£12.23.)
. L

. .1! frtv

v.1
0‘.

Li»:
I.

w:

:5

I

a)
. 1|

.5‘41 y...

‘1'. .l A?! .

g I.”

.
y

T. .

w

.

gunk.»
-. .

:151. .
a}... v.

.J. In an.

;

 

, . . . . . ‘ . u v , "f » . a
......L»._r‘.m Flue.“ .kkwbnrufih Jaivgg.

4 .1: l...’.5.i.h|u)v‘.

 

 

 

 

‘3-35818

MICHIGAN STATE

H l 1|le

 

ll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HUN!!!Pli'llWlWHlHHlNHi

W53411098

This is to certify that the

dissertation entitled

Ninterinq Strategies and Transmission Line
Mortality of Common Mergansers in Lower Saginaw Bay

presented by

Paul Irving Padding

has been accepted towards fulfillment
ofthe requirements for

 

 

Doctor of Philosophy (kgnwin Fisheries and wiidiife

  

 

Major professor

Dam: March 12, 1993

 

513L113 an Affirmative Aaron ’L‘qua/ Oppormmry Institution 0- 12771

 

 

 

 

 

K.

"fi

fi‘

LIBRAR r
"hunt! State
University

L
.J

 

 

_

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before dme due.

______________________———————7
DATE DUE DATE DUE DATE DUE

,ji 4..
::JL

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

L,_______

 

 

 

 

 

’fJ’J J

MSU Is An Affirmative Action/Equal Opportunity Institution
Wmea-p. t

 

 

WINTERING STRATEGIES AND TRANSMISSION LINE MORTALITY
OF COMMON MERGANSERS IN LOWER SAGINAW BAY

By

Paul Irving Padding

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

1993

ABSTRACT

WINTERING STRATEGIES AND TRANSMISSION LINE MORTALITY
OF COMMON MERGANSERS IN LOWER SAGINAW BAY

By

Paul Irving Padding

Age and sex specific differences in fasting endurance and behavioral
dominance have been advanced as hypotheses to explain the preponderance of
adult males often observed in waterfowl populations using northern wintering
sites. To assess these hypotheses, numbers, habitat use, time-activity budgets.
and foraging success of adult male and other common mergansers (Mergus
merganser) observed in lower Saginaw Bay were compared among 3 winters (Nov.
1988 - March 1991). Each year, 10,000-12,000 mergansers, of which 6-1 5%
were adult males, arrived in the area in late fall and early winter to feed on gizzard
shad fry (Dorosoma cepedianum) that aggregated in the Karn-Weadock power
plant's thermal discharge zone. More than 50% of the mergansers left the area in
late January, after the shad dispersed. During 2 winters, the proportion of adult
males in midwinter increased 2- to 3-fold as more than half of the adult males
stayed in the area while most of the other birds departed. Adult males spent less
time foraging in midwinter than other birds, as predicted by the fasting endurance
hypothesis. However, there was no change in the proportion of adult males from
early to midwinter during the third year, indicating that fasting endurance is not
solely responsible for the age and sex distribution of wintering mergansers. There
was no evidence to support the behavioral dominance hypothesis.

A collision mortality problem developed at the Karn-Weadock site as a

result of migrating and wintering common mergansers flying across transmission

lines that are concentrated in flight corridors. Flight patterns were observed to
identify problem corridors, and collision mortality was estimated from observed
collisions and numbers of carcasses found under transmission lines that bisected
flight paths. An estimated 624 collisions resulting in death and 907 resulting in
serious injury occurred during the 3 winters. In December and January 1989-90
and 1990-91 , an attempt was made to direct birds away from the lines by
obstructing key flight paths with 1.7-m diameter, helium-filled balloons placed at
transmission line height. Balloons were removed every other week to assess their
effectiveness. Obstruction of flight paths caused significant (P< 0.001) shifts in
merganser flight patterns that resulted in a reduction of documented mortality
from 41 fetal collisions when balloons were absent to none when balloons were

present .

ACKNOWLEDGEMENTS

This study was funded by Consumers Power Company. I extend my
thanks to the Consumers Power environmental planning staff, particularly Dr. Gary
A. Dawson and Mr. Dale Leinberger, for arranging on-site logistic and facilities
support for the project whenever they were needed.

I would like to express my deepest appreciation to Dr. Harold H. Prince, my
major professor, for his guidance and friendship. As my mentor during my years
of graduate study, he greatly enhanced my life both professionally and personally.
My sincere appreciation is also extended to the other members of my guidance
committee, Dr. Scott R. Winterstein, Dr. Donald L. Beaver, and Dr. Richard J.
Aulerich, for the constructive criticism and advice they provided me. I also thank
department secretary Carla Dombroski for her invaluable assistance on numerous
occasions and, equally importantly, for her friendship.

I would like to thank my field assistants, Joseph Hazewinkel, Douglas
Peterson, Karen Awrylo, and Matthew Gray, for their efforts. They were excellent
workers and good company all, and each of them came up with several ideas that
resulted in improvements to the project. I also thank fellow graduate students
Cathy Cook, Lou Bender, Susan Traylor, Gregg Hancock, Gary Roloff, John Kocik,
Dan Hayes, and Russ Brown for their friendship and support.

Finally, I thank my daughter Rachel and my parents Harold and Viola for
their love, encouragement, and moral support throughout my education. My

accommishments reflect their sacrifices and efforts as well as my own.

iv

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
PART I: WINTERING STRATEGIES

INTRODUCTION
STUDY AREA
METHODS
Counts
Habitat Use
Behavior
Foraging
RESULTS
Population Trends
Habitat Use
Behavior
Foraging
DISCUSSION

PART II: TRANSMISSION LINE MORTALITY

INTRODUCTION

STUDY AREA

METHODS
Daily Movements And Flight Patterns
Collision Mortality
Factors Affecting Collision Rates
Experiment

RESULTS
Counts
Flight Patterns
Collision Mortality
Factors Affecting Collision Rates
Experiment

DISCUSSION

BIBLIOGRAPHY

vi

viii

38

38
39
39
39
41
42
42
43
43
45
45
49
50
54

57

LIST OF TABLES

Proportion and maximum number of adult males in populations
of migrating and wintering common mergansers using lower
Saginaw Bay from November through March.

Mean number of common mergansers flying either toward inland
habitats (Inland) or toward the Karn-Weadock discharge plume
in lower Saginaw Bay (Plume) during winter (Nov. - Mar.).

Mean number of common mergansers flying either toward inland
habitats (Inland) or toward the Karn-Weadock discharge plume
in lower Saginaw Bay (Plume) in the evening during winter

(Nov. - Mar.).

Mean number/day of common mergansers using 3 habitats in lower
Saginaw Bay during winter, 1988-89 - 1990-91.

Proportion of adult males (P) in groups of common mergansers
observed in 3 habitats in lower Saginaw Bay from November -
March, 1988-91. Sample size (n) represents total common
mergansers identified during multiple independent observations.

Number of behavioral observations (scans) of fall migrating and
wintering adult male (AM) and adult female and immature (Other)

common mergansers in lower Saginaw Bay, Dec. 1988 - Mar. 1991.

Test statistic (F) and P-value results of analyses of variance
testing for differences in behavior of common mergansers between
adult males and other birds, among habitats, and between adult
males and other birds within habitats (A/s x H).

Estimated number of low altitude (_<_2m above transmission lines)
flights/day by common mergansers crossing transmission lines at
the Karn-Weadock power plant site during late fall and early
winter (Nov. - Jan.).

Mean number of flights/day and percent low altitude (52m above
transmission lines) flights by common mergansers using the
Karn-Weadock power plant site during late fall and early

winter (Nov. - Jan.).

vi

15

17

18

27

47

10.

11.

12.

13.

Estimated number of transmission line collisions by common
mergansers at the Karn-Weadock power plant site during late
fall and early winter (Nov. - Jan.) versus mid] and late winter
(Feb. - Mar.).

Estimated collision rates [(collisions / flight) x 100] of common
mergansers crossing transmission lines at the Karn-Weadock
power plant site during late fall and early winter (Nov. - Jan.)
versus mid/ and late winter (Feb. - Mar.).

Estimated number of flights/day by common mergansers using
treatment area and other flight paths during late fall and early
winter (Nov. - Jan.).

Estimated number of flights/day by common mergansers through
the treatment area during late fall and early winter (Nov. - Jan).

vii

51

52

LIST OF FIGURES

Lower Saginaw Bay, the Karn-Weadock (Essexville, Michigan)
power generating station's discharge channel, and the warm
water mixing zone (plume) created by thermal discharge from
the power plant.

Maximum counts (bars) of common mergansers in relation to ice
cover (dotted lines) in lower Saginaw Bay and the Saginaw
River, 1988 - 91.

Diurnal time-activity budgets of common mergansers in lower
Saginaw Bay during winter (November - March), 1988-89 -
1 990-91 .

Seasonal time-activity budgets of migrating and wintering
common mergansers in lower Saginaw Bay during winter
(November - March), 1988-89 - 1990-91.

Time-activity budgets of adult male (AM) versus adult female
and immature (Others) common mergansers in lower Saginaw
Bay during mid/ and late winter, 1989 - 1991.

Midwinter time-activity budgets of adult male (AM) versus
adult female and immature (Others) common mergansers using
2 habitats in lower Saginaw Bay, 1989 - 1991.

Number of forage fish caught using dip nets in the Karn-Weadock
(Essexville, Michigan) power plant's discharge channel during
late fall and early winter, 1990-91.

The Karn-Weadock (Essexville, Michigan) power generating
facility at the mouth of the Saginaw River.

Mean number/day of common mergansers using the Karn-

Weadock (Essexville, Michigan) power plant's discharge
channel at night.

viii

13

PART I: WINTERING STRATEGIES
INTRODUCTION

Although the majority of North American waterfowl winter in southern
United States and Central and South America (Bellrose 1980), populations of
several species winter farther north. Most North American sea ducks [eiders
(Somateria spp.). scoters (Melanitta spp.), oldsquaw (Clangula hyemalis), and
harlequin ducks (Histrionicus histrionicus” winter north of the 40th parallel along
the Atlantic and Pacific coasts (Bellrose 1980, Reed 1986). Likewise, many of
the black ducks (Anas rubripes) and canvasbacks (Aythya valisineria) that winter
in the Atlantic Flyway remain in northern coastal habitats throughout the winter
(Diefenbach er al. 1988, Serie at al. 1983). These populations exhibit fidelity to
specific coastal wintering sites from year to year, and stay in their traditional
wintering areas even during extended periods of extremely harsh weather (eg.
Longcore and Gibbs 1988).

