H represents the results of the Kruskal-Wallis test.

Station

Average
Depth m

1964

1965

1966

H

C-l

20

77

72

99

3.20

C-2

50

80

85

84

. 12

C-3

77

56

67

63

13.47**

C-4

108

46

51

49

.62

C-5

157

39

47

43

1.95

C-6

99

49

^58

49

2.63

C-7

55

65

72

70

. 73

** Significant at 1%

TABLE 11. Spearman rank correlation coefficients indicating the degree
of association that occurred among the patterns of seasonal abundance
of Pontoporeia affinis for stations C-2 thru C-7 for the three sampling
seasons.

C-2

C-3

C-4

C-5

C-6

C-3

C-4
C-5
C-6
C-7

648**

566*

258

637**

783**

578*

.297

350

594*

737**

765^

658**

721**

495

709**

* Significant at 5%
** Significant at 1%

96

TABLE 12. Spearman rank correlation coefficients indicating the degree
of association that occurred between the patterns of seasonal abundance
of Pontoporeia affinis for the stations of the C transect and then respec-
tive depth zones.

C-l

versus

10-35m

C-2

versus

36-65m

C-3

versus

65-105m

C-4

versus

> 105m

C-5

versus

> 105m

C-6

versus

66-105m

C-7

versus

36-65m

.422
.467*

.405
.. 515*
.432
. 175
.266

* Significant at 5%

PATTERNS OF SPATIAL ABUNDANCE

All observations from each station of the five cross -lake transects
were pooled for the three sampling periods to provide information on the
spatial standing crop of Pontoporeia (Figo 46), These results indicated
that its abundance at shallow inshore stations A-l and A- 6 was lower
than at other stations on the A transect., which were located at greater
depths and farther from shore. The remaining four transects show a
general tendency for average amphipod counts at the stations located in
the deeper portions of the lake to be less than the shallower inshore
stations. The standard deviations of amphipod density of the inshore
stations are generally greater than those of the deeper stations. Stations

97

O

u

Id

Z

2

O

m

o
o

o
in

«S

.02

o
in

o
m

8[5

-J

,o2

Si3

o2

o
m

m

2

0)

,3

1/)

4->

CD

z
C
u.

CO

<

CO

-M

<

a

CO

£

CD

n

•H

4->

*

e

4->

CD

z
O
n

I

CO

u.
O
q:

5

CD

4^

a

lj

O

Jh

^

2

D
Z

Li.
l/>
Z
<

CO

CD

Z

h-

•1-1

fH

<

O

2

7

O

^

CO

u

q:

1-

Ltl

O

DC
Ui

cr

SI

CD

CD

a

CO

w

M

• H

to

P-

CD
U

1-

O

3

z
u

LJ

CO

CL
LJ

b

a

t/)

O

Jh

X

<

^

<-H

>-

CO

PJ

CD

CD

CD
>

CD

H

CD
O

a

4^
CO
•r-4

TJ
0)

•
CO
+->
O

CD
CO

•

r{

•

vO

4->

CD

^

■J3

•

r-H

O

n

CD

>
•r-l

CD
4^

98

situated in the northern basin of the lake have, in general, more amphi-

2

pods/m than stations located in the southern basin that are situated at

comparable depths.

It is difficult to compare directly the amphipod counts at the 35
stations because they were located at different depths and it has been
shown that depth strongly influences the abundance of P. af finis in an in-
verse manner. In order to make direct comparison possible., the effect
of depth on amphipod density was removed by first calculating the density-
depth regression equation Y « 81 04 - 0. 232X where Y represents the
average number of amphipods per station stop, X the depth per station
stop, 81. 04 the intercept, and -0.232 the regression coefficient. A
second order polynomial relationship for depth was important in explain-
ing fluctuations in amphipod abundance when all environmental parameters
were considered. However, when the relationship between amphipod
density and depth was considered independently of the other environ-
mental factors, the squared term for depth was not significant.

The grand average density of amphipods and the average depth
must fall on the regression line. In removing the effect of depths the
deviations of Y about the regression line were shifted so these deviations
occurred about a regression line that had a slope of zero and an inter-
cept equal to the grand average density of amphipods. This was accom-
plished by modifying the above regression equation in the following
fashion: Y = y + 0. 232(X-85 9) where y was the observed average value

99

of amphipods per station stop, X the depth of the station stop, 0. 232 the
regression slope, 85, 9 the average depth, and Y the estimated density
of amphipod with the effect of depth removed.

Figure 47 shows that the pooled average density of Pontoporeia at
stations A-l and A- 6 is much lower than the remaining stations of the A
transect after the effect of depth was removed. The average abundance
of amphipods found at shallow station B-8 also appears to be lower than
the remaining stations of the B transect. The density of amphipods at
station C-2 appears to be somewhat higher than the other stations of the
C transect, and the density at stations D-6 and E-6 also appears to be
higher than the other stations on their respective transects.

100

o
u

z
<

to

in ->
I

^C

tn

to

Old

in _j

o
in

8

6

in

tj

^-^ — " " '

u

.Sm

V)

1

z

<

o:

H

^-

<

v ,

X,

o
tn

a

-OS

U
u

z
<

\ .

A,

/

D

/

, V

o2

-J

o2

o

in

o
m

o
o

z <->

to

u.

UJ U_

q: uj

EL

UJ LU

to

x t
< £

W

DISCUSSION

REPRODUCTION

The results of the present investigation indicate that at least three
distinct breeding populations of Pontoporeia occur in Lake Michigan,,
The data strongly suggest that amphipods living at a depth of about 10m
mature in one year, while those in approximately the 20- 35m zone re-
quire two years to mature. The reproductive season seems to be com-
pleted by late May-early June for these populations, Amphipods living
at depths greater than 35m may require at least three years to mature,
and the population there breeds intermittently throughout the spring,
summer, and fall. Insufficient data were obtained to determine whether
breeding also takes place in winter.

