.mpnn ..--...

         

 

 

 

   

 

   

 

,
‘ » _
. , :
‘ ',.A V ,_
I < .- , ' ~
' ‘0 .
I ' '3 I."» ‘ - _ . ‘ . .‘ ' .
I ‘ ‘ x ' . - - ‘ ‘ _ .
‘ . ' ... . ‘ '7 ‘ -“ A. _ , ,

 

   

 

 

 

 

‘.
. . f ‘ _ _ . .
' -. ‘ .. j _ .-¢- _ ‘ ' ' , '
. -, ~ ‘ , - ‘ ,
. V
4
l I
‘ ‘ X ‘ ‘ g .
‘ , .
x‘ '5‘ V
’ . .
l j l v
> ‘ r - ‘ “ a. V
‘ , ‘ ‘ _. }
‘ r , . .m >
‘ ‘ ‘j , ‘ H ‘u
H ‘ ‘_ Ll.“ .‘uun “dun.“ ..’»v..-£"ut

 

      
   

 
  

lllllllllllHHIIHJIHHHIHIlllllllllllIll!llllllNllllllll
1 31293017804802 .-

 

LIBRARY
Michigan State
Unlverslty

 

 

 

This is to certify that the

dissertation entitled

Assessment of Common Tern (Sterna hirundo)
Reproductive Success in Saginaw Bay
of Eastern Michigan

 

presented by

Kelly F. Millenbah

has been accepted towards fulfillment
of the requirements for

Doctor of Philosophy degreein Fish. & Wildl.

 

 

553%. UM

Major professor

Date ’2‘16‘ 97

 

MS U is an Affirmative Action/Equal Opportunity Institution 0-1277!

 

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

 

DATE DUE DATE DUE DATE DUE

 

i 9,
GJUEL at :022001

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ASSESSMENT OF COMMON TERN (ST ERNA HIRUNDO)
REPRODUCTIVE SUCCESS IN
SAGINAW BAY OF EASTERN MICHIGAN

By

Kelly F. Millenbah

A DISSERTATION

Submitted to
Michigan State University
in partial fi11filh'nent of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Fisheries and Wildlife

1997

ABSTRACT

ASSESSMENT OF COMMON TERN (ST ERNA HIR UNDO)
REPRODUCTIVE SUCCESS IN
SAGINAW BAY OF EASTERN MICHIGAN

By

Kelly F. Millenbah

Once common and widespread in the Great Lakes region, common tern (Sterna
hirundo) numbers in Michigan are currently estimated to be around 1,400 breeding pairs.
Reproductive success of common terns in the Great Lakes is below that on the eastern
seaboard suggesting that their continued existence in the Great Lakes region may be
severely jeopardized. Two common tern colonies were observed in Saginaw Bay of
eastern Michigan in the summers of 1995 - 1997 to determine impacts to reproductive
success using a monitoring program that included observations, trapping and banding,
and nest checks. Population projection models were developed to determine the long—
term survival of the species at these sites. The effectiveness of using nesting platforms to
increase reproductive success also was evaluated. Results from this study indicate that
the common tern population is declining. Major threats to survival include wave
inundation and predation. To reverse this declining trend it will be necessary to increase
the survival of common tern eggs and chicks, the primary variables that can be impacted
in Michigan. Recommendations for increasing reproductive success include predator

control efforts, use of nesting platforms, and manipulation of habitat.

ACKNOWLEDGMENTS

Funding for this project was provided by the US. Fish and Wildlife Service
(U SF WS), East Lansing Field Office; USFWS, Twin Cities Regional Office, Non-game
Conservation Project; Michigan State University, Department of Fisheries and Wildlife;
and George J. Wallace and Martha C. Wallace Endowed Scholarship. I am indebted to
my committee members and thank them for their knowledge and continued support of
this project: Dr. Scott Winterstein, major advisor, Dr. Henry Campa, III, Dr. Donald
Beaver, Dr. Frank Fear, and Dr. Lisa Williams. Additionally, I extend my deepest
gratitude to Dr. William Taylor and Charlie Wooley who have provided invaluable
mentoring, support, and friendship throughout this project.

This project would not have been completed without the field assistance of Betsy
S. Cook and Eric J. “Let’s go play Greenpeace with the fishing boats” Baka. Others who
have contributed to this project include Dave Best, Melissa Harrison, Hadiya White,
Jacquie Stout, Jen Beaman, Mike Shaw, Mark Johnson, Helen Hayes, Gordon Williams,
Mike Weaver, and the Bayport Fish Company. A very special thank you is extended to
my dad, Richard L. Millenbah, who took countless hours off of work and traveled many
miles to build the nesting platforms used in this project. Drawings in this dissertation
were provided by Chris Fetters of IM Graphics.

iii

TABLE OF CONTENTS

LIST OF TABLES ...................................................... vi
LIST OF FIGURES ..................................................... ix
INTRODUCTION ....................................................... l
OBJECTIVES ......................................................... 10
STUDY AREA ........................................................ 11

CHAPTER I: POPULATION ASSESSMENT AND REPRODUCTIVE STATUS OF
THE COMMON TERN IN SAGINAW BAY

INTRODUCTION ................................................ 19
OBJECTIVES ................................................... 22
METHODSn .................................................... 23
General Observations ........................................ 23
Predation ................................................. 24

Site Characteristics .......................................... 26
Assessing Reproductive Success and Determining Survival Estimates . 27
Population Projections ...................................... 29
RESULTS ...................................................... 33
General Observations ............... - ......................... 33
Predation ................................................. 44

Site Characteristics .......................................... 47
Assessing Reproductive Success and Determining Survival Estimates . 55
Population Projections ....................................... 59
DISCUSSION ................................................... 70
General Observations ........................................ 7O