Other wintering waterfowl utilize inland habitats in the north. Because
food availability is less stable and predictable in inland habitats than in coastal
habitats during winter (Diefenbach er al. 1988). these birds tend to be facultative
migrants (i.e., they remain in northern areas only as long as those areas provide
their winter requirements). Individuals may benefit from a flexible wintering
strategy such as this in several ways. Ketterson and Nolan (1976, 1979) noted
that energetic costs of migration are reduced for birds that winter farther north
than other conspecifics. Furthermore, wintering closer to theibreeding grounds
may enhance their timing of arrival on breeding grounds (Ketterson and Nolan

1

2

1976, 1979). Individuals of several species and subspecies of waterfowl have
adopted the facultative migrant strategy, including mallards (Anas plaryrhynchos)
(Bellrose and Crompton 1970, Nichols at al. 1983, Jorde at al. 1984). common
goldeneyes (Bucephala clangula) (Sayler and Afton 1981). common mergansers
(Mergus merganser) (Bellrose 1980), giant Canada geese (Branta canadensis
maximal (Hanson 1965), and to a lesser extent, black ducks (Bellrose and
Crompton 1970) and interior Canada geese (B. c. interior) (Cook 1990).

Technological and industrial development in the 20th century has resulted
in the creation of numerous artificial environments, including permanent open
water zones in areas well north of the traditional wintering grounds used by most
waterfowl. These potential wintering areas in northern latitudes provide
alternative winter habitats for some waterfowl species, particularly facultative
migrants. Significant numbers of mallards, goldeneyes, and mergansers currently
winter in open water zones created by hydro-electric dams (Reed and Bourget
1977, Thompson et al. 1988). municipal waste and water treatment operations
(Cooper and Johnson 1977), and warm water discharge from power generating
plants (Haymes and Sheehan 1982) and other industrial plants (Sugden er al.
1974, C00per and Johnson 1977).

The facultative strategy may result in increased survival and/or reproductive
success. However, physiological and ecological constraints may limit the ability of
some species, or demographic groups within species, to take advantage of these
habitats. Food resources must be adequate to meet the greater energetic
demands that are imposed by harsh winter climate in northern latitudes. Although
waterfowl species often minimize interspecific competition for food through

habitat segregation (Stott and Olson 1973, White and James 1978), limited food

3

supplies within habitat segments may result in intraspecific competition. In
addition to food requirements, wintering waterfowl need favorable habitats for
other activities. Requirements for roosting, resting, body maintenance activities,
and courtship may not be adequately met in feeding habitats (Gray at al. 1987). If
the energetic demands of movement between habitats are offset by survival or
reproductive benefits derived (eg. greater security or space for courtship and
pairing). optimal wintering areas may consist of several different habitats.
Wintering areas in northern latitudes are not equally suitable for all sex and
age classes within a species, as evidenced by a general pattern of skewed sex and
age ratios that has been documented in wintering populations of several species of
ducks. Adult males were predominant in populations of common mergansers
wintering in northcentral United States, whereas adult females and immature
mergansers were more abundant farther south (Anderson and Timken 1972).
Similar winter distribution patterns have been shown for common goldeneyes
(Sayler and Afton 1981), mallards (Jorde at al. 1984), and canvasbacks (Nichols
and Haramis 1980). Ketterson and Nolan (1976, 1979) hypothesized that in
addition to possible differences among age and sex classes in costs and benefits
associated with migration and timely arrival on breeding grounds, the social status
and body size of individuals may influence wintering distribution patterns. Sayler
and Afton (1981) concluded that differences in body size and weight between
males and females and/or intraspecific competition for food among sex and age
classes are among the causative factors underlying the skewed sex and age ratios
observed in many wintering duck populations. Larger, heavier adult males require
relatively less energy for maintenance than adult females and immature birds, and

their position of social dominance favors them in competition for limited food

4
resources (Sayler and Afton 1981). Perry of al. (1988) treated these as

competing hypotheses, terming the former the fasting endurance hypothesis and
the latter the behavioral dominance hypothesis. Although their studies of behavior
and weight loss in stressed captive canvasbacks did not provide support for either
hypothesis, they believed that their results were insufficient to conclusively refute
the hypotheses.

Little is known about the factors that influence the choice of wintering
sites by piscivorous species such as the common merganser. Migrating common
mergansers arrive in the Great Lakes region in November and early December
(Bellrose 1980). Although scattered small flocks of common mergansers
traditionally wintered in open water on inland lakes and streams and along the
coasts of Michigan, large groups were rare (Salyer and Lagler 1940). However,
when Consumers Power Company's Karn-Weadock power generating facility
commenced operations, the plant's thermal discharge created a previously
unavailable open water habitat in lower Saginaw Bay during winter. Migrating and
wintering mergansers began congregating at the Karn-Weadock site in response to
this new habitat; several thousand common mergansers have been observed there
each winter since 1976. This study was an investigation of the wintering
strategies of common mergansers using that permanent open water zone. The
following predictions were tested: 1) the age and sex composition of the
wintering population would be skewed in favor of adult males (consistent with
both the behavioral dominance and fasting endurance hypotheses); 2) adult males
would displace other birds from preferred foraging habitats (predicted by the
behavioral dominance hypothesis); 3) adult males would be more successful at

capturing prey than other birds (predicted by the behavioral dominance

5
hypothesis); and 4) adult males would spend less time foraging than other birds

(predicted by the fasting endurance hypothesis).

STUDY AREA

The study area is located in the vicinity of Bay City, Michigan, where the
Saginaw River drains into lower Saginaw Bay (Figure 1). With the onset of winter,
inland lakes, ponds, and streams that would otherwise provide suitable Open water
habitats for migrating and wintering common mergansers are eliminated by ice
cover. Water temperatures at the mouth of the Saginaw River average 0-2°C
during the winter months (December - March), and most of the river and lower bay
is also often ice-covered. Small, scattered pockets of open water, created by
discharges from several factories and a waste water treatment plant in Bay City,
are present for most of the winter in the Saginaw River.

The Karn-Weadock facility is situated at the mouth of the Saginaw River,
bordered by Saginaw Bay to the north and the Saginaw River to the west (Figure
1). Water imported from the river to cool the plant's turbines is returned to
Saginaw Bay via a discharge channel. This continuous discharge of heated water,
averaging 12-13°C during winter, creates a plume of permanent open water in the
bay (Figure 1) that varies in size from 80-770 ha throughout the winter. The size
and shape of the plume is influenced by temperature, wind velocity and direction,
and water currents and wave action in the bay. The power plant's thermal
discharge attracts some species of fish, including gizzard shad (Dorosoma
cepedianum). in late fall. During November and December, young-of-the-year

gizzard shad aggregate into large schools near the mouth of the Saginaw River and

 

 

3
LOWER
SAGINAW BAY 27
0 PLUME
’ DISCHARGE CHANNEL
ESSEXVILLE
SAGINAW RIVER
BAY CITY
a
o 2 4
9 SCALE (KM)

 

 

 

Figure 1. Lower Saginaw Bay, the Karn-Weadock (Essexville, Michigan)
power generating station’s discharge channel, and the warm water
mixing zone (plume) created by thermal discharge from the power plant.

7

in the power plant's plume (Pope 1976). The timing of these shad movements

coincides with the early winter arrival of migrating common mergansers.

METHODS

Activities of common mergansers were observed 4 days/wk during the
winters of 1988-89 (December-March), 1989-90 (November-March), and 1990-91
(November-March). The objectives of this study were to: 1) describe seasonal
changes in population size and the age and sex composition of the population; 2)
compare seasonal habitat use by adult males versus other birds; 3) construct and
compare seasonal time-activity budgets for adult males versus other birds; and 4)
document and compare foraging success of adult males versus other birds.
Counts

Common mergansers in the area were counted at midday 1-4 days/wk,
using binoculars or a spotting scope. During early winter 1988-89, midday counts
covered the power plant's discharge channel and the plume. When other sites
used by mergansers in the area were identified as the winter progressed, they
were included in midday counts. As a result, by late winter midday counts were
also taken at the mouth of the Saginaw River and along a 20-km segment of the
river extending south from the river mouth. Midday counts in 1989-90 and 1990-
91 covered all sites that were identified as potential merganser habitat in 1988-
89.

Mergansers using inland habitats at night were also counted. Initial
observations indicated that most of the birds in the area were spending the night
in the discharge channel, therefore either evening or early morning counts were

conducted 4 days/wk from vehicles parked on the dikes along the channel.

8

Evening counts consisted of enumerating mergansers in the channel in late
afternoon and birds that flew into the channel from late afternoon until 0.5 hr
after sunset. During early morning, birds swimming or flying out of the channel
were counted from 1-1.5 hr before sunrise to 0.5 hr after sunrise. It was
assumed that mergansers did not enter or leave the discharge channel during the
night. In February and March, mergansers in the Saginaw River were counted at
night from the shore 21 night/2 wk.

The sex and age composition of the population was estimated by sampling
randomly selected groups of birds after counts were completed and during
behavioral observations. Adult females, immature males, and immature females
could not be reliably differentiated, thus birds were identified as either adult males
or others. Estimates of the number of adult males present were extrapolated by
multiplying the proportions of adult males in the sample groups by the
corresponding total counts of mergansers.

Habitat Use

Movement Patterns. Flights of common mergansers were documented 4
days/wk during 3 time periods: morning (0.5 hr before sunrise to 2.5 hr after
sunrise); midday (2.5 hr after sunrise to 2.5 hr before sunset); and evening (2.5 hr
before sunset to 0.5 hr after sunset). Three 15-min observations/time period/day
were conducted from vehicles parked at the mouth of the discharge channel.
Number of birds and flight direction (inland or toward Saginaw Bay) were
recorded. Although movement between inland habitats (river and discharge
channel) and the plume was recorded, movement within habitats was not
documented. The 3 flight observations/time period/day were pooled to estimate

total flights/time period/day.

9
Avarlabfl'ty. Based on observations during the first year of the study, the

discharge channel. the plume, and Saginaw River were identified as habitats
suitable for common mergansers during winter. However, the availability of the
plume and the river was limited by ice cover, the extent of which varied as a
function of weather conditions. Changes in the amount of open water in those 2
habitats were noted when they occurred, providing estimates of percent
availability throughout the winter.