Juday and Birge (1927) found that the breeding season of Ponto-
poreia extended from December to May in Green Lake, Wis. , and there
the population required two years to mature. Bousfield (1958) reported
ovigerous females from November to April in Canadian oligotrophic
lakes, and indicated that the life span was slightly more than two years,
Larkin (1948) found that this amphipod had a life span of 2 and perhaps
3 yea^s in Great Slave Lake. Teter (I960) felt that P. af finis matured
in one year in Lake Huron while Cooper (1962) found a one year life
cycle at a depth of 12 m and a two year life cycle at a depth of 35 m in

101

102

South Bay, Lake Huron, Green (1965) found a maturation period of 2
years in Cayuga Lake, New York, and SegerstrSle (1950) reported that
P. af finis matured in one year at a depth of 3 m and required 3 years to
mature at a depth of 35 m in the Baltic Sea, Summer breeding., which
appears to be common in North American oligotrophic lakes, was first
recorded in the Baltic Sea in 1967 by SegerstrSle and represents the only
record of summer breeding in Eurasia.

Observed variations in the duration of the breeding season and
development are likely attributable to corresponding variations in water
temperature. In Lake Michigan, the average bottom temperature of the
10-15 m depth zone, from March to October, is approximately 9 C„ The
average bottom temperature at 20-35 and 36-86 m depth zones for the
same time interval is approximately 7°C and 5 C respectively. Depths
beyond 86 m have an average bottom temperature that is slightly less
than 4 C from March to October* An examination of bottom temperature
records of the present study indicated that Pontoporeia did not breed at
any depth when the bottom temperature exceeded 7 C, Samter and
Weltner (1904) observed that temperatures greater than 7 C prevented
egg production and concluded that the type and locality of a lake deter-
mined to a large extent the time and duration of the reproductive periodo

PATTERNS OF SPATIAL DISTRIBUTION
Short- Terra Study Area

Living organisms are subject to a multitude of variable factor s,

103

both intrinsic and extrinsic, that affect their behavior, performance,
and survivalo The effects of these influencing factors are not independ-
ent but exhibit numerous complex interactions of which we know little.
Within any defined geographical area, some factors will remain constant
for a short time while others may vary a In general, the smaller the
area the greater will be the number of constant factors. In the short-
term investigation a small study area was sampled rapidly in order to
hold constant as many environmental factors as possible.

Populations are either randomly or nonrandomly distributed depend-
ing on the effects of the environmental factors. There are two categories
of departure from randomness: a contagious distribution where the indi-
viduals of the population tend to be aggregated, and a regular dispersal
where the individuals of the population tend to be uniformly space* It is
generally observed that practically all biological populations are conta-
giously dispersed and that few populations are randomly spaced.,

Several theoretical distributions have been proposed to explain
these dispersal patterns and apparently some of these conform to empir-
ical data. In the short-term study area two such theoretical distributions
were considered, the normal and the Poisson* If members of a popula-
tion conform to a Poisson distribution then the density of individuals is
low relative to the possible density that cound exist in that area, and the
members of the population are considered to be randomly spaced. If the
effects of all environmental factors are small or the environmental fac-

104

tors are randomly distributed, the individuals will be dispersed as a
normal distribution. This distribution can also be considered the result
of one or more organisms within a sample unit influencing the probability
of other organisms occurring in the same sample*

In this investigation of the distributional patterns it was necessary
to examine the number of individuals in several quadrats of constant size.
However, there is a major disadvantage in quadrat analysis in that changes
in frequency distribution occur when the quadrat size is altered. When
the quadrat size is increased it is more likely to contain many individuals
and less likely to contain no individuals. Preston (1948) has indicated
that in order for the Poisson distribution to be valid the quadrat must con-
tain more space than is actually utilized by the organisms „ As the quad-
rat size is increased, the environmental factors become more variable
and it becomes more difficult to discriminate between patterns of micro-
dispersion, L e within a large quadrat highly populated areas are pooled
with sparsely populated areas. The sampling area of the hand coring
device in the short-term study area contained more space than was uti-
lized by the macrobenthic invertebrates but was sufficiently small to
measure adequately the microdistribution of the organisms.

Replicated random sampling was selected as one method of data
collection because it not only obtains an estimate of the mean but also
an estimate of the variability of the mean. Because the mean is independ-
ent of the variance in the normal distribution and equal to the variance

105

in the Poisson distribution, the variance-mean regression is a useful
tool in determining the distributional patterns. It is apparent from the
short-term study area that small and total Pontoporeia conformed to the
normal distribution while large Pontoporeia fitted the Poisson distribution,,
In spite of the existance of only 15 bivariate observations and 13 degrees
of freedom, the trends of these variance-mean regressions strongly
indicated these two distributional patterns.

Systematic sampling was chosen as the other method of data
collection because it provided an easier method of obtaining underwater
samples. However, systematic sampling has two potential disadvantages.
If the populations or environmental factors contain a periodic variation,
and if intervals between successive systematic samples happen to corre-
spond with this variation, the resulting mean may be seriously biased.
Further, with systematic sampling there is no reliable method of esti-
mating the standard error of the sampling mean,

Six intersecting transects were used to provide information about
the linear distribution of the organisms and to sample extensively the
inner region of the sampling area. It appeared that none of the amphipod
categories showed any interaction between adjacent sampling sites.
Thus, each sample appeared to be independent of the others.

Since 45 samples were randomly chosen and the 43 systematic
samples appeared to be independent of each other, they were pooled in

106

constructing the frequency histograms. The chi square test for good-
ness of fit comparing the theoretical distributions with the observed
frequency histograms showed that small and total Pontoporeia followed
the normal distribution while large Pontoporeia conformed to the Poisson,

Henson (1954) showed in a study of Cayuga Lake that Pontoporeia
larger than 6 mm in length were randomly distributed, those 4 to 5 mm
in length were distributed on the "borderline" between randomness and
contagiousness* and those smaller than 3 mm in length were contagiously
distributedo Green (1965), also in an investigation of Cayuga Lake,
found that adult Pontoporeia were randomly distirubted while the 1-2 mm
size group were strongly clumped.