Duck Island ......................................... 7O

Confined Disposal Facility ............................. 71

Predation ................................................. 72

Site Characteristics .......................................... 77

iv

Assessing Reproductive Success and Determining Survival Estimates . 79
Population Projections ....................................... 81

CHAPTER II: USE OF NESTING PLATFORMS TO INCREASE COMMON TERN

REPRODUCTIVE SUCCESS

INTRODUCTION ................................................ 85

OBJECTIVES ................................................... 89

METHODS ..................................................... 90

RESULTS AND DISCUSSION ..................................... 99
SUMMARY AND RECOMMENDATIONS ................................ 101
LITERATURE CITED ................................................. 105

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

LIST OF TABLES

Population projection calculations used to predict the long-term survival of
common terns on the Confined Disposal Facility (CDF) in Saginaw Bay,
Michigan. 30

Greatest number of active common tern nests on a given day, suggesting
the number of breeding pairs, on Duck Island and the Saginaw Bay
Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan,
summers 1995 - 1997. 38

Mean clutch size (number of eggs per nest) of corrunon tern nests on the
Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of eastern
Michigan, summers 1995 - 1997. 39

Number of common tern young fledged per pair (out of those that
successfully hatched) on the Confined Disposal Facility (CDF) in Saginaw
Bay of eastern Michigan summer 1997. Only values obtained from
enclosures in 1997 are reported due to the high rate of censoring in 1995
and 1996 and outside of the enclosures in 1997. 40

Number of common tern young surviving to hatching and/or fledging per
clutch (%) on the Confined Disposal Facility (CDF) in Saginaw Bay of
eastern Michigan, summer 1997. Only values from enclosures used in
1997 are presented. 41

Number of common tern chicks banded in each age class (aged according
to criteria presented by Nisbet and Drury (1972)) on the Saginaw Bay
Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan,
summers 1996 - 1997. 43

Number of dead common tern chicks recovered and last age when known
alive in each age class (aged according to criteria presented by Nisbet and
Drury (1972)) on the Saginaw Bay Confined Disposal Facility (CDF) in
Saginaw Bay of eastern Michigan, summers 1996 - 1997. 45

vi

Table 8

Table 9

Table 10

Table l 1

Table 12

Table 13

Table 14

Table 15

Table 16

Table 17

Mean (SD) vegetation characteristics at common tern nests of combined
successfiil and unsuccessful hatching on the Saginaw Bay Confined
Disposal Facility (CDF) in Saginaw Bay of eastern Michigan, in summers
1996 and 1997. 48

Mean (SD) vegetation characteristics at common tern nests of successful
and unsuccessful hatching on the Saginaw Bay Confined Disposal Facility
(CDF) in Saginaw Bay of eastern Michigan, in summer 1997. 50

Mean (SD) vegetation characteristics at common tern nests of successful
and unsuccessful hatching on the Saginaw Bay Confined Disposal Facility
(CDF) in Saginaw Bay of eastern Michigan, in summer 1996. 51

Mean (SD) vegetation characteristics at common tern nests of successful
hatching on the Saginaw Bay Confined Disposal Facility (CDF) in
Saginaw Bay of eastern Michigan, in summers of 1996 and 1997. 52

Mean (SD) vegetation characteristics at common tern nests of
unsuccessful hatching on the Saginaw Bay Confined Disposal Facility
(CDF) in Saginaw Bay of eastern Michigan, in summers of 1996 and
1997. 53

Mean (SD) vegetation characteristics at common tern nests of successful
and unsuccessful fledging (at least one young from a nest fledged) on the
Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of eastern
Michigan, in summer 1997. 54

Daily (SD) and period (Sp) survival rates (variance; Mayfield 1961) of
common tern eggs on the Saginaw Bay Confined Disposal Facility (CDF)
in Saginaw Bay of eastern Michigan, summers 1995 - 1997. 56

Daily (SD) and period (8,.) survival rates (variance; Mayfield 1961) of
common tern eggs on Duck Island in Saginaw Bay of eastern Michigan,
summers 1995 - 1997. 57

Daily (SD) and period (8,.) survival rates (variance; Mayfield 1961) of
common tern chicks on the Saginaw Bay Confined Disposal Facility
(CDF) in Saginaw Bay of eastern Michigan, summers 1996 - 1997. No
estimates of chick survival are available for 1995. 58

Overall period (8,.) survival rates (variances; Mayfield 1961) of common
terns from initiation to fledging on the Saginaw Bay Confined Disposal

Facility (CDF) in eastern Michigan, summer 1996 - 1997. 60

vii

Table 18

Table 19

Table 20

Table 21

Table 22

Table 23

Combinations of common tern egg survival (Mayfield estimate, Mayfield
1961), chick survival (Mayfield estimate), and mean clutch size on the
Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan
used in population projections that yielded A 2 1, indicating a stable or
increasing population. 61

Population projection of common terns on the Confined Disposal Facility
(CDF), in eastern Michigan. Projection represents the most representative
case scenario using egg survival = 0.4461 , chick survival = 0.637, and
mean clutch size = 2.34. See Table 1 and associated text for description of
calculations. 64

Population projection of common terns on the Confined Disposal Facility
(CDF), in eastern Michigan. Projection represents the worst case scenario
using egg survival = 0.3571, chick survival = 0.598, and mean clutch size
= 2.04. See Table 1 and associated text for description of calculations. 66

Population projection of common terns on the Confined Disposal Facility
(CDF), in eastern Michigan. Projection represents the best case scenario
using egg survival = 0.8887, chick survival = 0.932, and mean clutch size
= 2.61. See Table l and associated text for description of calculations. 67

Manipulation of common tern egg survival (8,), chick survival (SP), and
mean clutch size to generate A. = 1.0. Mean values generated for egg
survival, chick survival, and mean clutch size were used as the basis of the
manipulations. 68