Use. Estimates of the mean number of birds using each habitat during the
day were derived from midday counts, whereas estimates of night habitat use
were based on a combination of counts and flight observations. Although night
counts of birds in the discharge channel were considered accurate, complete
counts of mergansers in the Saginaw River often were not obtained. Therefore,
when flight estimates indicated that more birds were using inland habitats at night
than were actually counted there, the additional birds were assumed to be using
the river. Night use of the plume was estimated indirectly by subtracting the
number of birds using inland habitats at night from the total number of mergansers
in the area, as determined from midday counts.

Day and night estimates were averaged to examine the factors affecting
diel habitat use. The effects of year, season, and age-sex group on the number of
birds using each habitat were analyzed with linear model Chi-square analogues of
analysis of variance (SAS Institute Inc. 1985). When full-model analyses indicated
that interaction effects were significant, subsets of the data were used to analyze

the main effects.

10
Behavior

Instantaneous sampling was used to construct time-activity budgets.
whereas continuous observations were used to estimate duration of dives and
foraging success. Tacha at al. (1985) found that continuous observation of
individuals gives more accurate and precise estimates of percent of time spent in
various activities, especially those that occur infrequently or are of short duration,
than instantaneous methods. Continuous observation also provides reliable
information on the duration of various activities (Altmann 1974). However, scan
sampling of groups (an instantaneous method) allows observation of many more
individuals compared to the continuous method, and is therefore a more valid
basis for comparisons at the population level.

Birds were observed 4 days/wk from parked vehicles, using binoculars or a
spotting scope. Observation posts were established along the shores of the
discharge channel (10 posts). the plume (10 posts).and Saginaw River (17 posts).
Scans were only conducted when the activities of 2_ 20 birds could be reliably
ascertained, thus the distance from which birds were observed varied depending
on light and other visibility factors as well as water turbulence, particularly in the
plume. Groups of mergansers were scanned ; 10 min after the vehicle was
parked, and the behavior of each bird at the instant it was seen was recorded.
Merganser activities were assigned to the following categories: foraging
(searching for food, diving, or handling prey); resting (sleeping or inactive);
locomotion (swimming); social (agonistic or courtship behavior); comfort
(preening, bathing, or other body maintenance activities); and alert.

Diurnal behavior was observed during the same morning, midday, and

evening time periods that were established for flight observations. In 1989-90

11

and 1990-91, observations of mergansers in the discharge channel and the
Saginaw River were also conducted during 3 night time periods: 0.5 hr after
sunset to 3.5 hr after sunset: 2300 hrs - 0200 hrs: and 3.5 hr before sunrise - 0.5
hr before sunrise. Birds in the plume were not visible from the shore at night.
preventing observation of night behavior in that habitat.

To minimize the loss of independence among samples, data used in
analyses were restricted to 1 scan/observation post/time period/day. For each
scan, the proportion of birds engaged in each activity was calculated separately
for adult males and other birds. After proportions of 0 and 1 were adjusted as
recommended by Snedecor and Cochran (1980). they were converted to angles
using the arc sine square root transformation. The transformed data were then
weighted by the sample size (n) for each scan (Snedecor and Cochran 1980) prior
to analysis. Univariate and multivariate analysis of variance were used to examine
the effects of year, season, age/sex group, time of day, and habitat on occurrence
of the various activities. As with analysis of habitat use. when full-model analyses
indicated that interaction effects were significant, subsets of the data were used
to analyze main effects. The Statistical Analysis System (SAS Institute Inc. 1985)
was used to conduct the analyses.

When groups of undisturbed mergansers were observed from _<_ 100 m, a
single bird was randomly selected for continuous observation after the scan. The
duration of each of the subject's activities was recorded for 10 (1988-89) or 15
(1989-90 and 1990-91) min. Birds that flew out of sight were assumed to be
flying for the remainder of the observation period, but when a subject swam out

of sight the observation was aborted. An observation was also aborted if the

12

subject was ”lost" among a group of birds during the observation period.
Continuous observations were not conducted at night.
Foraging

In 1990-91, dip netting was used to provide an index of the relative
abundance of forage fish in the discharge channel over time. Species and length
were recorded for fish caught during 10 dipping attempts/day at each of 2 fixed
locations along the channel. Sampling was conducted at midday, 4 days/wk.
Comparisons with the previous 2 winters are based on casual observations of
schools of forage fish seen at the surface of the water in the discharge channel.
Availability of forage fish in other habitats was not estimated.

Percent of time spent diving, duration of dives, and foraging success rates
were estimated from data collected during continuous behavioral observations.
Chi-square tests were used to compare foraging effort and success rates of adult
males versus other birds. Crops from carcasses collected during the 1988-89 and
1989-90 transmission line mortality study (see Part II below) were used to
determine food habits of migrating and wintering common mergansers. The
esophagus and proventriculus of each bird was removed and preserved individually
in diluted 2-propanol for later analysis. Crop contents were weighed and prey

items were counted and, when possible, identified to species and measured.

RESULTS
Population Trends

Cormts. Common mergansers began arriving in lower Saginaw Bay in
November (Figure 2). Most of these early arrivals were observed in the plume and

on the Saginaw River. In 1988 and 1989, mergansers continued to move into the

 

 

    

No. birds % ice
tzooo —
10.000
3.000
6,”
4,000
zooo
0
‘ 100
12'“ _ ......................... .
1989-90
10.000 .0
8.00!) so
(3.000
40
4.000
20
2.000
0 0
....................... - 100
1 000 .
z r 1990-91
so
60
40
20
0

 

 

 

Nov 16-!) Dec1-15 00:16-81 JIn1-15 Jln16-31 Feb1-14 Feb 15-28 M 1-15 Mal 16-31
Dates

Figure 2. Maximum counts (bars) of common mergansers in relation to ice
cover (dotted lines) In lower Saglnaw Bay and the Saginaw River, 1988 - 91 .

14
area in December, with peak numbers of 12,000 (1988-89) and 11,000 (1989-

90) occurring in late December and early January. The peak number (10,000) did
not occur until January in 1990-91, probably because unusually warm weather
during late fall and early winter influenced chronology. Each year more than 50%
of the mergansers departed in mid/ to late January. At least 3,400 birds remained
throughout February, when most of the wintering birds left the area. All were
gone by the end of March.

These counts provided the basis for partitioning the winter into 4 segments
according to the chronology and status of wintering common mergansers using
the area. The segments are defined as:

1) Late fall - birds begin arriving in the area (16 November - 15 December

in 1988 and 1989 and 16 November - 31 December in 1990);

2) Early winter - peak numbers of birds present (16 December - 15

January in 1988-89 and 1989-90 and 1-31 January in 1991);
3) Midwinter - wintering birds remain in the area (16 January - 28
February in 1989 and 1990 and 1-28 February in 1991); and

4) Late winter - birds present during March.

Age and Sex Composition. Although fewer adult males used the area
during late fall and early winter in 1988-89 than in the 2 subsequent years, the
proportion of adult males among early arrivals and at peak numbers was similar
within years (Table 1). In 1988-89 and 1989-90, the proportion of adult males
increased during midwinter as most of the adult males present during peak
numbers also wintered in the area. Of those, about half remained in the area in
late winter, after most of the other wintering birds had left. The pattern was

different in 1990-91. The proportion of adult males remained the same from late

15

Table 1. Proportion and maximum number of adult males in populations of migrating and
wintering common mergansers using lower Saginaw Bay from November through March.

 

 

 

 

Pr0portion Number
Season 1 988-89 1 989-90 1 990-91 1 988-89 1989-90 1990-91
Late fall 0.09 0.17 0.15 180 1020 390
Early winter 0.06 0.13 0.1 5 720 1430 1 500
Midwinter 0.20 0.25 0.1 6 700 875 540

Late winter 0.40 0.39 0.42 400 390 420

 

16

fall through midwinter, indicating that most of the adult males left during the
general early winter exodus of mergansers from the area. However, both the
proportion and the number of adult males remaining in late winter were similar to
those of the previous 2 years.

Habitat Use

Movement Patterns. Daily movements were marked by large groups of
mergansers flying between inland habitats and the plume during the morning and
evening periods (Table 2). Three main patterns were evident: flights to inland
habitats in the evening and back out to the plume the following morning; flights to
inland habitats in the morning and back out to the plume the same evening; and
no movement betWeen inland habitats and the plume. Flights between inland
habitats and the plume occurred most frequently during late fall and early winter
(Table 3). primarily by mergansers spending the night in inland habitats. This
pattern continued through mid/ and late winter in 1989, but comparatively little
movement between habitats was observed during mid/ and late winter of 1990
and 1991.

Availability. When mergansers began arriving in late fall, all suitable
habitats in the area were ice-free and completely available. However, when
freeze-up occurred in early to midDecember in 1988 and 1989 (Figure 2), ice
cover reduced the size of the plume and virtually eliminated the river as an
available habitat during early winter. Open water in the river was restricted to
small, intermittently thawed pockets created by municipal and industrial thermal
discharges. In 1990 freeze-up occurred in late December, but pockets of
permanent open water were present in the Saginaw River throughout early winter.

Thus although availability of the river was reduced by > 95%, it was not

17

 

 

 

 

 

 

 

 

 

 

 

Zn 2 R «2 2.! 5a 83 «no a Km .3 econ can 4 F.» as»
mow «3 can m3. 0.! 8m 32 an: 3 F «.3 mun menu 9.29m
.2 not on o3 8. Z» en 3 cm Bu 2: mum >822
:: 9o 3 new no! «new we 8. new. 33 I 3" 8.522
mm m on m mm m um m mm m an m >2. 3 as...
25: Es... 25: Eat. 25: 82:.
8.32 8-82 3.82

 

Ace! - .52. .853 05.5..
.25.... >3 333cm .25. 5 08:... 8.2.0.5 xooueegéix .5 Pete“ 3 Scot: 833a... acct. Eaton .053 9.3: Bees-9.2: c8580 3 59.5: see: .« Joe...

18

 

 

 

 

 

 

 

 

 

 

 

O on o 0— or 0. or ON on 00 com com .353 8a..
ON on On 3 0m cm 00 om 0: on» 03 0mm P .853 2.2
o 3.. on: one 00a — SN 03 o 3 P om: ON _. own 00a can» ESE! 2.5m
our 0: 0: 00¢ om cm on on 0: con Dean 03* =3 8....
mm H mm m mm mm mm m mm m mm m 30> .0 E2.
083$ 9.2:. 0:5... acct. Ear.— neat.
36am _. comma — aoéaap

 

.303. - 5oz. .353 0:22. 9.23. 05 c.
.eEsrz >3 3235 530. c. 2.51 8.33% sooeeeBEex 2: Bake. 8 9:02: 8852.. acct. Bake. 35% 05>: Ease-9!: c9533 3 59:2 cos! .n 038.