The investigations of Henson and Green and the present study all
indicate that larger Pontoporeia are randomly distributed within the sed-
iments. It was difficult to compare the distributional patterns of the
smaller amphipods from the three studies because Henson utilized two
small-size categories. Both Henson and Green used the variance-mean
relationship to determine if the size groups were rendomly distributed.
The smallest size group of the short-terra investigation of the present
work was tested in the same manner and also found to be contagiously
distributed.
Long- Term Study Area

Sampling variability (heterogeneity of error) may be classified as
irregular and regular. Irregular error is characterized by certain

107

samples possessing considerably more variability than others with no
apparent relation between means and variances, Regular heterogeneity
usually arises from some type of nonnormality in the data with the vari-
ability of the samples being related to the sample means (Steel and
Torrie I960).

It is often difficult to interpret the causes of variance-mean inter-
action but the variance of either the Poisson or binomial distribution is
functionally related to the mean. Evans (1952) pointed out that if density
is defined as the number of individuals per quadrat, density becomes
directly proportional to quadrat size with a doubling in value occurring
for each two-fold increase of quadrat area. The larger the quadrat the
more likely it will contain many individuals and less likely to contain
fewer individuals. A population sampled with one quadrat size may have
its members distributed at random, yet the same individuals may be con-
sindered highly aggregated when another quadrat size is usedo The
Pontoporeia sampled in the long-term study area were converted to
individuals kx& because the two samplers used had different grab areas.
If samples are multiplied by a constant, the mean of these samples is
increased by the multiple of the constant while the variance is increased
by the square of the constant, which can affect the variance-mean rela-
tionship. It was indicated earlier that different size groups of Poptoporeia
were either contagiously or randomly distributed. The relative propor-
tion of these size groups will affect the distributional patterns.

108

The usual purpose of a transformation is to change the scale of
measurements in order to make statistical analysis more valid. If the
variance tends to change with the mean of the measurements, the vari-
ance will only be stabilized by a suitable change in the scale of the meas-
urements., The transformed scale should be one for which real effects
are linear and additive with the variation being normally distributed.
Bartlett (1947) pointed out that when data consist of an integer type
(whole numbers) with heterogeneity, especially if the data have been
collected under field conditions, the square root transformation may be
an appropriate transformation.

INTERSPECIFIC AND INTRASPECIFIC ASSOCIATIONS

The use of correlation coefficients as an index of biological asso-
ciation was questioned by Cole (1946) because the frequency distribution
of organisms in samples commonly differ so widely from a random dis-
tribution that the validity of applying ordinary correlation methods to
such data becomes highly questionable,, Alley and Anderson (1968) showed
that the Sphaeriidae and Chironomidae were randomly distributed while
the Oligochaeta were distributed as a negative binomial distribution in
the short-term study area. Thus 5 biological associations derived from
correlation coefficients between large Pontoporeia , sphaeriids, and
chironomids would not violate the objections of nonrandomness. In this

investigation there was only concern for a significant positive or negative
association, and not the degree of association, so it is felt that these

109

relationships are attributable to either biological or environmental
phenomena.

It is often difficult to interpret the implications of biological asso-
ciation because the term itself is a loose definition which makes no dis-
tinction between relationships derived from mutualism, parasitism,
symbiosis, and predation with those based on similar or dissimilar hab-
itat requirements. The patterns of association are also influenced by
the size of the units from which the samples are obtained. The relation-
ships between organisms may vary from time to time,

SegerstrSle (1965), in an investigation of the Baltic Sea, found an
inverse relationship between the successful recruitment of the bivalve
Macoma baltica and the abundance of Pontoporeia affinis. He also found
a significant negative relationship between the abundance of the priapulid
worm Halicryptus spinulosus and jP. affinis . Segerstrale concluded that
Pontoporeia ingested the spat of Macoma and the eggs of Halicryptus as
it fed on detrital materiaL Although in Lake Michigan the larvae of the
Sphaeriidae develop within the mantle cavity of the adult clam, the im-
mature individuals, when released from the marsupia, are often small
enough to be ingested by large amphipods» Because little is known about
the complete feeding habits of Pontoporeia it is difficult to determine if
this amphipod actively seeks small clams and the eggs of the oligochaetes-

The positive associations between total Pontoporeia and the other
macrobenthic groups of the long-term study area probably reflect the

110

differences attributable to a larger quadrat sampling area and the
greater variability between the 35 stations. It is felt that these results
merely reflect the fact that, in general* a fairly large quadrat will mask
patterns of microassociations and those regions within the lake that are
preferred by Pontoporeia are also selected by the other macrobenthic
groups.

ENVIRONMENTAL RELATIONSHIPS
Environmental Parameters Treated Individually

Depth : Eggleton (1937), in a study of the Lake Michigan macro-
benthos, also found a concentration of Pontoporeia at the 30-40 m depth
zone. The density increased from 10 to 30 m, and beyond 40 m the den-
sity sharply decreased. Merna (I960), also in an investigation of the
Lake Michigan macrobenthos, did not find the 30 m concentration zone
of Pontoporeia , however, his shallowest sampling depth was approxi-
mately 20 m<, Marzolf (1963) found no correlation with depth in a study
of Grand Traverse Bay, but his sampling depths began at 20 m<,

Bottom Temperature : Gordeev (1952) reported that the optimum

temperature range for Pontoporeia , as determined through laboratory

studies, is 8. to 12. C. The upper limit of tolerance according to

field observations is 14. to 20. 0°C as reported by Samter and Weltner

1904; Ekman 1915; Thienemann 1928; Segerstrale 1937; and Bousfield

1958 . It is evident from the present Lake Michigan field observations
that this amphipod can tolerate an extreme range of bottom temperatures

Ill

and does not show any specific temperature preference.