Supplies, including prices and quantities, for building one common tern

nesting platform used near Duck Island in Saginaw Bay, Michigan in
summer 1995 - 1997. Prices are given in 1997 dollars. 98

viii

Fig. 1

Fig. 2

Fig. 3

Fig. 4

Fig. 5

Fig. 6

Fig. 7

Fig.8

Fig. 9

LIST OF FIGURES

Location of common tern colonies (Confined Disposal Facility (CDF) and
Duck Island) in Saginaw Bay, Michigan in summers 1995 - 1997. 12

Aerial view of Duck Island located in Wildfowl Bay State Game Area of
Saginaw Bay, Michigan showing location of common tern nesting colony
in relation to vegetation and the water’s edge. 13

Side view of Duck Island in Wildfowl Bay State Game Area in Saginaw
Bay, Michigan showing height of island fiom water line and vegetation
height. 14

Side view of Saginaw Bay Confined Disposal Facility (CDF), Michigan
showing the design of the island including water level, outer retaining
wall, interior dike wall, and interior cell. 16

Aerial view of Saginaw Bay Confined Disposal Facility (CDF), Michigan
showing location of common tern colony and other nesters including ring-
billed and herring gulls, black-crowned herons, and caspian terns. 1 7

Diagrarnmatic representation of infra-red camera system connected to time
lapse recorders used on the Confined Disposal Facility (CDF) in Saginaw
Bay, Michigan to record predation activities within the CDF common tern
colony summer 1996 - 1997. 25

Number of common tern clutches, eggs, and chicks on the Confined
Disposal Facility (CDF) in Saginaw Bay, Michigan summer 1997. 34

Number of common tern clutches, eggs, and chicks on the Confined
Disposal Facility (CDF) in Saginaw Bay, Michigan summer 1996. 35

Number of common tern clutches and eggs on the Confined Disposal
Facility (CDF) in Saginaw Bay, Michigan summer 1995. 36

ix

Fig. 10

Fig. 11

Fig. 12

Fig. 13

Location of Duck Island and other islands of Wildfowl Bay State Game
Area in Saginaw Bay, Michigan. 87

Duck Island in Saginaw Bay, Michigan and placement of common tern
nesting platforms in summer 1995 - 1997. 91

Diagrammatic representation of one common tern nesting platform used
near Duck Island in Saginaw Bay, Michigan summer 1995 - 1997.
Illustration depicts platform completely assembled and installed but
lacking substrate. 92

Diagrammatic representation of one common tern nesting platform used
near Duck Island in Saginaw Bay, Michigan summer 1995 - 1997.
Illustration depicts platform partially disassembled. 96

INTRODUCTION

The common tern (Sterna hirundo) breeds throughout a wide band of latitudes in
the Northern Hemisphere (Morris et a1. 1980). In Michigan, nests can be found on all of
the Great Lakes with a few birds nesting on inland lakes (Scharf 1991). A notable gap in
their distribution in Michigan occurs along the western shore of the state, where preferred
nesting habitat is lacking. Tems nest in highest numbers on human-made islands in the
Saint Mary’s River, in northern Lake Michigan, and in Saginaw Bay of Lake Huron
(Scharf and Shugart 1985, Shugart and Scharf 1983, F. Cuthbert, University of
Minnesota, pers. commun.)

Throughout their range common terns are colonial nesters and require three
features for colony nest site selection including physical isolation from predators, a
constant supply of food nearby, and flat and relatively open habitat with good visibility
and sparse vegetation (Austin 1929, Palmer 1941 , and Burger and Gochfeld 1991). Tems
prefer nest sites on peninsulas or islands presumably to limit or prevent access by ground
predators (Courtney and Blokpoel 1983, Scharf 1981). Colonies are typically located in
estuaries, bays, lakes, rivers, and occasionally marshes (Bent 1921 , Palmer 1941).
Colony sites located throughout the Great Lakes have included sandbars, shoals,
breakwaters, and artificial islands. Selected nests sites are usually located on well

1

2

drained substrate including gravel, sand, dredge spoil, chipped concrete, iron slag, bare
rock, and rocky spits. Nest sites can be characterized by early plant succession with 10 -
30% vegetation cover (Burger and Gochfeld 1991). If vegetation is widely scattered
throughout the colony site common terns will select a nest site near an individual plant to
provide shade for chicks after hatching. Additionally, terns often use objects that
protrude from the substrate as focal points when returning to the nest site. Unobstructed
vision also is an important factor in site selection (Blokpoel et al. 1978, Burger and
Gochfeld 1991). If colony members can not easily see or hear each other, then fear
reactions (i.e., panic flights) cannot be communicated among individuals within the
colony (Blokpoel et al. 1978). Although too much vegetation can negatively impact
survival, some vegetation provides refuge from predators and needed thermal cover
(Palmer 1941).

Palmer (1941) stated that a constant food supply nearby is influential in colony
site selection. Tems display a tendency to return to the same colony each year if
reproductive efforts have been successful at the site (Austin and Austin 1956, Courtney
and Blokpoel 1983). Site tenacity in common terns increases as birds age (Austin 1942).
Because food availability influences reproductive success and common terns are site
tenacious, especially when reproductive efforts have been successful, food availability
should influence site selection.

Colony size can be quite variable with colonies as small as 10 nests or as large as
several thousand nests (Burger and Gochfeld 1991). Inter-nest distance for common terns

has been reported to vary between 45 - 300 cm (Burger and Gochfeld 1991 ). Common

3

terns may nest with other species [e.g., black Skimmers (Rynchops niger); Burger and
Gochfeld 1991] or with other species nearby [i.e., ring-billed (Larus delawarensis) and
herring gulls (Lams argentatus) in Saginaw Bay; Winterstein and Millenbah 1996].