19

effectively eliminated as during the previous 2 years. Although the size of the
plume was reduced during early winter, 2 200 ha of open water was present each
year.

Availability of the river and plume habitats increased as the result of a
midwinter thaw in late January each year (Figure 2). The Saginaw River was ice-
free for at least 10 days during late January and early February each year, as was
the entire plume and most of the bay that was visible from the shoreline. When
cold temperatures returned, Open water habitat in the bay was again confined to
the plume, which was reduced in size until the spring thaw in late March. In the
Saginaw River, however, pockets of permanent open water at thermal discharge
sites persisted throughout the remainder of the winter during all 3 years.

Use. During late fall, when all habitats were completely available, > 50%
of the mergansers in the area used both inland habitats and the plume in 1988-89
and 1990-91. Equal numbers of birds used the discharge channel and the
Saginaw River ()F=0.69, 1df, P=0.406) (Table 4), indicating that the 2 inland
habitats were equally preferred. In 1989-90, > 20,000 herring (Larus argentatus)
and ring-billed gulls (L. delawarensis) appeared in the study area in early
November, apparently in response to unusually large and/or accessible
aggregations of gizzard shad fry. The gulls exploited this food source for nearly 2
months before leaving the area in late December. The constant presence of
thousands of gulls in the air and on the water at inland habitats seemed to disrupt
the normal movement patterns of mergansers; of the mergansers that were
observed attempting to land in the discharge channel in late fall, most were
unsuccessful and returned to the plume. Thus most birds spent nearly all their

time in the plume until early winter.

 

20

Table 4. Mean number/day of common mergansers using 3 habitats in lower Saginaw Bay
during winter. 1988-89 - 1990-91.

 

 

 

 

 

Discharge Discharge Saginaw
channel plume River
Year and
season No. birds % No. birds % No. birds %
1 988-89
Late fall 500 25 1000 50 500 25
Early winter 3275 27 8725 73 0 0
Midwinter 1 00 2445 70 955 27
Late winter 25 500 50 475 48
1989-90 ‘
Late fall 275 6 4200 93 25 1
Early winter 2775 25 7975 73 250 2
Midwinter 25 1 795 47 2010 52
Late winter 15 3 305 51 280 47
1 990-91
Late fall 275 1 5 1295 69 300 16
Early winter 75 1 7565 76 2360 24
Midwinter 25 1 2000 59 1 375 40
Late winter 15 2 505 63 280 35

 

21

When merganser numbers peaked in early winter, about 50% of the birds
used both inland habitats and the plume. The Saginaw River was ice-covered in
early winter the first 2 years, therefore birds using inland habitat were confined to
the discharge channel. However, in 1990-91 the Saginaw River was only partially
ice-covered during early winter, and nearly all birds using inland habitat selected
the river over the discharge channel (Table 4). In all 3 years, mergansers that
remained in the area avoided the discharge channel after early winter. Movement
between habitats decreased, and sedentary wintering birds showed a preference
for the plume over the river during both midwinter (X’=442.9, 2df, P< 0.001) and
late winter 06:34.1, 2df. P< 0.001 ).

The proportion of adult males present did not differ significantly among
habitats in either late fall (X’=1.35, 2df, P=0.509). early winter (X" = 1.78, 2df,
P=0.410), or late winter 06:043. 1df, P=0.513), indicating that habitat use
was similar for adult males and other birds during those 3 seasons (Table 5).
However, a difference ()6 =62.6, 1df, P< 0.001) in habitat use between adult
males and other birds was observed during midwinter; 23% of the mergansers
observed in the plume were adult males compared to 17% of the birds using the
river (Table 5). Thus in midwinter, adult males preferred the plume whereas one
or more of the other age-sex groups preferred the river.

Behavior

Four hundred three scans of common mergansers in all habitats (Table 6)
were used to estimate the percent of time birds spent in each of the 6 behavioral
categories. Because most attempts to scan birds at night failed due to high

densities of very active birds and/or poor lighting, few observations of nocturnal

22

Table 5. Proportion of adult males (P) in groups of common mergansers observed in 3 habitats
in lower Saginaw Bay from November - March, 1988-91. Sample size (n) represents total

common mergansers identified during multiple independent observations.

 

Discharge channel

Discharge plume

 

 

Saginaw River

 

 

Time of year P n P n P n

Late fall 0.12 574 0.15 3118 0.15 909
Early winter 0.09 4773 0.11 6272 0.13 4833
Midwinter 0.28 153 0.23 4384 0.17 7127
Late winter 0213 46 0.40 1970 0.41 810

 

23

Table 6. Number of behavioral observations (scans) of fall migrating and wintering
adult male (AM) and adult female and immature (Other) common mergansers in lower

Saginaw Bay, Dec. 1988 - Mar. 1991 .

 

Season and habitat

Number of birds observed

 

 

Scans AM Other Total
Late fall
Channel 10 60 640 700
Plume 24 301 1511 1812
River 1 1 147 667 814
Total 45 508 281 8 3326
Early winter
Channel 54 1 64 2249 241 3
Plume 50 809 4401 5210
River 34 668 4004 4672
Total 138 1641 10654 , 12295
Midwinter
Channel 24 80 699 779
Plume 53 1 270 2968 4238
River 87 1 249 7090 8339
Total 164 2599 10757 13356
Late winter
Channel 6 1 7 1 73 190
Plume 33 698 1074 1 772
River 17 412 598 1010
Total 56 1 127 1845 2972
Total 403 5875 26074 31949

 

24

behavior were quantified. Therefore only diurnal time-activity budgets were
estimated; observations of night activities are presented qualitatively.

Daily and Seasonal/lotivr‘ty Patterns. During the day, mergansers were
most active (foraging and locomotion) during the morning and least active (resting)
at midday (Figure 3). Much of the locomotory activity consisted of swimming
between dives, and was therefore associated with foraging. Diurnal activity
patterns of adult males were similar to those of adult females and immatures
(Age/sex x time of day interaction effect; F =2.81 , 2df, P=0.061 for social
activity; F g 0.92, 2df, P; 0.397 for all other activities).

Time-activity budgets also varied among seasons (Figure 4). Foraging
activity was constant during late fall and early winter, dropped off significantly in
midwinter (F =9.93, 3df, P=0.0001). and returned to its previous level in late
winter. An inverse relationship between resting and locomotion was apparent;
mergansers spent more time resting (F =8.34, 3df, P=0.0001) and less time in
locomotory activities (F =9.06, 3df, P=0.0001) during early and midwinter than in
late fall and late winter. Seasonal variations in time allocated to comfort, social,
and alert behavior were not consistent from year to year.

The percent of time birds engaged in foraging, resting, locomotory, and
comfort activities did not differ between adult males and other birds in late fall and
early winter (Table 7). but the patterns of behavior diverged thereafter. In both
mid/ and late winter, adult males devoted more time to locomotion and social
behavior and less time to foraging than did other birds (Figure 5, Table 7). Resting
time was similar for all birds in midwinter, but adult males spent less time resting
than other birds in late winter (Figure 5, Table 7). Much of the mid/ and late

winter increase in locomotory and social activity by adult males was due to

25

 

 

 

 

 

    

Percent Time
60
50 -
- Morning
[:1 Midday
40 -
Evening
30 ~—
20 —
10 ~
.
O _...... -M
Rest Locomotion Comfort Social Alert

 

Activity

Figure 3. Diurnal time-activity budgets of common mergansers in lower
Saginaw Bay during winter (November - March), 1988-89 - 1990-91.

 

 

26

 

 

 

 
 
 
  

  
 
   
   
  
  
  
  
   
  
 
   

   
 
 
 
 
 
 
  
 
  

 
 
  
 
  
  
  
  
 
 
  
 
 
    

    
  

      

 
     
   
   
  
   
  
  
  
  
  
  
  
  
  

  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
    

  

 

 

 

Percent Time
60
50 —
. ‘ I Late fall
_ :e
90‘ -
0.4 @ Early wrnter
90‘
4° - 90‘
g0: [:| Midwmter
> 0
— to:
:03 Late winter
30 — M ‘
VO‘
. ’0‘
90‘
90‘
00
2° — :e
90‘
P0‘
- T... V V
90‘ ’0’1
00 >0
0% 90‘
>0 00
00 >0
>0 .4
00 >0
0% 90‘
>0 01
00 >0
>0 00
90‘ ’0‘4
00 >0 .
’090 30‘. ==>:-.-:=- -_ _-m tart-w
Rest Locomotion Comfort Social Alert
Activity

 

Figure 4. Seasonal time-activity budgets of migrating and wintering common
mergansers in lower Saginaw Bay during winter (November - March), 1988-89 -
1990-91.

27

 

 

 

 

 

 

 

 

000.0 00.0 000.0 0.0 000.0 00.0 000.0 00.0 000.0 00.0 000.0 8.0 . 10.2
000.0 00.. 000.0 00.0 000.0 00.0 000.0 .0.0 .000 00.. 000.0 00.0 . 5.00:
000.0 00.0 .00.0v 00.00 000.0 0... 000.0 00.0. 000.0 00.0 000.0 00.0 . 08.02
have?» Duo.—
0000 0... .00.0v 00... 0.0.0 00.0 000.0 00.0 000.0 00.0 000.0 00.0 . 1.32
000.0 00.0 .00.0v .000 000.0 00.0 000.0 00.0 000.0 0.0 .00.0v 00.0. . 00.30:
000.0 00.0. .00.0v 00.00. 000.0 00.0 .000 00... 00.0 «0.0 0.0.0 00.0 . 3.8.8.0
.8532:
000.0 00.0 000.0 .00 4.00.0 00.0 2.0.0 00.0 000.0 00.0 000.0 8.0 0 0.0.2
000.0 .0.0 .000 00.0 0.0.0 00.0 000.0 8.0 000.0 00.. 000.0 00.. 0 .330...
000.0 00.0. .00.0v 00.00 000.0 00.0 .000 0.... 000.0 00.0 000.0 00.0 . 000.03
.353 >000
000.0 00.0 0.0.0 00.0 000.0 00.0 000.0 00.0 000.0 0.0 000.0 0.0 0 1.32
2.0 00.0 000.0 00.0 000.0 00.0 000.0 00.. 000.0 00.. 000.0 00.0 0 .830:
.00.0v 00.00 .00.0v 00.0. 000.0 00.0 000.0 00.0 000.0 00.0 000.0 00.0 . 08.00..
.383
a m a .4 a m m .0 a .4 e. u .0 0 “Hum”
:02 1.80 00.500 502.08.. .000 0020“.