Substrate Relationship : Marzolf (1963), in a laboratory study,
found that Pontoporeia selected sediment particle sizes smaller than
0. 50 mm in diameter but did not discriminate between smaller sizes .
Sand is composed of particles ranging from 1. mm to 0* 062 mm in dia-
meter, silt from 0. 062 to 0. 0039 mm, and clay less than o 0039 mm.

Examination of the percent organic carbon content of the sediment
in the present study revealed that Lake Michigan sand has an average
carbon percentage of 0. 14 with a range of 0. 04 to 0. 22, sandy silt an
average of 0.47 with a range of 0. 26 to 0. 71, silt an average value of
2. 82 with a range of 1. 70 to 4. 02, and the upper stratum of layered sed-
iment an average value of 1„ 08 with a range of 0. 68 to 1„ 69° These
ranges of percent organic carbon indicate that the largest values of sandy
silt overlap with the lowest values of layered sediment, and the largest
values of layered sediment overlap with the lowest values of silt,

Marzolf (1963) concluded from laboratory experiments that the
selection of a substrate by Pontoporeia was significant only when organic
matter was present as a surface film on the sediments. He also found a
significant association between the density of this amphipod and sedimen-
tary bacterial numbers and the amount of organic matter in the sediments.
Zobell and Feltham (1942) showed that certain species of bacteria are
capable of degrading cellulose, chitin, and lignin which most animals
find difficult, if not impossible, to digest. They earlier (Zobell and

112

Feltham 1938) established that a variety of marine organisms can utilize
bacteria as an energy source. Baier (1935) suggested that most detritus
feeders are nourished by the bacteria which decompose the detritus rather
than by the detritus upon which they appear to be feeding, Marzolf found
that in the laboratory Pontoporeia actively selected the sediments that had
been "conditioned" by the growth of bacteria.

The gut contents of several Pontoporeia collected from the lake and
of several maintained in the laboratory were examined in the present
study. Aliquots of cultured algae were added weekly to the aquaria of the
laboratory reared amphipods. The foregut of the lake samples contained
sediments coated with a layer of organic material and an occassional
diatom frustule, while the hindgut contained only sediment and diatom
frustuleso The foregut of the laboratory reared amphipods contained
sediment coated with organic matter and intact algal cells. The hindgut
contained only the sediments and intact algal cells. While the organic
matter on the sediment appeared to be digested, the cellulose walls of
the algal cells appeared to be little affected by the passage through the
digestive systeiru
Stepwise Multiple Regression Analysis of the Environmental Parameters

Perhaps the greatest single problem confronting the analysis of a
biological system is the confounding effects that can occur between
environmental parameters. Within Lake Michigan,, depth is ultimately
related to all the physical parameters measured (with day of year being

113

the exception). In the two dimensional presentation of data, such as
Pontoporeia density versus distance from shore, it is impossible to
determine how much of the apparent relationship is actually attributable
to depth, bottom type or other environmental parameters. Multiple linear
regression is a useful statistical tool in the study of biological systems
because it has the ability of mathematically removing the effects of
confounding environmental factor s.

A remarkable aspect of the multiple regression analysis was the
high multiple correlation coefficients that were obtained at the four depth
zones: 0. 756 for 10-35 m, 0. 599 for 36-65 m, 0. 677 for 66-105 m, and
0.826 for greater than 105 m. The high multiple correlation coefficients
imply that the physical and biological parameters "explain" a large
proportion of the changes in amphipod counts.

Within the 10-35 m depth region of the lake, the multiple regression
equation showed a significant interaction of depth, sediment type, and
distance from shore which indicated that inshore regions of the lake pre-
sent an unsuitable environment for Pontoporeia . The turbulence attrib-
utable to waves and water currents of this region have the ability of
resuspending and redepositing the finer sediments and detritus into other
areas that are less mixed. These shallow areas contain surficial sedi-
ments of coarse beach sand and hard bottom types, and fewer amphipods
are found there.

The necessity of including the depth squared term in the multiple

114

regression equation indicated that the relationship between Pontoporeia
and depth is not simple or linear. This second order polynomial term
probably implies that, in this shallow region of the lake, depth is ulti-
mately associated with the abundance of amphipods, but is also associated
to some extent with the other environmental parameters.

At first glance it is surprising that surface temperature should be
so important for explaining amphipod abundance in the shallower portions
of the lake,, Within this area of the lake the thermocline is usually either
absent or not well established^ so bottom temperature is usually a func-
tion of surface temperature. The positive effect of water temperature
on the abundance of this amphipod is probably attributable to the effect
that temperature has on growth rate. As the amphipods increase in size
they are probably better able to survive in this environment.

Beyond 36 m there was little interrelationship between Pontoporeia
and the physical environment. At the 36-65 m depth zone, sediment
showed the same inverse relationship as it did in the shallower regions
and had about the same influence on amphipod abundance. Depth had a
slight inverse effect on amphipod counts, and beyond 105 m depth had a
very strong negative influence on amphipod density

Within the deeper portions of the lake the rain of particulate matter
is slowed considerably by the water column, Bruun (1957), in an inves-
tigation of the ocean, indicated that over great depths this rain of particles,
which is mainly derived from plankton, is rapidly consumed by the animals

115

living at greater depths. Raymont (1963) indicated that although large
quantities of organic matter are present in bottom muds, this richness
is misleading because the benthos is very much dependent on plankton
supply» This is particularly interesting because no association was
found between percent carbon in the sediment and the abundance of
Pontoporeia .

It is particularly difficult to interpret the significance of the posi-
tive statistical associations that occurred between this amphipod and the
other macrobenthic groups unless they indicate that the taxonomic groups,
and especially the sphaeriids, respond to environmental changes in the
same manner as Pontoporeia . The Sphaeriidae were positively associ-
ated with Pontoporeia both at the four depth zones and when all depths
were combined, while the oligochaetes showed a positive relationship
with Pontoporeia beyond 65 m and no apparent association when the lake
was treated as a unit. The Chironomidae showed a positive association
with this amphipod only at the 66-105 m depth zone, However, the
chironomids did show a positive relationship when all depths were com-
bined.