Two or three eggs are commonly laid by common terns with the third egg being
the smallest and least likely to survive (Ehrlich et a1. 1988). Common terns have been
observed with up to 5 eggs in a nest, but this is rare. Typically common terns produce a
single brood and may attempt to raise a second brood but are rarely successful (Hays
1984, Wiggins et a1. 1984). Initial nest loss is frequent and may be compensated by
second nesting. Common terns are monogamous within a breeding season and hatching
within a colony is synchronous. Food availability and precipitation determine the timing
of initial nesting (Becker et a1. 1985, Burger and Gochfeld 1991) with most egg-laying
commencing in mid- to late-May and ending by early June (Austin 1932, Nisbet 1973).
In highly disturbed sites, initiation of egg laying may continue into late July (Winterstein
and Millenbah 1996).

Once abundant and widespread in the Great Lakes and many inland lakes
(Barrows 1912), common terns were first threatened with extinction in the early 1900's
due to plume and feather trade. After passage of the Migratory Bird Treaty Act, the
common tern population in Michigan rebounded, but has never recovered to previous
levels (Brewer 1991). Currently, common terns in Michigan are designated as a State
threatened species. Other states in the Great Lakes region that provide designation to
common terns as threatened or endangered include Ohio, Illinois, Wisconsin, Minnesota,

and New York. Additionally, the entire Great Lakes population of common terns is

4

currently undergoing a status assessment for possible listing as Federally endangered (S.
Lewis, U. S. Fish and Wildlife Service (USFWS), pers. commun.)

Recent population estimates for common terns in Michigan number around 1,200
- 1,400 breeding pairs (Evers 1992, F. Cuthbert pers. commun.). Banding records
indicate that the Great Lakes population is fairly discrete with little movement to and
from other areas (Austin 1953, Haymes and Blokpoel 1978), thus the size of the
population and its persistence will largely be determined by natality and mortality.
Unfortunately, the reproductive success of the common tern in the Great Lakes is below
that of colonies on the eastern seaboard of North America (Nisbet and Drury 1972,
Morris et al. 1976), suggesting that their continued existence may be severely
jeopardized. Several factors have been identified that may negatively impact
reproductive success and survivorship, including loss of habitat, competition with gulls,
predation, and contaminant effects.

Loss of available beaches in Michigan to high waters in the 1980's, plant
succession, and human disturbance have caused a dramatic decline in the number of
common tern breeding pairs (Scharf and Shugart 1985). Elevated lake levels have
eliminated much of the natural island nesting habitats where many long-time colony sites
have been washed away (Morris and Hunter 1976, Scharf 1991). Because terns require
nest sites with sparse, low vegetation that are isolated from mammalian predators
(Courtney and Blokpoel 1983), they often nest on bare sandy, gravelly parts of islands or
peninsulas highly susceptible to wave damage and erosion (Dunlop et a1. 1991). This

situation has been exacerbated during the past several decades by large populations of

5
ring-billed gulls which have forced common terns into marginal habitat close to the water
line (Blokpoel and Scharf 1991).

Loss of habitat also can be attributed to residential and commercial development
along the Great Lakes shoreline. This loss of natural habitat has resulted in a shift to the
use of human-made islands (Scharf 1991). Other human activities that adversely affect
common tern nesting success include recreation in or near colonies and egg collection.
Birders and photographers cause disturbances at nest sites and may keep brooding adults
off nests thus exposing eggs and chicks to thermal stress and/or opportunistic predators
(Courtney and Blokpoel 1983, Burger and Gochfeld 1991).

Vegetation encroachment also reduces the amount of area available for nesting.
Tems prefer to nest in locations with sparse vegetation (Palmer 1941), however, as plant
growth progresses and increases in density, terns may be forced to abandon nest sites
(Courtney and Blokpoel 1983, Shields and Townsend 1985). For example, seasonal
growth of sandbar willow (Salix exigua) forced terns to abandon nests at a Minnesota
colony (McKeaman and Cuthbert 1989) while terns along the Maumee River colony in
Ohio were physically unable to reach their nests as a result of seasonal grass growth in
1980 (Shields and Townsend 1985).

As indicated previously, available nesting sites for common terns also have been
depleted due to the expanding ring-billed gull population (Courtney and Blokpoel 1983,
Shugart and Scharf 1983, Ludwig 1991). Morris and Hunter (1976) suggest that the
widespread decrease in common tern numbers has been related to the occupancy of

nesting areas by ring-billed gulls which results in terns nesting in less preferred sites that

6

may be more vulnerable to human disturbance, pollution, and predation. Because gulls
arrive on their breeding sites 2 - 3 weeks prior to terns, out-competing terns for nesting
sites is not difficult (Morris and Hunter 1976). In addition to the physical displacement
of common tern colonies, the vegetation structure within a colony can be damaged due to
nesting gulls. Vegetation can be killed or removed by acidic gull fecal matter and pulling
and/or trampling of plants within the colony (McBrayer et a1. 1995).

Predation also contributes to the declining common tern population. Although
some authors (Hatch 1970, Bumess and Morris 1992) indicate that ring-billed gulls often
have a large impact on the reproductive success of terns in terms of predation, Courtney
and Blokpoel (1980) found that ring-billed gulls have little to no effect on common tern
reproductive success. Rather, ring-billed gulls posed more of a threat by excluding terns
from otherwise quality nesting habitat (Morris and Hunter 1976).