 

.f x 02. 3330.. £53. 09.3 .050 0:0 00.0... :30 0003000 0:0 .8830...
9.0....- 623 .050 0:0 00.0.... 0.2.0 000300.. 2000090.: 005500 00 003200 0. 00002050 .0. 9500» 0008.2. .0 0002000 5 8.302 03130 0:0 5 090930 000.. .s. 030k

28

 

 

 

 

 

 

 

 

 

MIDWINTER
Percent Time
60
50 .
I AM
40 [3 Others
30
20
10
0 _-=._:1_
Forage Rest Locomotion Comfort Social Alert
Activity
, LATE WINTER
Percent Time
60
50 —
o I AM
4 H [:1 Others
30 L
20 —
iL—fi

 

 

 

 

 

 

 

 

 

 

Forage Rest Locomotion Comfort Social Alert
Activity
Figure 5. Time-activity budgets of adult male (AM) versus adult female and

immature (Others) common mergansers in lower Saginaw Bay during mid/ and
late winter, 1989 - 1991.

29

participation in courtship parties, some of which were observed each year by
midFebruary.

In late fall and early winter 1988-89 and early winter 1989-90, thousands
of birds using the discharge channel at night periodically engaged in what can best
be described as feeding frenzies. These episodes were observed only during the
first 1-2 hours after nightfall, and may have occurred in other habitats as well.
Adult males and other birds appeared to be equally active, and many of the birds
seen diving exhibited food handling postures upon reemerging at the water
surface. These late fall and early winter feeding bouts did not occur in 1990-91.
During all 3 years, nocturnal foraging activity was rarely seen in mid/ and late
winter.

Habitat Effects. Time-activity budgets of mergansers were similar among
habitats during late fall and early winter (Table 7, habitat main effect), and
behavior of adult males did not differ from that of other birds in any habitat during
those seasons (Table 7, age/sex x habitat interaction effect). In midwinter, both
adult males and other birds spent more time foraging and alert in the Saginaw
River than in the discharge plume (Figure 6, Table 7). The midwinter increase in
social activity by adult males occurred mainly in the plume (Figure 6). as courtship
parties of up to 6 adult males vied for the attention of a single adult female.
Differences in late winter time-activity budgets between adult males and other
birds were not attributable to habitat effects (Table 7). However, by late winter
many of the adults had paired, and most pairs were observed in the river. Thus
most of the social activity in the river involved pairs, whereas social behavior in
the plume primarily consisted of interactions between members of courtship

parties.

 

Q Others in river

0|
9
.N
r
.m
M
A

I AM in plume
[3 Others in plume

 

3O

Wmnmnwwuwuw«uwnwumuwnmnwuwnwuwm"WWW.

«04.40104940404040400049
50.900.000.000...00990000090900

 

 

 

   

              

wnmnwnwnwnwnwnwnm

090900400.

  

   

    

 

Percent Time

500
4o—
30—-

 

Alert

Social

Comfort
Activity

Locomotion

Rest

Forage
female and immature (Others) common mergansers using 2 habitats in lower

Figure 6. Midwinter time-activity budgets of adult male (AM) versus adult
Saginaw Bay, 1989 - 1991.

31
Foraging
Food Availability. Large schools of gizzard shad fry (50-80 mm long) were

frequently seen at the water surface and in the shallows of the discharge channel
in early December of both 1988 and 1989. Most of the shad apparently left the
channel in midDecember; no concentrations of fish were seen in late December or
thereafter. Schools of shad fry were not seen in the channel in 1990-91.
suggesting that fewer were present than in the previous 2 years. Despite this, 79
fish, 58 (73%) of which were gizzard shad fry, were caught during 32 days of dip
net sampling from 11 December 1990 - 28 February 1991. During that period,
forage fish were most abundant in midDecember, and none were caught after 3
January (Figure 7).

Foraging Effort and Success. Continuous observations of 294 mergansers
indicated that the percent of time they devoted to daytime foraging activities did
not differ (P>O.25) for the entire winter between adult males (8.5%) and other
birds (9.5%). However, adult males stayed submerged longer during dives than
other birds. The mean duration of 363 dives by adult males was 14.2 sec
(SE =O.4 sec) compared to 11.1 sec (SE =O.2 sec, n=809 dives) for other birds.
The maximum observed time submerged was 55 sec for adult males and 40 sec
for other birds. Despite this apparent advantage, adult males did not have greater
foraging success than other birds (X2 =O.23, P>O.50); adult males captured prey
during 18 of 363 observed dives (5.0% success rate) compared to 35 captures in
809 dives (4.3%) for other birds. Daytime success rates were higher (X‘ = 22.07,
P< 0.001) for birds foraging in the Saginaw River (1 1.2%) than for those foraging

in the plume (3.0%) or the discharge channel (3.5%).

32

 

Fish/Day

 

 

 

 

13-18 19-24 25-30 31-4 5-10 11-16 17-22 23-28 29-3 4-9
December January February

Dates

Figure 7. Number of forage fish caught using dip nets in the Karn-Weadock
(Essexville, Michigan) power plant's discharge channel during latede and
early winter, 1990-91.

33
Food Habits. Crops of 146 mergansers collected in late fall and early

winter were examined; 84 (58%) contained some fish (Ra-25.0 g, SE =2.8 g)
and the remaining 62 (42%) were empty. Six of 9 (67%) adult male crops
examined were empty compared to 56 of 137 (41%) for other birds. Although
prey species included alewife (Alosa pseudoharengus), yellow perch (Perca
flavescens), and shiners (Notropis spp.). 270 (94%) of 287 identifiable fish in the

crops were gizzard shad, > 90% of which were young-of-the-year.

DISCUSSION

Lower Saginaw Bay at the mouth of the Saginaw River has become an
important winter habitat for common mergansers in recent years as the birds have
responded to gizzard shad fry concentrations that occur there annually in late fall.
During this study, migrating mergansers remained in the area for up to 6 weeks,
spending much of their time engaged in foraging activities. Birds apparently left
the area only after the shad dispersed in Saginaw Bay or became inaccessible
when they moved under the ice pack. Thus the duration of the stopover seemed
to be based solely on food availability.

Common mergansers are primarily sight hunters (Sioberg 1987) and
therefore probably spend little time foraging at night. However, the Karn-Weadock
discharge channel provided an ideal nocturnal feeding site because it is shallow,
confined, and partially illuminated by street lamps and other lights around the
power plants. When gizzard shad aggregations moved into the channel, the
available light combined with an extreme density of prey apparently made the
channel a preferred night feeding site. Mergansers using the discharge channel

had better foraging success than migrating and wintering common mergansers

34
collected by Timken and Anderson (1969) during fall and winter. They examined

165 crops, 24% of which contained fish, that were removed from birds shot in
Nebraska, South Dakota, and Minnesota; at the Karn-Weadock site, 58% of the
mergansers killed by collisions with transmission lines near the channel had fish in
their crops. Although the collection methods differ, both samples were taken from
low-flying birds killed near feeding areas. Comparison with Timken and
Anderson's (1969) results suggests that the discharge channel was a very high
quality feeding site during late fall and early winter.

There were no significant differences in habitat selection, behavior, or
foraging success between adult males and other mergansers during late fall and
early winter. This suggests that adult males did not exclude other birds from
optimum habitats, nor did they disrupt the foraging of other birds, either through
aggression or competition.

Mergansers that remained in lower Saginaw Bay in midwinter shifted their
emphasis from moving between habitats and feeding to energy conservation,
spending less time foraging and more time resting than they did in late fall and
early winter. There was no evidence of the large aggregations of gizzard shad fry
seen in late fall and early winter, thus foraging effort was apparently directly
influenced by food availability. Again, there were few differences in habitat
selection, behavior, and foraging success between adult males and others. Adult
males exhibited a greater preference for the plume than other birds, but spent
most of their time resting there. Contrary to the prediction of the behavioral
dominance hypothesis, they did not displace other birds from preferred feeding
habitats or otherwise outcompete other birds for food resources. However, adult

males did spend less time in foraging activities than other birds. This indicates

35

that adult males needed less food for maintenance than adult females and
immatures. as predicted by the fasting endurance hypothesis.

In late winter, foraging effort by all mergansers increased to levels equal to
those observed in late fall and early winter. Differences in behavior between adult
males and other birds were a result of increased courtship activity among adults.
Most of the courtship activity by adult males occurred in the plume, indicating that
unpaired females preferred open space, perhaps to facilitate their selection of
mates. In contrast, paired adults preferred more secluded areas available in the
Saginaw River.

Available habitats apparently provided all the requirements of wintering
common mergansers: abundant and accessible food in late fall and early winter
(primarily in the discharge channel), adequate food in mid/ and late winter (the
plume and Saginaw River), secure resting habitat (primarily the plume), space for
courtship in mid/ and late winter (the plume) and seclusion for pairs (Saginaw
River). Despite this, each year most birds left the lower Saginaw Bay area in mid/
to late January. In 1988-89 and 1989-90, the proportion of adult males in the
midwinter population increased 2- to 3-fold as more than half of the adult males
stayed in the area while most of the other birds presumably resumed their
migration. There was no evidence in support of the behavioral dominance
hypothesis as an explanation for this, either before or after migrating birds left the
area. However, in 1990-91 the age and sex composition of the migrating and
wintering populations were the same, indicating that the fasting endurance
hypothesis alone is also inadequate for predicting annual winter distributions of

mergansers.

36

During the first 2 winters of the study, most birds arrived in the area when
their food source (gizzard shad fry) was most abundant and accessible. This was
not the case in the third winter; shad fry aggregations were much reduced from
the previous 2 years, and most birds arrived in the area after the shad had
dispersed in the bay. Despite this, the size of the wintering population of
mergansers was similar to those in 1988-89 and 1989-90. This suggests that the
condition of birds upon arrival had some influence on whether they remained in
the area throughout the winter.