Robertson (1967) reported that 12 species of the Sphaeriidae were
found at our sampling stations,. Eight of these species, Sphaerium
striatinum form acuminatum , Pisidium casertanum , P. compressum,
P- nitidum form pauper culum , P. fallax , P. henslpwanum , P. idahoense ,
and F\ amnicum occurred between 10 and 20 m. Pisidium lilljeborgi and

116

S. corneum were found at 10 to 50 m while Sphaerium nitidum occurred
between a depth of 20-30 m. Pisidium conventus was found at depths
that ranged from 30 to 200 m„

Hiltunen (1967), in an investigation of the oligochaetes of Lake
Michigan, found 26 species of Tubificidae* 12 species of Niadidae, and
1 species of Lumbriculidae in three regions of the lake. The areas were
located in Green Bay, the southern end of the lake and Ludington harbor.
In the southern part of the lake practically all naidids were collected
at depths that ranged from 5 to 20 m with a few individuals of
Vejdouskyella intermedia being found as deep as 37 m. Members of the
families Lumbriculidae and Tubificidae were found at all sampling depths

(5-74 m) in the southern areas of the lake,

2

Go R, Brenniman, J, A, Soyak, and L L. Curry have examined

the depth distribution of the Chironomidae in Lake Michigan,, They iden-
tified 38 species of midge larvae in the lake with three species ,
Cryptochironomus digitatus , Procladius adumbratus^ and Prodiamesa
(Monodiamesa ) bathyphila being found throughout the lake. The species
of genus Hydrobaenus are found only in the profundal regions of the lake,
while larval forms of Tendipis plumosus , T\ attenuatus , and T\ riparius
are found in the shallower near shore and harbor areas

2

Distribution of Midge larvae (Tendipedidae: Diptera) in Lake

Michigan, Paper presented at the 9th Conference on Great Lakes

Research. Chicago, 111. , 1966,

117

It can be seen, then, that the macrobenthic groups associated with
Pontoporeia represent a composite of several species* Each species has
behavioral characteristics and environmental requirements that are
unique for its niche. These positive associations, as was indicated
earlier, may merely reflect the fact that the larger sampling areas of
the Ponar and Smith- Mclnty re benthos samplers mask the patterns of
the microassociations and those regions within the lake that are pre-
ferred by Pontoporeia are also selected by the other macrobenthis groups .

SEASONAL DISTRIBUTION

During the present study the density of amphipods has been found
to change very little over a 24-hour period of time, and there appear to
be little changes with time in the average standing crop of amphipods
when all stations are combined.

Eggleton (1937) combined the monthly Pontoporeia counts for the
samples collected from Lake Michigan in 1931. He found a slight decline
in the density from May to June, a marked increase from June to July,
a seasonal maximum in August, and a sharp decline in the counts from
August to November. Treating his 1932 samples in the same manner he
found a reverse trend,, The abundance of amphipods declined slightly
from April to June, decreased sharply from June to July, and then in-
creased considerably from July to September. He concluded that the
dissimilarities which existed between the seasonal samples of the two
years reflected the natural tendency of any biological system to vary

118

from year to year,. However, Eggleton had only 167 samples, collected
over a two year period, available for his analysis while more than 1, 700
samples were collected in the three years of the present study. Further,
his 1932 samples were from a different region of the lake than those taken
in 1931.

Henson (1954), in an investigation of a single produndal station in
Cayuga Lake that was approximately 100 m deep, reported that in 1952
there was a significant increase in the total crop of Pontoporeia occurring
from May to October, He found that the 1953 standing crop was signifi-
cantly greater than that in 1952. He attributed the 1953 increase to the
production of young and the immigration of amphipods from the shallower
depths.

Although there was no significant seasonal variation in the abundance
of amphipods at the 10-35 m zone in Lake Michigan there was considerable
variation within sampling seasons, The numerical decrease from March
to April 1966 represents the death of spent females. The low density of
amphipods in the late spring and early summer is probably attributable
to losses arising from the elutriation- screening device that separates the
organisms from the sediment. Large numbers of newly released amphi-
pods may have been swept through the separating screen. As the small
amphipods increased in size, they were less inclined to be lost in this
manner. The remaining three depth zones do not appear to show any
well defined patterns within the sampling periods.

119

It is difficult to interpret the similar patterns of amphipod abun-
dance that occurred at stations C-2 through C-7. Samter and Weltner
(1904), Ekman (1913), and Henson (1954) have shown that as smaller
oligotrophic lakes warm in the spring and summer, Pontoporeia has a
tendency to migrate to the colder hypolimnion. The investigations of
Wells (I960) and Marzolf (1963) have shown that larger Pontoporeia are
active swimmers and these larger amphipods could possibly represent
transient groups that move from area to area. However, a reexamination
of the size frequency histograms showed that there appeared to be little
exchange of size groups within the shallower waters of the lake. These
fluctuations in abundance also did not appear to be the result of repro-
duction.

It is possible that these changes in abundance were the result of
sampling error. The ship logs were examined to determine if the fluc-
tuations in the amphipod counts resulted from samples being taken during
periods of high seas. Sampling in rough weather often caused the dredges
to close prematurely. There appeared to be no association between the
state of the sea at the time the samples were taken and the variation in
the amphipod counts. All persons involved in collecting and processing
the samples were well trained in the sampling procedures, thus the col-
lecting and processing of the samples should not represent a source of

error.

Smith (1968) indicated that adult alewives are densely concentrated

120

on the bottom in the deepest parts of Lake Michigan in mid-winter where
they seek the warmest water of the lake. In late winter and early spring
they move shoreward in dense schools, spawn near shore, in rivers, and
in bays during the late spring and summer, and begin to migrate back to
open waters of the lake in the fall. He feels that when the alewife is
concentrated inshore in the summer, the deeper water, previously in-
habited by fishes that were replaced by the alewife, are essentially un-
occupied and the invertebrate fauna is incompletely utilized* He feels
that similar situations occur during other seasons as the alewife moves
from zone to zone. '~*~~

Morsell and Norden (1968) reported that as the length of alewives
increased beyond 119 mm they progressively relied more and more on
P. af finis as a source of food. They found, by stomach analysis, that
Pontoporeia constituted nearly 80 percent of the dry weight for the total
stomach content of alewives living in the littoral and sublittoral regions
of the lake.