Herring gulls also are known to cause declines in common tern reproductive
success. In Maine, herring gull predation on common tern eggs and chicks was suggested
as the major factor in lowered reproductive success of a colony (Hatch 1970).
Additionally, black-crowned night herons (Nycticorax nycticorax) have been identified as
a nocturnal predator of common tern colonies preying primarily on eggs and l - 3 day-old
chicks (Collins 1970, Morris and Hunter 1976). Similarly, great horned owls (Bubo
virginianus; Morris and Wiggins 1986) and ruddy tumstones (Arenciria interpres;
Farraway et a1. 1986, Morris and Wiggins 1986) are known to prey on tern chicks. Great
horned owls have removed over 100 common tern chicks from a single colony in a given

breeding season (Courtney and Blokpoel 1983). Other avian predators include American

7
crows (Corvus brachyrhychos; LaBarr 1995) and Canada geese (Branta canadensis;
Courtney and Blokpoel 1983).

Mammalian predators also have been documented to have significant impacts on
common tern reproductive success. Norway rats (Rattus norvegicus) have been identified
preying on common tern eggs, chicks, and adults in colonies along the Atlantic coast
(Austin 1932, Palmer 1941). Feral cats (F elis cattus) have interrupted incubation by
preying on eggs (Shields and Townsend 1985) while mink (Mustela vison) have been
observed preying on eggs and half-grown chicks (Courtney and Blokpoel 1983, Penning
and Cuthbert 1993). Racoons (Procyon lotor), striped skunks (Mephitis mephitis), and
red foxes ( Vulpes vulpes) also have been identified as predators of common tern eggs and
chicks in colonies in Ohio (Stricker 1995), Lake Ontario (Courtney and Blokpoel 1983),
Minnesota (McKeaman and Cuthbert 1989), and along the Atlantic coast (Palmer 1941).

Avian and mammalian predation of common terns is most often described in the
literature but reptilian and invertebrate predation also have been reported. Predation by
garter snakes (T hamnophis spp.) has been recorded at New England colonies (Lyon 1927)
while Stricker (1995) found evidence of fox snake (Elaphe vulpena) predation on
common terns in Ohio. Additionally, ants have been reported entering pipping common
tern eggs as well as preying upon newly hatched chicks (Austin 1932). Newly hatched
chicks have been killed and/or blinded by ants in a Massachusetts common tern colony
(Shields and Townsend 1985). Many ant predations coincide with predations by owls
and subsequent nocturnal desertion of the colony by adult common terns (Nisbet 1972).

Nisbet and Welton (1984) suggested brooding adults are able to keep ants out of pipping

8

eggs and away from chicks, but nocturnal desertion allows enough time for ants to access
eggs and chicks and kill them.

Contaminants have also been identified as a possible cause of the declining
common tern population. High concentrations of some contaminants were reported in
common tern eggs from the Great Lakes in the 1970's (Gilbertson and Reynolds 1972,
Morris et a1. 1976). Additionally, a study conducted from 1969 - 1973 by the Canadian
Wildlife Service showed that the eggs of common terns nesting in the Great Lakes
contained relatively high levels of several organochlorine contaminants (Weseloh et a1.
1989). This coincided with a high incidence of chick deformities and a lower
reproductive success. However, common terns may be exposed to various chemical
contaminants not only on their breeding grounds but also on wintering grounds and along
migration routes. Fox (1976) reported low concentrations of some contaminants in the
summer food of common terns and assumed that bioaccumulation in these birds resulted
from the intake of contaminated food on wintering grounds. Although high
concentrations of some contaminants were reported in the 1970's, a study conducted in
1989 indicated that organochlorine pollution levels had been reduced in the Great Lakes
suggesting that they are no longer an important factor in the population dynamics of the
Great Lakes common tern (Weseloh et al. 1989).

From previous research it is obvious that one or many factors may be impacting
the continued survival of the common tern. To make the most appropriate management
recommendations for promoting the continued existence of the species, information must

be collected at each colony to understand those factors that either threaten or may

9

potentially threaten their survival. In Michigan, two common tern colonies in Saginaw
Bay, located in Lake Huron, have been monitored by the USFWS for several years. In
1994, a human-made island in Saginaw Bay (Confined Disposal Facility; CDF) supported
approximately 100 breeding pairs of common tems. Few if any young were believed to
have fledged from the site (L. Williams, USFWS, pers. commun.). A second island in
Saginaw Bay, Duck Island, was observed in 1993 to fledge few if any young from the 75
- 100 breeding pairs at the site (D. Best, USFWS, pers. commun.). In each case, USFWS
biologists were uncertain as to the cause of the poor reproductive success. Thus it was
necessary and imperative that investigations be undertaken to identify potential causes of
declines at these sites so that appropriate steps could be taken. The purpose of this
research was to determine the factors affecting the reproductive success of the common
tern in Saginaw Bay, Michigan and to make appropriate management recommendations

that could ensure the continued existence of the species.

1)

2)

3)

4)

5)

OBJECTIVES

Specific objectives of this study were to:

determine the reproductive success of common terns in Saginaw Bay;
determine the frequency and impacts of mammalian and avian predators on
common tern reproductive success;

predict the long term success of common terns using population projection
models;

determine the effectiveness of nesting platforms in increasing common tern
reproductive success; and

make recommendations for managing common terns for increased reproductive

success on natural and human-made islands.

10

STUDY AREA

Common tern colonies were observed at two locations in Saginaw Bay, Michigan
during the summers of 1995 ~ 1997: 1) Duck Island and 2) Saginaw Bay CDF. Duck
Island (< 0.5 ha) is located in Wildfowl Bay State Game Area between Heisterrnann and
Maisou Islands (43 °50'30"N 83 °27‘30"W; Figure 1). Due to the location (nearest
landforrn > 400 m away) and small size of the island, no permanent avian or mammalian
residents are found on the island. Common terns nest on the southwest side of the island.
No nests are found on the northeast side (Figure 2). Duck Island is surrounded by
shallow water (< 1.5 m) but can be easily flooded in storm events or completely
submerged in high water years. Vegetation is clumped near the center of the island with
common species including aspen (Populus spp.; < 4 m in height) and reed canary grass
(Phalaris arundinacea; Figure 3).