Anderson et al. (1974) observed that transient common mergansers spent
little time feeding during brief stops en route to wintering grounds. Presumably
the condition of each bird on arrival in lower Saginaw Bay was a function of the
distance it traveled to get there. Although most of the breeding range of common
mergansers lies well north of the Great Lakes (Bellrose 1980), numerous pairs
breed in northern Michigan (Brewer et al. 1991) and throughout the northern Great
Lakes region (Bellrose 1980). Those southern-breeding birds and their offspring
have less distance to travel to reach lower Saginaw Bay than birds that breed
farther north in Canada, consequently they should arrive in better condition.
Furthermore, condition may be positively related to foraging success when
thousands of birds feed in a confined area, as was often the case in the discharge
channel. If so, southern-breeding mergansers would likely be in better condition
than northern-breeding birds in midwinter, and would be better equipped to
survive the rest of the winter in lower Saginaw Bay.

The change in the proportion of adult males from early to midwinter in
1988-89 and 1989-90 suggests that in some years, the fasting endurance

hypothesis is consistent with the observed wintering strategies. Although adult

37

males were not more successful or efficient foragers than other birds, they
apparently were able to assimilate and/or conserve energy more efficiently when
food was abundant. This advantage probably resulted in rapid enhancement of
their condition during late fall and early winter relative to other birds and, coupled
with their lower energy requirement during mid/ and late winter, allowed most of
the adult males to remain throughout the winter.

However, the sex and age composition of the 1990-91 midwinter
population suggests that the fasting endurance hypothesis does not hold true for
mergansers in lower Saginaw Bay if food is scarce when they arrive in early
winter. When this occurs, the condition of a bird upon arrival may be the principal
determinant of whether it stays the entire winter or continues migration. Thus an
individual's wintering strategy may depend to a large extent on where it spent the
preceding summer.

Future research on this subject should consider migration distance as well
as energetic and behavioral aspects of wintering waterfowl, but should also focus
on quantifying the reproductive benefits (or lack thereof) of wintering in northern
latitudes. This would provide the basis for a cost-benefit analysis that could

generate testable predictions involving several hypotheses of wintering strategies.

PART II: TRANSMISSION LINE MORTALITY
INTRODUCTION

As increasing numbers of migrating and wintering common mergansers
began congregating at the Karn-Weadock site, a collision mortality problem
developed as a result of birds flying across the transmission lines that are
concentrated in the area. Numerous cases of birds colliding with transmission
lines have been recorded in the United States (Stout and Cornwell 1976, Avery et
al. 1980) and abroad (eg. Ogilvie 1967, McNeil et al. 1985). Although in some
instances many individuals are killed (99. Malcolm 1982), collision mortality is
usually not considered a major factor at the population level for most species. For
example, Stout and Cornwell (1976) reported that collisions accounted for 0.07%
of the documented nonhunting mortality of ducks in North America from 1930-
1963. However, the significance of collision mortality is greater for some
populations. Collisions have been identified as the primary cause of mortality of
fledged whooping cranes (Grus americana) (Lingle 1987) and of mute swans
(Cygnus olor) in Great Britain (Ogilvie 1967).

The probability of colliding with a transmission line may be affected by the
biology of the bird, weather, and type and location of the transmission line. The
numerous potential factors involved, variability among species and sites, and lack
of comprehensive data complicate attempts to assess the significance of collision
mortality. Proposed methods of alleviating the problem are currently in the
experimental evaluation stage (eg. Morkill and Anderson 1991). The purpose of

this study was to design a method of reducing collision mortality of common

38

39

mergansers at the Karn-Weadock site. The initial goal was to determine the
magnitude of the collision problem, the timing and locations of collisions, and the
factors that influence collision rates at the site. This information provided a basis
for designing a treatment, the testing and evaluation of which was the objective of

the study.

STUDY AREA

The Karn-Weadock property consists of about 500 ha of low, level land.
Most of the northwest half of the property is developed for power plant facilities
and a series of ash settling ponds (Figure 8). The discharge channel cuts through
the middle of the property, and the southeast half of the site consists primarily of
wetlands dominated by common reed (Phragmites australis). Gravel roads on an
extensive dike system provide access to most of the site. Most of the collision
mortality occurs at the transmission lines that extend across the ash ponds, the
discharge channel, and the wetlands (Figure 8). These lines bisect flight paths
that are used by mergansers traveling back and forth between the discharge

channel and the plume.

METHODS
Dally Movements And Flight Pattema

Preliminary observations suggested common mergansers flying to or from
the discharge channel suffered most of the collision mortality. As described
previously, birds using the channel were counted either during early morning or
evening, and at midday. Those daily counts were used to provide biweekly

estimates throughout the winter of the mean number of mergansers at risk.

4O

33¢ Busmmw 05 .0 5:2: 05 .0 >580. 9:90:00 .033 E09525. .0___>x000wv 0.00.00.03.me 02.0 .o 0591..

0000.. 00000< '00:: co_00_E0:0._._. ..... 00.00.00; I 00:00 :0< 0E0... .030... I

 

 

 

 

 

 

«EB:
3553

   

><m >><z_6<m

 

 

 

41

Flight observations were conducted near transmission lines from designated
points along the discharge channel. For all common mergansers seen crossing
transmission lines, flock size, flight direction (toward the power plants or toward
Saginaw Bay), altitude (S. or > 2 m above the lines), and flight path (over the ash
ponds. over the discharge channel, or over the wetlands) was recorded. Flight
observations were separated by time period, altitude, and flight path. Data
collected during the 3 15-min observation segments/time period/day were pooled
and used to estimate daily flight rates (flights/day) for each subcategory.

Collision Mortality

Before and after each time period, the area was searched for collision
victims. Species, sex, age, location, and estimated time of collision were recorded
for each carcass found. Documented mortality (DM) was adjusted for scavenging
bias and crippling bias to estimate total collisions (James and Haak 1979). Some
carcasses were replaced after examination to provide an estimate of scavenging
bias based on the proportion of sample carcasses removed within 48 hr. Crippling
bias was estimated based on observed collisions. Cripples were considered to be
birds that struck lines and fall, but were not found dead during the subsequent
search. Birds that struck a line and kept flying without appearing to be injured
were not considered cripples. Scavenging and crippling bias estimates (SB and
CB, respectively) were calculated using James and Haak's (1979) equations,
where

58

1 / (1 - proportion of sample carcasses removed)
and

CB 1 / (1 - proportion of crippled collision victims).

42

Estimated mortality (EM) was then calculated as
EM = DM x SB

and total collisions (TC) were estimated by
TC = EM x CB.

Collision rates (collisions/flight) were then calculated based on total
collision estimates and estimates of total flights derived from flight rates. Collision
victims that fell into the discharge channel were rarely detected and recovered
before they floated out to the bay. Therefore mortality on the channel flight path
was estimated based on the calculated collision rate for the ash pond flight path.
Factors Affecting Collision Rates

Correlation analysis was used to evaluate the relationships among weather
conditions, the number of mergansers using the discharge channel, and the
number of collisions within specific time periods. Weather conditions for the area
were obtained from the National Weather Service, Saginaw River Station, located
<1 km from the study site. Visibility, wind velocity and direction, air
temperature, and barometric pressure are recorded every 2 hr at the station.
Percent cloud cover was recorded at the study site.

Experiment

Observations during the first year of the study suggested that birds often
could not see the transmission lines as they approached them, which sometimes
resulted in collisions. Thus the goal of the experiment was to alter the flight
patterns of mergansers by forcing them to avoid the transmission lines where
most collisions occurred. However, objects could not be attached to existing lines
to increase their visibility, because of the potential for excessive additional stress

on the lines and support structures that might result from ice storms. Instead, a

43

series of 1.5-1.7-m diameter, helium-filled balloons was erected to create visible
obstacles in key areas. The balloons were tethered in a position at transmission
line height at 40-m intervals across the treatment area. Because most of the
collisions occurred as birds flew from the discharge channel toward the bay,
balloons were placed inland from the transmission lines. The treatment area
during the first year was most of the ash pond flight path and was expanded to
include the channel flight path the second year. The test was conducted in 2-wk
blocks; balloons were placed in the treatment area for a week and removed the
following week. Chi-square tests were used to evaluate the effects of the
treatment on flight paths mergansers selected, their flight altitudes, and collision

mortality.

RESULTS
Counts

In 1988 and 1989, most of the area froze up in early to midDecember
(Figure 2). The only open water remaining in the bay was the plume created by
the plant's thermal discharge, thus confining mergansers to the plume in early
winter. During late December, most of these birds visited the discharge channel
every night to feed and roost, and use of the channel was also high in early
January (Figure 9). In 1990-91, freeze up was delayed until early January, and
birds began arriving in the area about 2 weeks later than in the previous 2 winters.
As a result, early winter use of the discharge channel was reduced. Mergansers
virtually stopped using the channel during the midJanuary thaws that occurred
each winter, and did not resume using the channel after the area froze again in

early February. Their avoidance of the channel coincided with the disappearance

44

 

 

 

 

 

 

 

No. birds
1 988-89
1 989-90
1 990-91

Dates

Figure 9. Mean number/day of common mergansers using the Karn-
Weadock (Essexville, Michigan) power plant’s discharge channel at night.

45

of the large aggregations of gizzard shad that were present during late fall and
early winter.
Flight Patterns

When use of the discharge channel was highest (November-January),
mergansers flew to the channel each evening, spent the night there, and flew
back out to Saginaw Bay early the next morning. Flight intensity across
transmission lines was greatest during early morning and evening, and lower for
the midday period (Table 2). Most birds leaving the channel flew out during the
first 0.5 hr after sunrise, often in large flocks of up to 2000. The evening pattern
was different, as smaller groups of 5-50 mergansers flew to the channel from 1 hr
before sunset to 0.5 hr after sunset. Most of the low altitude flights observed in
1988-89 and 1989-90 were by birds flying out of the channel in the morning, but
a large proportion of the mergansers moving between the plume and the channel
at midday also crossed transmission lines at low altitudes (Table 8). Few low
altitude flights were observed during 1990-91. The ash pond and channel flight
paths were used most frequently in 1988-89, but use of the wetland flight path
increased in subsequent winters (Table 9). Most of the low altitude flights
occurred on the channel path, whereas mergansers rarely flew low over the ash
ponds (Table 9). Although some low altitude flights occurred over the wetlands,
mergansers were never observed flying through or under transmission lines on that
flight path.
Collision Mortality

Documented collision mortality of birds at the site included 210, 105, and
28 common mergansers in 1988-89, 1989-90, and 1990-91, respectively (Table

10). Over 95% of the mortality occurred from midNovember through midJanuary,

46

Table 8. Estimated number of low altitude (52m above transmission lines)
flights/day by common mergansers crossing transmission lines at the Karn-
Weadock power plant site during late fall and early winter (Nov. - Jan.).