The seasonally discriminant exploitation of Pontoporeia by alewives,
as well as amphipod predation by other bottom feeding fishes, could
possibly contribute the observed seasonal fluctuations.

PATTERNS OF SPATIAL ABUNDANCE

Powers and Robertson (1965) found the standing crop of Pontoporeia

higher in the inshore areas of the northern basin of Lake Michigan than
comparably placed stations located in the southern basin. Their results

121

generally agree with the findings of the present study.

Whenever the effect of depth is statistically removed from the
abundance of Pontoporeia , the average density appears to be highest at
the most western inshore stations in the northern basin of the lake,
Ayers et al. (1958) have shown that areas on the western margin of the
lake, in the vicinity of stations E-6 and D-6, are periodically subjected
to upwelling and currents sweeping down from Green Bay, The increased
productivity attributable to upwelling plus the added nutrients carried
south from the Green Bay region could make the western protions of the
lake more productive than would normally be expected. The inshore
regions of the eastern portions of the lake., in proximity to the C 9 D,
and E transects, are subjected to upwelling caused by short-lived NE
winds and the normal outflow current which is enriched by local runoff
and nutrients sweeping north from the southern basin. The amphipod
density also appears to be higher within these regions „

Station B-8, located approximately 2 miles from Waukegan, Wis, 9
station A-6 situated 9 miles from Chicago, and A-l located 2 miles from
Benton Harbor, Mich, , are found within the regions of the southern lake
basin that have been classified as grossly polluted by the Federal Water
Pollution Control Administration (1968), Although it is currently fashions-
able to regard areas in the lake with low Pontoporeia densities as areas
of gross pollution, it must be pointed out that the sediment types in these
regions are quite variable, ranging from fine beach sands to glacial tills,

122

and these shallow areas are subjected to considerable turbulence during
periods of intense storms. While there is considerable truth to the fact
that these regions have been subjected to pollution., they are probably-
natural regions of low amphipod productivity.

CONCLUSIONS

This amphipod, Pontoporeia affinis , is a remarkable organism.
The field observations of this study indicate that Pontoporeia is capable
of enduring a wide range of environmental conditions. In Lake Michigan.,
it is found from the sublittoral to the deepest profundal areas and tolerates
water temperatures that range from 1 to 19 centigrade. Although
Pontoporeia is a burrowing amphipod it is also found, in low densities,
within hard bottom surficial sediments . Juday and Birge (1927), in an
investigation of Green Lake, Wis. , found that Pontoporeia was able to
survive within the sediments when the water contained 0, 72 cc/1 dis-
solved oxygen one meter above the bottoiru

The results of this study strongly indicate that Lake Michigan can-
not be treated as a unit in the analysis of the interrelationships of
Pontoporeia and its environment . This amphipod has the extraordinary
capability of adjusting to the ambient environmental conditions and has
even adapted its reproductive cycles to comply with these local conditions.

Within the sublittoral regions of the lake a Pontoporeia has developed
two distinct reproductive patterns that appear to be related to the water
temperature,, It matures in one year in the warm shallow inshore regions
of the sublittoral and requires two years to mature in the deeper portions
of this zone. Pontoporeia of the sublittoral are geared for a late winter-

123

124

early spring period for reproduction; in other depth zones they breed
year round.

The greatest environmental stresses for Pontoporeia also occur
within the sublittoral. Later visits to the sublittoral short-term study
area (in August and November) showed that molar action attributable to
storms and bottom currents had caused considerable shifting of the bottom
sediments and had resuspended the detrital material. The greater vari-
ability in the amphipod counts of the sublittoral zone found throughout
our studies probably reflects the disrupting forces attributable to ambient
meteorological and limnological conditions and to the general environ-
mental heterogeneity of this region.

Pontoporeia reaches its maximum density at the junction of the
sublittoral and profundal zones. This region corresponds roughly to the
lower limit of the thermocline. The surficial sediments within this area
are little affected by turbulence, and the range of bottom temperatures is
not so extreme as in the shallower inshore waters. In this area and the
adjacent profundal zone, Pontoporeia matures and breeds intermittently
throughout the year. There appears to be a greater proportion of breed-
ing females in the amphipod population that inhabits this area and the
contiguous profundal areas.

The junction of the sublittoral and profundal zones represents a
reasonably stable environment where Pontoporeia can effectively utilize
the rain of suspended particulate: matter that is created through the enrich-

125

meat by upwelling^Loical:^ shore

current. It must be emphasized that this junction changes with the sea-
sons, and its location within an area is dependent on the mo rpho metric
features of the lake and the prevailing meteorological conditions.

The profundal zone represents the most stable environment of the
lake with all areas of the profundal sharing a common low water temper-
ature. The Pontoporeia of this region probably require three or more
years to mature, and although breeding is intermittent throughout the
year only a very small portion of the population appears to be mature at
any particularvtime. The density of Pontoporeia in this zone shows a
strong inverse relationship with depth, which most likely reflects the
negative effect that depth has on the rain of suspended particulate matter.
The lack of available nutrients is probably the most restrictive element
of the profundal environment.

^ s Pontoporeia is the only macrobenthic invertebrate that is ubiq-
uitously distributed throughout the sublittoral and profundal zones of the
hike, it should play an increasingly important role as an indicator of
water pollution. Cook and Powers (1964) showed that a benthic community
subjected to the effluent of the St. Joseph River contained few Pontoporeia
and an abundance of oligochaetes. On the other hand, a benthic commun-
ity adjacent to Little Sable Point, which has no major tributaries, con-
tained a large population of Pontoporeia and few oligochaetes. Ayers
and Huang (1967) included macrobenthic sampling in a pollution study of

126

Milwaukee Harbor and the adjacent embayment. Inside the breakwater
they found a high density of oligochaetes, no amphipods, and considerable
environmental degradation. The embayment immediately outside the
breakwaters also showed environmental degradation, high densities of
oligochaetes, and occasionally Pontoporeia.