The CDF is an artificial kidney shaped island (118.5 ha) near the mouth of the
Saginaw River (43 °40'N 83 °50'30"W; Figure 1). This structure was built around the
Channel and Shelter Islands by the Army Corps of Engineers in 1978 to house
contaminated dredge material (fine silt, sand) from the Saginaw River and Saginaw Bay
shipping channel (Hodgkins 1993). From 1978 - 1996 disposal of all hydraulically
dredged material occurred in late summer and early fall to avoid impacting breeding bird

ll

12

 

Lake Huron

Tuscola

Fig. 1. Location of common tern colonies (Confined Disposal Facility (CDF) and Duck
Island) in Saginaw Bay, Michigan in summers 1995 - 1997.

l3

 

Common Tern
Nesting Sites

Fig. 2. Aerial view of Duck Island located in Wildfowl Bay State Game Area of Saginaw
Bay, Michigan showing location of common tern nesting colony in relation to vegetation
and the water’s edge.

14

Emma: nouflowg EH on: $33 50¢ 283
no Emma. mega... .3222 «am 8&8 s 8:. ease 25 am 332$, s sin 0.25 co 32> 206 .m .3

 

 

 

 

15

use of the facility. Since its construction, the CDF has been a haven for migratory birds
[i.e. black-crowned night herons, semipalrnated plovers (Charadrius semipalmatus),
spotted sandpipers (Actitis macularia), caspian terns (Sterna caspia), common terns, red-
winged blackbirds (Agelaius phoeniceus), ring-billed gulls, herring gulls, great egrets
(Ardea albus); Hodgkins 1993]. It also supports mammals [white-tailed deer (Odocoileus
virginianus), muskrat (Odontra zibethicus)] and herptiles [American toad (Bufo
americanus), unidentified garter snake (Thamnophis spp.); Hodgkins 1993].

The outer retaining wall of the CDF rises approximately 3 - 5 m above water level
with the interior dike wall, which completely surrounds the periphery of the CDF,
approximately 0.5 m below the top of the retaining wall (Figure 4). Dropping steeply 2 -
2.5 m from the interior dike wall is the interior cell where dredge spoil material is
deposited (Figure 4). Large boulders constitute the outer retaining wall while coarse
packed gravel and rock chips make up the interior dike wall.

Common terns nest in the east-central section of the CDF (Figure 5). Most nests
occur on the interior dike wall with few nests occurring in the interior cell and on the
outer retaining wall. Common tern colony dimensions are approximately 75 m long by
the width of the interior dike (approximately 2 m). During the breeding seasons of this
study the colony was bordered on the north by herring and ring-billed gulls and on the
south by herring gulls (Figure 5).

Common vegetation species in the CDF colony include bittersweet nightshade
(Solanum dulcamara), spotted knapweed (Centaurea maculosa), field peppergrass

(Lepidium campestre), field pennycress (Thlaspi arvense), rough cinquefoil (Potentilla

16

.=oo .852: 23 £85 8% Stat: .325 9:558 .830 ._o>o_ 833 memes—2:
ease as co Ea“. 2: mesons 5min: .958 3:5 583 Bungee am 5&8 do is 2% .4 a;

:35

=u>> o
9:582... .850 BMW—c.

=00 3:3:—

\||}\II<I.}\ . 9.... {Q ..
, . , soooqwuo a... $5.. ... ...___....._.
, . . usflwfieug ._ f: t

 

 

 

 

17

.253 Swans can .898: EH: 3:329x83 .m=:m math: 28 3:59.: 3622: $88: .550 98 EMA—co
E3 :oEEoowo 5:32 wagon.» semi—£2 .EQUV mum—moan 28235 3:580 mam 39:me .«o 33> Hive. m . m

           

230 uchI

duo EDEN—.2.

9.6 .880

.. . «=8 33sec

Q2—

".00

18

norvegica), and white sweet clover (Melilotus alba).

Caspian terns nest in the interior cell on the west-central section of the CDF while
black-crowned night herons, great egrets, and great blue herons (Ardea herodias) are
common nesters in trees on the northwest side of the CDF (Figure 5). Trees in this area
are remnants from the original Channel and Shelter Islands. The closest trees supporting

these nesters occur no closer than 100 - 150 m from the common tern colony.

CHAPTER I
POPULATION ASSESSMENT AND REPRODUCTIVE STATUS OF
THE COMMON TERN IN SAGINAW BAY

INTRODUCTION

Declining numbers of common terns in the Great Lakes have caused considerable
concern over the future of the species. So great is this concern that the USFWS recently
ordered a status assessment of common terns for the entire Great Lakes population for
possible listing as Federally endangered (S. Lewis, pers. commun.). Regional declines in
nesting pairs have been noted in the Canadian region of the southern Great Lakes
(Blokpoel et al.1978) and in northern Minnesota (Penning and Cuthbert 1993). In
Michigan, estimates for common terns have declined from approximately 6,000 breeding
pairs in the early 1960's (Ludwig 1962) to approximately 1,400 breeding pairs in the
1980's (Scharf and Shugart 1991). More alarming, however, was the noticeable decline
in the number of breeding colonies from 31 to 23 in the Great Lakes region during that
time period. Those areas that experienced a particular reduction in the number of
breeding colonies included Duluth-Superior of Wisconsin and Minnesota, Green Bay of
Wisconsin, and Thunder Bay and Saginaw Bay of Michigan (Scharf and Shugart 1991).