 

 

 

 

 

1988-89 1989-90 1990-91
% of % of % of
Direction and total total total
time of day No. flights No. flights No. flights
Toward power plant
Morning 44 17.0 31 17.1 3 1.3
Midday 38 1 1.5 26 28.0 1 0.7
Evening 87 3.2 65 1.9 2 0.2
Total 189 5.7 122 3.3 6 0.4
Toward plume
Morning 368 23.8 445 18.9 3 0.4
Midday 57 27.6 6 2.0 2 0.4
Evening 1 1 3.6 46 8.7 2 0.2
Total 436 21.3 497 15.6 7 0.3

 

47

 

 

 

 

 

 

to m 5 88 9m 5: F 8% o. 2 93. «can .83
no 8. 3.3 a; «mm «5: .3 man com 2282.
Z 3 N? «at 3» $8 mdw Rm. 003 .255
«.0 mm. 89 md one «3: m; a? Sou 22. fi<
83:.» mm m 83% mm m 83:... mm m 58 22.“.
32 8 26— 3 30. .X.
5.88 8-38 3.83

 

.35.. - $02. .353 2.3 can =8 2w. 2:2. 2.» Eu... .033 3.0833-on 2: as»... 93599.:
5888 .5 8:9: 32... 5.3.6855 gone E NW. 2553 26. «593 one 528:2: *o .363... :32 d can...

48

 

O

0000

twp

0000 O

hm

O

0000

NPOO

O

0000

hum

VN

o 3‘
ha

O

0000

up

h:
mo

0000 0

wow

0..

an
on

"*0"

New

§°3

V

PPON

non

50

Oh—
cm—

0‘)

"00°!

NON

NM

map

.33

.25
.8283
.8ch

28 .3.

.52 in...
as»
.26

23.3
.856
28. :2

d... - .>oz

 

20.....8 tires

3:85.:

80.....8

3:83.:

3.3.3

80.....8

EttoE

BaotoE

5:. in...

836.5 3256 822.88 .8853 886.3 828.88 8868b 885.3 322.88 .8. coo-om

 

 

 

wag F 8-00m F aoéoa F

 

.332 - do": .853 one. be. 3...: 39.0)
.6... - $02. .85! >5. use .1. 8.. 055v 8.. Eel .252. soon-03.502 .5 «- 23:090.: .3588 >9 0:01.28 I... conga-8:2. 3 89.5: tee-Ewen .o— 03-...

49

primarily on the channel and ash pond flight paths. No collisions were observed
on the wetland flight path, nor were any carcasses found there.

Forty collisions were observed directly. Although in most cases it could not
be determined which line a bird struck. collisions with both conductor and ground
lines were observed. All of the observed collisions were suffered by birds flying
toward the bay. Six of 8 (75%) collisions over the ice-covered ash ponds resulted
in mortality, whereas only 9 of 32 (28%) of the birds that struck lines over the
channel and dikes were killed. Therefore different crippling bias adjustments were
used to estimate total collisions for the ash pond (C8 = 1.333) and channel
(CB = 3.556) flight paths. Of 114 carcasses that were placed under transmission
lines, 16 (14%) were removed by scavengers ($8 = 1.163). Documented mortality
and adjustments for recovery, scavenging. and crippling biases produced estimates
of 870, 527, and 134 common mergansers killed or wounded by collisions during
the 3 winters (Table 10). This amounted to about 7%. 5%, and 1% of the peak
population of common mergansers in the area in 1988-89, 1989-90, and 1990-
91, respectively.

Factors Affecting Collision Rates

Mergansers appeared to make every effort to avoid the power lines if they
could see them. The collisions happened when birds did not see the lines at all or,
more commonly, saw them too late to take effective evasive action. Adult males
and other mergansers were equally susceptible; the number of adult male collision
victims did not differ from that expected (based on the proportion of adult males
among birds observed in the discharge channel) during either late fall ()6 =0.43,
1df, P> 0.50) or early winter (X‘ =2.15, 1df, P> 0.10), when most collisions

occurred. Although most collisions occurred in the morning, collision rates

50

(collisions/flight) were highest during midday (Table 11). There were very few
collisions by birds flying into the channel in the evening. The number of
mergansers using the discharge channel was positively correlated with the number
of documented collisions, especially during the early morning (1’ =0.46, n = 58,
P< 0.001) and midday periods (r’=0.49, n=49, P<0.001). The correlation was
weaker for the evening period (r2 =0.15, n = 51, P=0.005). No consistently
significant relationships were found among weather conditions and flight patterns
or collisions.
Experiment

In 1989-90, the balloon treatment was tested for 32 days during December
and January: 12 days with balloons present in the test area and 20 days with
balloons absent. The test was terminated when mergansers stopped using the
discharge channel. The following winter, the test covered 8 weeks from late
November through midJanuary: 27 days with balloons in the expanded treatment
area and 28 days without balloons. During both test years, the proportion of
mergansers using treatment area flight paths was lower when balloons were
present (Table 12). Balloons had the most pronounced effect on flight path
selection during early morning [)(’= 155.0 (1989-90) and 102.4 (1990-91), 1df,
P< 0.001)] and midday 06:70.7 (1989-90) and 70.2 (1990-91), 1df,
P< 0.001 )1. 0f the mergansers that selected treatment area flight paths, fewer
flew at low altitudes when balloons were present (Table 13). The difference was
much greater during 1989-90 (X’=21.2, 1df, P<0.001) than in 1990-91
W =3.5, 1df, P< 0.10). Documented collision mortality in the treatment areas
was also significantly reduced when balloons were present ()0 = 7.0, 1df,

P< 0.01 ). No collision victims were found in the treatment areas during either test

 

51

 

 

 

 

 

 

95.0 3 _. .0 000.0 000.0 000.0 «00.0 03.0 c 50.0 .vN . .0 :32

500.0 5 50.0 .v 50.0 000.0 000.0 000.0 500.0 5 .00 «No.0 05:05

«50.0 00m.0 :«0 000.0 000.0 000.0 050.0 N 5.0 50.0 >322

5 5.0 m5 —.0 00 p .0 000.0 000.0 P 50.0 00 p .0 05 5.0 5 5.0 0550.2

3000. 00-0005 00-000. 5000. 00-000. 00-000. .0000. 00-0005 00-000. >3 .0 25...
.853 =< :26! - Ego“. 5.53:2. . 896302

 

+5.2 - do“: .853 3a. ace 3...: can...) .62. - $02. .853 Ewe ace =2 So. 055.. 3% 8a.:
326: sooueogécex 05 «a 8:: 5.3.8388 05820 23:38:. 5868 .o .00. x .22.. . «causes. «32 33:8 336:3 .3 can...

 

52

005 0055 5 005 00 035 cop 5..5N .vm N9: ow mmN— .905.

 

 

 

 

 

 

 

 

 

 

8, one we new no Em co, 3? mm «mm 5 NS 8.55m
8. 2h 9 m: 3 8. 8. 3K «5 5% mm Bu >825.
oo. mum on can No mam co. m 5 mm «.9 mm o .N 8.522
6-83
09 82 B on: 9 mom 2: «m3 S 88 m mt .58
8. P2. 8 van 3 8. 8. m9 3 33 n 3 223m
8. me Q m. «.5 mm 8. 08 5 v9 9 9 F 262.2
co. 8o 8 85 5 8 oo— mosv mm 53 F 3 82.22
8-32
3 85.: 8 8%..“ .x. 3.3.: .x. 35.: .x. «£2: .x. 8%.: >3 .o
2:: new 50>
9.33 =< «53 350 new ~85 «£8 =< 2:8 .050 new :5.
«coma» 2.00:3 Samoa 950.30
.22. - <62.

.35? 2.3 9... =2 8a. 8:2. «58 £9: .28 P... 33 8253: 8.»: 83:38:. 3683 S 52850... B .363: 38:53 .N. 038.

 

53

0.8. 53' 0.0 0 N00 000— 0.005 3N P «.0 N 0.8. «m2. .305

 

 

 

 

 

 

 

 

 

0.8. .00 0.0 a 0.8 $0 0.8. «00 ..0 . 0.8. .00 05:30

0.8. 00. 00 c 0.50 «0. 0.8. 08 0.0 0 0.8. 08 .68.:

0.8. 30 0.0 u 0.8 000 0.8. 0 .u 0.0 . 0.00 03 8.50.2
.080.

0.8. 00a ..0. .0 0.00 ct 0.8. 2.. N. u 0.00 2.. .30..

0.8. 50. 0.0 0 0.3 .0. 0.8. v. 0.0 0 0.8. c. 9.230

0.8. 00 5.0 u 0.3 an 0.8. 0. . 0.0 . .8 c. . .68....

0.8. no 0.00 an 0.00 3 0.8. 3 ad . 5.3 9 8.52...
8-000.
0.. .20.... 0.. 8.0.... 0.. .20.... 8 .20.... 0.. .20.... i .20... >00 .0 2....
20 30>

.80. .02.... ovaui .80» 28....- 8a...-
33 00.: 33 :0...
Eco..- ocooiu 0:820 3.00.10

 

4...... . $02. .80.! 3.00 v... =8 30. 0550 .20 €0.50... 05 £0325 202.090... .8558 >a. 82850.: 00 39.3.. $80.55 .0— 030.5

54

year when balloons were present, compared to 41 carcasses found when balloons
were absent. Outside the treatment areas, 4 collision victims were found when

balloons were present and 13 when balloons were absent.

DISCUSSION

Most collisions occurred as birds flew toward the bay. apparently because
transmission lines were less visible to birds flying in that direction. Collision rates
were highest during midday, indicating that more light did not necessarily increase
the visibility of the lines. Likewise, weather conditions had no detectable
influence on collision rates. The positive correlation between collisions and
number of mergansers using the discharge channel suggests that vulnerability to
collisions did not decrease as birds became more familiar with the area.

Mergansers often did not appear to notice the lines until they were too
close for effective evasive action. The birds frequently flared as they crossed
transmission lines, rather than during their approaches. In contrast, dabbling
ducks and gulls usually flared before reaching the lines, or reduced their flight
speeds and avoided the lines while flying through them. Thus mergansers were
especially susceptible to collisions because of their inability to detect transmission
lines from a distance combined with their lack of maneuverability during flight.