The low density Pontoporeia, as a criterion for water pollution, is
very unsatisfactory because the abundance of this amphipod is dependent
on many environmental conditions, some of which bear no relation to
water pollution. This study showed that bottpm sediments can strongly
affect the density of Pontoporeia. High water temperatures and low
oxygen tensions, which were previously considered limiting environmental
factors for this amphipod, are not necessarily a deterrent for their hab-
itation. Before an area can be classified as polluted, by using the low
levels of Pontoporeia as the indicator of pollution, a very careful exam-
ination must be made of the environment to determine if low densities are
attributable to pollution or the disrupting effects of an environmental
factor.

LITERATURE CITED

Adamstone, F, B. , 1928, Relict amphipods of the genus Pontoporeia,
Trans, Am, Micros, Soc. 47: 366-371-

Alley, W. P., and R. F. Anderson. 1968. Small- scale patterns of
spatial distribution of the Lake Michigan macrobenthos, Proc.
11th Confo Great Lakes Res. ; Internat. Assoc, for Great Lakes
Res, In press.

Ayers, J, D, , D. C. Chandler, G. H. Lauff, C, F, Powers, and
E, B. Henson, 1958, Currents and water masses of Lake
Michigan, Univ. Michigan, Great Lakes Res, Div, Spec. Rep,
No, 3, 169 p

, and J. L. Hough, 1964, Studies on Water movements and sedi-
ments in southern Lake Michigan. Part II. The surficial bottom
sediments in 1962-1963, Univ. Michigan, Great Lakes Res. Div.
Spec. Rep. No. 19, 47 p.

, and D. C. Chandler, 1967, Studies on the environment and

eutrophication of Lake Michigan, Univ. Michigan, Great Lakes
Res. Div, Spec. Rep. No, 30, 415 p.

, and J. C. K. Huang. 1967. Studies of Milwaukee harbor and

embayment, Univ, Michigan, Great Lakes Res. Spec, Rep,
No, 30: 372-394,

Baier, C, R. , 1935. Studien zur Hydrobakteriologic stehender
Binnengewasser, Archiv. f, Hydrobiol. 29: 183-264,

Bcirtlett, M. S. , 1947. The use of transformations. Biometrics
3: 39-52,

Bousfield, E, L , 1958, Freshwater amphipod crustaceans of glaciated
North America, Can, Fid, Nat, 72: 55-113,

Bruun, A, F, , 1957, Deep sea and abyssal depths. In Treatise on

Marine Ecol, and Paleocol, Vol, I, Ecol. , J. W. Hedgpeth, Ed,
Geol, Soc. Amer, Memoir, 67: 641-672,

Cole, L. C. , 1946, A theory for analyzing contagiously distributed
populations. Ecology 27: 329-341.

127

128

Cook, G. W. , and R. E» Powers, 1964. The benthic fauna of Lake
Michigan as affected by the St. Joseph River,, Proc, 7th Conf.
Great Lakes Res. ; Univ. Michigan, Great Lakes Res. Div.
Spec. Rep. No. 11: 68-76.

Cooper, J. E, , 1962. Seasonal changes with depth in population of
Pontoporeia affinis (Amphipoda) in South Bay* Lake Huron.
Proc. 5th Conf. Great Lakes Res. ; Univ. Michigan^, Great Lakes
Res. Div. Spec. Rep. No. 9: 173. (Abstr. )

Eggleton, F. E. , 1936. The deep-water bottom fauna of Lake Michigan.
Pap. Mich. Acad. Sci. , Arts, and Lett. 21: 599-612.

. 1937. Productivity of the profundal benthic zone in Lake

Michigan. Pap. Mich. Acad, Sci., Arts, and Lett. 22: 593-611.

Ekman, S. , 1 91 3 C Zwei neue europUische Arten der Amphipodengattung
Pontoporeia Krgfyer. Arkiv. F. Zool. 8.

Evans, F. C. , 1952. The influence of size of quadrat on the distributional
patterns of plant populations. Univ. Michigan, Cont. Lab. Vert.
Biol. No. 54: 1-15.

Fager, E. W. , A. O. Flechsig, R. F. Ford, R. I. Clutter, and R. J.
Ghelardi. 1966. Equipment for use in ecological studies using
SCUBA Limnol. and Oceanog. 11: 503-50.9.

Federal Water Pollution Control Admin. 1968. Water quality investi-
gations, Lake Michigan basin* A technical report containing back-
ground data for a water pollution control program. FWPCA,
Great Lakes Region, Chicago, 111. 40 p.

Gordeev, O, N. , 1952. Biology and ecology of the relict crustacean
Pontoporeia affinis Lindstrbm in the lakes of Karelia. Uchon.
Zap. Karelo-Finsk. Univers. , Biol. Nauk. 4: 98-109.

Green, R„ H. , 1965. The population ecology of the glacial relict

amphipod Pontoporeia affinis Lindstrbm in Cayuga Lake, New
York. Ph.D. Thesis. Cornell Univ. 110 p.

Henson, E. B. , 1954. The profundal bottom fauna of Cayuga Lake.
Ph.D. Thesis. Cornell Univ. 108 p.

. 1966. A review of Great Lakes benthos research. Univ.

Michigan., Great Lakes Res. Div. Pub. No. 14: 37-54.

129

Hiltunen, J. K, , 1967. Some oligochaetes from Lake Michigan. Trans .
Amer. Micros, Soc, 86: 433-454.

Juday, C. , and E. A. Birge. 1927. Pontoporeia and Mysis in
Wisconsin lakes. Ecology 8: 445-452.

Larkin, P. A. , 1948, Pontoporeia and Mysis in Athabaska, Great
Bear, and Great Slave Lakes. BulL Fish. Res, Ed, Canada
78: 1-33.