The threats that have caused this continuing decline in common terns both in

19

20

Michigan and in the Great Lakes region are well documented and include such things as
loss of habitat due to vegetation encroachment (Palmer 1941, Courtney and Blokpoel
1983, Shields and Townsend 1985, McKearnan and Cuthbert 1989) and rising water
levels (Morris and Hunter 1976, Scharf and Shugart 1985, Scharf 1991), competition for
nesting space with ring-billed gulls (Courtney and Blokpoel 1983, Ludwig 1991, Shugart
and Scharf 1983), predation (Austin 1932, Collins 1970, Courtney and Blokpoel 1983,
Farraway et a1. 1986, Hatch 1970, LeBarr 1995, Morris and Wiggins 1986, Penning and
Cuthbert 1993, Shields and Townsend 1985, Stricker 1995), and contaminants
(Gilbertson and Reynolds 1972, Morris et a1. 1976, Weseloh et a1. 1989). With this
continued decline in common tern numbers many authors have attempted to determine
the minimum population parameter values necessary to maintain stable populations of
common terns under the given conditions. For example, Austin (1942) and Austin and
Austin (1956) used a life table approach to estimate that to maintain a stable common tern
population in Massachusetts adult annual survival must be between 71 - 75%.
Additionally, they predicted that at least 20% of the chicks fledged in a year must return
to breed at age 4 (common terns do not reach breeding maturity until 3 - 4 years of age) if
the population is to maintain itself. Nisbet (1978) pointed out errors in the above
calculations and re-calculated required armual survival for adult common terns to be 87%
with 10% of the fledged chicks surviving to enter the breeding population to maintain a
stable population. However, DiConstanzo (1980) indicated that the above information
should be interpreted with care as the estimates were based on populations that were

declining and may not be representative of a stable population. DiConstanzo (1980),

21

therefore, modeled a stable population for common terns in New York and suggested that
annual adult survival needed to be 92% with 14% of the fledglings returning to breed at
age 4 to maintain a stable p0pulation. Penning (1993) developed a deterministic model to
explore the population dynamics of common terns in Minnesota. He found that to
maintain a stable population of common terns in Minnesota they must have an age-
specific birth rate of 1.10, adult survival of 92%, sub-adult survival of 15%, and 12.5% of
sub-adults must breed. No estimates of common tern survival or population longevity
have been generated for Michigan colonies.

Before estimates of survival or predictions of population longevity can be
generated, one must first understand current threats to a colony and how these threats
may potentially impact survival. The purpose, therefore, of this research was to
determine the reproductive success of two common tern colonies in Saginaw Bay of
eastern Michigan and to identify threats to their continued survival. From this,
population projections can be generated to predict the survivorship of these colonies
under existing threats. Duck Island and the Saginaw Bay CDF (Figure 1) were chosen as
sites for this research because of interest from the USFWS in regards to an ongoing
Natural Resource Damage Assessment of Saginaw Bay. Additionally, during much of
the 1980's, coinciding with high water levels, the Saginaw Bay CDF supported
approximately one quarter of Michigan’s nesting population of common terns (Evers

1994), making it a potential stronghold for Michigan’s breeding population.

1)

2)

3)

4)

OBJECTIVES

Specific objectives of this part of the study were to:

determine the reproductive success of common terns in Saginaw Bay;
determine the frequency and impacts of mammalian and avian predators on
common tern reproductive success;

predict the long term success of common terns using population projection
models; and

make recommendations for managing common terns for increased reproductive

success.

22

METHODS

General Observations

Common tern colonies were observed from May - August 1995 - 1997 on the
CDF and Duck Island (Figure 1). Observations of sites were conducted to identify
potential predators to the colonies. Colonies were monitored approximately every other
day to minimize disturbance to the sites. Sites were not visited in inclement weather.
Due to the small size of Duck Island, observations were made from a boat anchored
approximately 50 m from the island. The CDF colony was observed from the center dike
>100 m from the colony in 1995. In 1996 and 1997, observations were made from blinds
located 10 m from each end of the colony. Observations were made with either a spotting
scope or binoculars.

Researchers randomly selected a group of terns or a nest site for observations. An
attempt was made to observe selected sites for a minimum of 2 h. Behavior (e. g.,
aggression) among and between terns and other species was recorded. Disturbances (e.g.,
presence of potential predator) and disturbance length were also recorded when possible.
No observations were made during the evening hours for any year. In 1995, observations
were made randomly from sunrise to sunset. Researchers were attempting to obtain a
better understanding of common tern behavior, thus times of visitation varied to provide

23

24

the greatest spectrum of different behaviors. Observations in 1996 and 1997 were limited

to 2 time periods (representing peak bird activity): 1) sunrise to 4 h after sunrise, and 2) 4

h prior to sunset to sunset.

Predation

Two and four infra-red cameras connected to time-lapse VHS recorders (Figure 6)
were placed at selected nests on the CDF in 1996 and 1997, respectively, to record
nocturnal predation attempts and other activities. Cameras and recorders were powered
with a 12 volt deep cycle marine battery. Batteries operated each system for 2 days
before recharging was required. Cameras were placed at nests when at least one egg or
young was present in the nest cup. Recorders operated for 12 h from 2000 hr to 0800 hr.
Every attempt was made to operate cameras on consecutive evenings until a nest
successfully fledged or was abandoned or destroyed. Because of the high incidence of
wave inundation at Duck Island, no cameras were placed at this site.

Two infra-red Trail Master trail monitors connected to Olympus 35 mm cameras
were placed at both ends of the CDF colony in 1996 and 1997 to record movement of
terrestrial predators and other animals into the colony. Cameras and monitors were set to
encompass the width of the colony (approximately 2 m). Any motion breaking the infra-
red monitor beam signaled the camera to take a picture, thus recording potential predators
on 35 mm slide film. Cameras ran for 24 h. Every attempt was made to operate cameras
on consecutive days. Because of the high incidence of wave inundation at Duck Island,

no cameras were placed at this site.