The attempt to divert flying mergansers away from the transmission lines
at the site was partially successful. Some birds responded to the presence of
balloons by increasing their approach altitudes and others moved to different flight
paths. These changes in merganser flight patterns resulted in a significant decline
in mortality. Morkill and Anderson (1991) observed similar avoidance reactions by

sandhill cranes flying toward transmission lines that had been marked by aviation

55

balls. They too documented significantly less mortality at lines that were marked
compared to unmarked lines.

There were some deficiencies in the balloon method that are not reflected
by the data. The obstruction of the channel flight path was not enough to divert
some mergansers. The reduction of low altitude flights over the treatment area
when balloons were present was much greater during the first test year. This
suggests that the balloons were more effective on the ash pond flight path than
on the channel path. Low-flying individuals and small flocks that ignored the
balloons and flew through transmission lines over the channel were observed
several times. Thus some undetected collisions probably did occur over the
channel while balloons were present during the last winter of the study. This
problem could probably be corrected by increasing the density and varying the
heights of obstacles along the channel flight path.

Although balloons were useful because they could be set up and removed
at will, they are not practical as a long-term solution to the collision problem.
Balloons are not durable enough to withstand stress produced by high winds. The
average life span of a balloon during this study was <4 days, and replacing
balloons was costly and time consuming. However, a seasonally (November-
January) permanent series of similar obstacles, placed inland from the
transmission lines that bisect the ash ponds and discharge channel, would
probably be equally effective in reducing collisions at the site.

Beaulaurier (1981) suggested that attaching highly visible markers to
transmission lines where numerous collisions occur is the most cost-effective
technique for reducing collision mortality. The method does produce the desired

result, as demonstrated by Morkill and Anderson (1991). However, in some

56

circumstances safety and/or economic considerations preclude direct attachment
of objects to the lines. In this study, placement of markers near transmission lines
provided an "early warning system“ that obtained results that compared favorably
to those of Morkill and Anderson (1991). Although this method presents more
logistic problems and is more costly than attaching objects to transmission lines, it
is a viable alternative that should be considered in cases where the transmission

lines can not be modified.

BIBLIOGRAPHY

BIBLIOGRAPHY

Altmann, J. 1974. Observational study of behavior: sampling methods.
Behaviour 49:227-267.

Anderson, 8. W., M. G. Reader, and R. L. Tlmken. 1974. Notes on the feeding
behavior of the common merganser (Mergus merganser). Condor 76:472-
476.

, and R. L. Timken. 1972. Sex and age ratios and weights of common
mergansers. J. Wildl. Manage. 36:1127-1133.

Avery, M. L., P. F. Springer, and N. S. Dailey. 1980. Avian mortality at man-
made structures: an annotated bibliography (revised). U.S. Fish Wildl.
Serv. FWS/OBS-80/54. 152pp.

Beaulaurier, D. L. 1981. Mitigation of bird collisions with transmission lines.
Bonneville Power Admin., Portland, OR. 82pp.

Bellrose, F. C. 1980. Ducks, geese, and swans of North America. Third ed.
Stackpole Books, Harrisburg, PA. 540pp.

_. and R. D. Crompton. 1970. Migrational behavior of mallards and black
ducks as determined from banding. lll. Nat. Hist. Surv. Bull. 30:167-234.

Brewer, R., G. A. McPeek, and R. J. Adams, Jr. 1991. The atlas of breeding
birds of Michigan. Michigan State Univ. Press, East Lansing, MI. 594pp.

Cook, C. L. 1990. Habitat selection by overwintering Canada geese in south-
central Michigan. MS. Thesis, Michigan State Univ., East Lansing. 57pp.

Cooper, J. A., and M. A. Johnson. 1977. Wintering waterfowl in the twin cities.
Loon 49:121-138.

Diefenbach, D. R., J. D. Nichols, and J. E. Hines. 1988. Distribution patterns of

American black duck and mallard winter band recoveries. J. Wildl.
Manage. 52:704-710.

57

58

Gray, 8. T., D. H. Gordon, and R. M. Kaminski. 1987. Functional attributes of
coastal wetlands for waterfowl: perspectives for research and
management. Pages 205-222 in W. R. Whitman and W. H. Meredith, eds.
Waterfowl and wetlands symposium: proceedings of a symposium on
waterfowl and wetlands management in the coastal zone of the Atlantic
flyway. Delaware Coastal Management Program, Delaware Dep. Nat. Res.
Environ. Control, Dover, DE.

Hanson, H. C. 1965. The giant Canada goose. Southern Illinois Univ. Press,
Carbondale. 226pp.

Haymes, G. T., and R. W. Sheehan. 1982. Winter waterfowl around Pickering
Nuclear Generating Station. Can. Field-Nat. 96:172-175.

James, B. W., and B. A. Haak. 1979. Factors affecting avian flight behavior and
collision mortality at transmission lines. Bonneville Power Admin., U.S.
Dept. of EnergY. Portland. OR. 109pp.

Jorde, D. G., G. L. Krapu, R. D. Crawford, and M. A. Hay. 1984. Effects of
weather on habitat selection and behavior of mallards wintering in
Nebraska. Condor 86:258-265.

Ketterson, E. D., and V. Nolan, Jr. 1976. Geographic variation and its climatic
correlates in the sex ratio of eastern wintering dark-eyed juncos (Junco
hyemalis hyemalis). Ecology 57:679-693.

, and . 1979. Seasonal, annual, and geographic variation in sex ratio of
wintering populations of dark-eyed juncos (Junco hyemalis). Auk 96:532-
536.

 

Lingle, G. R. 1987. Status of whooping crane migration habitat within the Great
Plains of North America. Pages 331-340 in J. C. Lewis, ed. Proc. 1985
Crane Workshop, Grand Island, NE.

Longcore, J. R., and J. P. Gibbs. 1988. Distribution and numbers of American
black ducks along the Maine coast during the severe winter of 1980-1981.
Pages 377-389 in M. W. Weller, ed. Waterfowl in winter. Univ. Minnesota
Press, Minneapolis, MN.

Malcolm, J. M. 1982. Bird collisions with a power transmission line and their
relation to botulism at a Montana wetland. Wildl. Soc. Bull. 10:297-304.

McNeil, R., J. R. Rodriguez 8., and H. Ouellet. 1985. Bird mortality at a power
transmission line in northeastern Venezuela. Biol. Conserv. 31:153-165.

Morkill, A. E.. and S. H. Anderson. 1991. Effectiveness of marking powerlines to
reduce sandhill crane collisions. Wildl. Soc. Bull. 19:442-449.

Nichols, J. D., and G. M. Haramis. 1980. Sex-specific differences in winter
distribution patterns of canvasbacks. Condor 82:406-416.

59

_, K. J. Reinecke, and J. E. Hines. 1983. Factors affecting the distribution of
mallards wintering in the Mississippi Alluvial Valley. Auk 100:932-946.

Ogilvie, M. A. 1967. Population changes and mortality of the mute swan in
Britain. Wildfowl Trust 18:64-73.

Perry, M. C., J. D. Nichols, M. J. Conroy, H. H. Obrecht Ill, and B. K. Williams.
1988. Sex specificity of behavioral dominance and fasting endurance in
wintering canvasbacks: experimental results. Pages 103-121 in M. W.
Weller, ed. Waterfowl in winter. Univ. Minnesota Press, Minneapolis, MN.

Pope, G. F. 1976. Fisheries survey of Saginaw Bay related to the thermal effects
of the Karn-Weadock generating stations. Beak Consultants, Inc.,
Mississauga. Ontario. 49pp.

Read. A., ed. 1986. Eider ducks in Canada. Canadian Wildl. Serv., Rep. Ser. 47.

_. and A. Bourget. 1977. Distribution and abundance of waterfowl wintering
in southern Quebec. Can. Field-Nat. 91 :1-7.

Salyer, J. C., II, and K. F. Lagler. 1940. The food habits of the American
merganser during winter in Michigan, considered in relation to fish
management. J. Wildl. Manage. 4:186-219.

SAS Institute Inc. 1985. SAS user's guide: statistics, version 6 edition. Cary,
NC. 1028pp.

Sayler, R. D., and A. D. Afton. 1981. Ecological aspects of common goldeneyes
Bucephala clangula wintering on the upper Mississippi River. Ornis Scand.
1 2:99-108.

Serie, J. R., D. L. Trauger, and D. E. Sharp. 1983. Migration and winter
distributions of canvasbacks staging on the upper Mississippi River. J.
Wildl. Manage. 47:741-753.

Sjoberg, K. 1987. Food selection, food-seeking patterns and hunting success of
captive Goosanders Mergus merganser and Red-breasted Mergansers M.
serrator in relation to the behaviour of their prey. Ibis 130:79-93.

Snedecor, G. W., and W. G. Cochran. 1980. Statistical methods. Seventh ed.
Iowa State Press, Ames, IA. 507pp.

Stott, R. S., and D. P. Olson. 1973. Food-habitat relationship of sea ducks on
the New Hampshire coastline. Ecology 542996-1007.

Stout, l. J., and G. W. Cornwall. 1976. Nonhunting mortality of fledged North
American waterfowl. J. Wildl. Manage. 40:681-693.

Sugden, L. G., W. J. Thurlow, R. D. Harris, and K. Vermeer. 1974. Investigations
of mallards overwintering at Calgary, Alberta. Can. Field-Nat. 88:303-31 1.

6O

Tacha. T. C., P. A. Vohs, and G. C. lverson. 1985. A comparison of interval and
continuous sampling methods for behavioral observations. J. Field
Ornithol. 56:258-264.

Thompson, B. C., J. E. Taber, and C. L. Turner. 1988. Diurnal behavior patterns
of waterfowl wintering on the Columbia River, Oregon and Washington.
Pages 153-167 in M. W. Weller, ed. Waterfowl in winter. Univ. Minnesota
Press, Minneapolis, MN.

Timken, R. L., and B. W. Anderson. 1969. Food habits of common mergansers in
the Northcentral United States. J. Wildl. Manage. 33:87-91 .

White, D. H., and D. James. 1978. Differential use of freshwater environments
by wintering waterfowl of coastal Texas. Wilson Bull. 90:99-111.

MICHIGAN STRTE UNIV. LIBRARIES
lllWWllllllllll”ll”lll“NWllllllllllllllmlll
31293009141098