Lindstrdm, G. , 1855. Bidrag till KUnnendomen om Ostersjdns
Invertebrat-fauna. Ofv. Kgl. Vet. Ak. Fbrh. , Arg. 12,

Marzolf, G, R. , 1963, Substrate relations of the burrowing amphipod
Pontoporeia affinis LindstrOm, Ph. D 8 thesis* Univ. Michigan,
92 p.

. 1965, Vertical migration of Pontoporeia affinis (Amphipoda) in

Lake Michigan. Proc. 8th Conf. Great Lakes Res. ; Univ.
Michigan, Great Lakes Res. Div. Spec, Rep, No. 13: 133-140.

Miclntyre, A. D. , 1954, A spring-loaded bottom sampler. J. Marine
Biol. Assoc, U. K. 33: 257-264,

Merna, J. W, , I960. A benthological investigation of Lake Michigan,
M.S. Thesis, Michigan State Univ. 74 p.

Morsell, J. W. , and C. R. Norden. 1968, Food habitats and seasonal
condition of the alewife in Lake Michigan,, Proc. 11th Conf, Great
Lakes Res, ; Internat, Assoc for Great Lakes Res, In press.

Norton, A, H. , 1909. Some aquatic and terrestrial crustaceans of the
State of Maine, Proc, Portland Soc, Nat, Hist, 2,

Powers, C, F. , and A, Robertson, 1965, Some quantitative aspects of
the macrobenthos of Lake Michigan. Proc. 7th Conf, Great Lakes
Res, ; Univ, Michigan, Great Lakes Res, Div, Spec, Rep. No. 13
153-159.

, and W, P. Alley. 1967. Some preliminary observations on the

depth distribution of macrobenthos in Lake Michigan. Univ.
Michigan, Great Lakes Res. Div. Spec, Rep. No. 30: 112-125.

, and A. Robertson. 1967. Design and evaluation of an all-purpose

benthos sampler. Univ. Michigan^ Great Lakes Res. Div, Spec.
Rep. No. 30: 126-131.

130

Powers, C. F. , A. Robertson, S. A. Czaika, and W, P. Alley* 1967,
Lake Michigan biological data, 1964-66,, Univ. Michigan, Great
Lakes Res. Div. Spec* Rep. No. 30: 179-227.

, and . 1968o Subdivisions of the benthic environment of the

upper Great Lakes* with emphasis on Lake Michigan, J. Fish.
Res, Bd. Canada 25: 1181-1197.

Preston, F. W. 1948. The commonness and rarity of species.
Ecology 29: 254-283.

Raymont, J. E. G. 1963. Plankton and productivity in the oceans.
Macmillan Co. , N. Y. 659 p.

Robertson, A. , and W. P. Alley. 1966. A comparative study of Lake
Michigan macrobenthos. LimnoL and Oceanog. 11: 576-583.

. 1967. A note on the Sphaeriidae of Lake Michigan. Univ.

Michigan, Great Lakes Res. Div. Spec. Rep. No. 30: 132-135.

Samter, M. , and W. Weltner. 1904. Biblogische Eigentlimlichkeiten
d er Mysis relicta , Pallas iella quadrispinosa und Pontoporeia
af finis erklfelrt aus ihrer eiszeitlichen Entstehung. Zool. Ans.
27: 676-694.

SegerstrSle, S„ G, , 1937. Studien liber Bodentierwelt in

Siidfinnlkndischen Kustengewassern III. Zur Morphologie und
Biologie des Amphipoden Pontop oreia affinis, , nebst einer Revision
der Pontoporeia -Systematik. Soc.Sci. Fenn. , Comm. BioL
7: 1-183.

. 1950. The amphipods of the coast of Finnland-some facts and
problems. Soc. Scio Fenn. Comm. Biol. 10: 1-28.

. 1959» Synopsis of data on the crustaceans Gammarus locusta,
Gammarus oceanicus , Pontoporeia affinis g and Corophium
volutator (Amphipoda Gammaridea), Soc. Sci. Fenn. Comm.
Biol. 20: 1-23.

1965 Biotic factors affecting the vertical distribution and
abundance of the bivalve, Macoma baltica (L. ), in the Baltic Sea
Botanica Gothoburgensia III. Proc. 5th Marine Biol. Symp. ,
195-204 p.

9 1967. Observations of summer-breeding in populations of the

glacial relict Pontoporeia affinis Lindstr. (Crustacea Amphipoda),
living at greater depths in the Baltic Sea, with notes on the repro-
duction of P. femorata Krbyer. J. Exp. Marine Biol. EcoL 1: 55-64,

131

Smith, S. H. , 1968* Species Succession and Fishery Exploitation in
the Great Lakes. J. Fisho Bd, Canada, 25; 667-693*

Smith, S, I. , 1874. The Crustacea of the fresh waters of the Unites

States. Rep* Comm. Fisho and Fish, , 1872 and 1883: 637-665.

Steel, R. G D. ,and J. H, Torrie, I960. Principles and procedures
of statistics* McGraw-Hill Book Co. , Inc. No Y, 481 p c

Teter, H, E. , I960. The bottom fauna of Lake Huron, Trans Amer,
Fisho Socv, 89: 193-197,

Thienemann, A..H28. Die Reliktenkreb Mysis relicta , Pontoporeia affinis,
Pallasiella quadrispinosa und die von ihnen bewohnten nortdentschen
Seen, Arch. 'Hydrobiol. , 19: 521-582.

Wells, Lo I960* Seasonal abundance and vertical movement of

planktonic Crustacea in Lake Michigan, Fish, Bull. U. S. Fish.
Wildl. Serv. , 60: 343-369*

Zobell, C. E, and C. B, Feltham. 1938, Bacteria as food for certain
marine invertebrate. Jour, Mar. Res, 1: 312-327,

, and . 1942, The bacterea flora of a marine mud flat as an

ecological factor. Ecology 23: 69-78,