25

.33 - 39 SEES .328 EB
8888 "EU 2: :53 8:338 :33on 282 9 53:32 Sam 3m£mam E EOE mum—moan ~38me ecumeoo
05 co eon: 80282 8%. 2h: 2 38058 8393 8083 3.735 mo messaged?" uzaEEEwED .o .wE

.8963”. 3%.. 80:50
oEF 2:825.

 

 

 

 

 

 

 

26

Site Characteristics

Vegetation variables were measured at a minimum of 50 nests per year per site to
determine if survivorship of eggs or chicks was related to habitat attributes.
Measurements were made as soon as a completed clutch (no additional eggs laid > 5 days
after last egg laid) of eggs was present. Vegetation variables were measured in 1996 and
1997 using a modified 50 x 50 cm Daubenmire frame (Daubenmire 1959) centered on the
nest and included: percent total, live, dead, grass, forb, and woody canopy; percent bare
ground; and percent litter cover. Additional measurements included distance to the
closest adjacent tern nest (cm) and distance to the water line (m).

Comparison of vegetation characteristics between 1996 and 1997 for all nests
(successful and unsuccessful hatching combined) were made using a Mann-Whitney U
test («X = 0.0125; adjusted with Bonferroni, Snedecor and Cochran (1980)). Additionally,
a Mann-Whitney U test (rx = 0.0125; adjusted with Bonferroni) was used to compare
vegetation characteristics between successful (hatched) and unsuccessfiil (did not hatch)
nests within a year and between years. Nests were considered successful if at least one
egg hatched from the nest. Similar comparisons (Mann Whitney U test, o< = 0.0125,
adjusted with Bonferroni) were made for those nests that successfully or unsuccessfully
fledged young in 1997. Nests were considered successful if at least one young fledged

from the nest.

 

27

Assessing Reproductive Success and Determining Survival Estimates

To determine reproductive success and collect data needed for population models,
all nests were checked every 2 - 4 days. Nesting variables recorded at each nest included
date of egg laying, number of eggs present each day, date of completed clutch, number of
eggs lost and reason(s) for loss, hatching date, number of chicks hatched, number of
chicks lost and reason(s) for loss, fledging date, and number of chicks fledged. Each nest
was given a number using a permanent marker on a nearby rock to allow nests to be
followed until hatching, fledging, or until the nest was destroyed or abandoned. An
attempt was made to follow all nests to fledging. To facilitate following chicks to
fledging, chicks were hand captured and banded using a USFWS #2 steel band and two
stripe and one solid celluloid color bands when applicable (leg length is long enough to
accommodate four bands). At a minimum chicks received a USFWS steel band and one
black/white split band. No color bands were used in 1995 and chicks received an
aluminum rather than steel USFWS #2 band. Chicks were banded only after they were
completely dry. In 1996 and 1997, chicks were aged according to criteria presented by
Nisbet and Drury (1972). Appropriate banding permits were obtained and capturing and
banding procedures were approved by Michigan State University’s All-University
Committee on Animal Use and Care (AUF # 07/95-099-00).

In 1997 three enclosures were erected within the CDF colony to assist in
following chicks to fledging. Enclosures were erected not to exclude predators but to
enclose chicks which assisted in following chicks to fledging. Large boulders on the

CDF served as excellent hiding cover for chicks, making them difficult to follow to

28

fledging and making survival estimates unreliable, thus enclosures prevented chicks from
hiding under the large boulders. Enclosures were constructed of 1.3 cm X 1.5 cm mesh
plastic cintoflex E fencing and were evenly distributed throughout the colony.
Dimensions of each enclosure were 1.5 m wide X 10 m long X 1 m high. Enclosures
were anchored at the corners and at various locations along the length using 1.5 cm X 1.5
m pieces of conduit. Fencing was attached to conduit using twist-ties. Approximately 20
cm of fencing was rolled to the inside of each enclosure along the bottom edge and
covered with rocks and gravel. This prohibited chicks from being able to escape under
the enclosure. All enclosures were open on the top to permit adults to enter and leave the
area freely.

Daily and period survival estimates for eggs and chicks were calculated using the
Mayfield method (Mayfield 1961) to assist with population projections. Survival
estimates were made for two time periods: 1) from initiation of incubation to hatching
and 2) from hatching to fledging. Nests were used as the experimental unit from
initiation of incubation to hatching and individual chicks were used as the experimental
unit from hatching to fledging. Nests were considered active if at least one egg or chick
was present in the nest cup. Abandoned nests were incorporated into the estimate using a
maximum of 12 exposure days for these nests. Period survival rates for eggs and chicks
were calculated using 24 days (L; Mayfield 1961) as the mean number of incubation days
and 28 days (L) as the mean number of days from hatching to fledging. Chicks were
considered fledged if they were 2 19 days old at last capture. Any chick < 19 days old at

last capture, not found dead in subsequent checks, was considered censored.

29

Population Projections

Populations projections were developed to predict the long-term survival of
common terns in Saginaw Bay. Assumptions of the projections were a 1 :1 sex ratio and
that only females were used in the projections. Only values obtained from the CDF were
used in the population projections. Only birds 2 4 years of age contributed to age 0
individuals in the projection as birds younger than this have not reached breeding age yet.
Although common terns reach breeding age by 3 years of age, research indicates that 3
year-old breeders are rarely successful in nesting attempts and contribute few if any
young to the population (Austin and Austin 1956, DiConstanzo 1980, Nisbet 1978).
Additionally, no birds were projected to live past 10 years of age. (Although common
terns have been reported living up to 15 - 20 years, Austin and Austin (1956) suggest that
birds > 10 years of age comprise only a small portion of the total breeding population.

Year 0 values for projections were calculated as follows and were used as starting
values for all projections generated. Age 0 numbers were generated by calculating the
number of young fledged from the three enclosures in 1997 and were extrapolated to the
entire colony to determine the total number of young that fledged that year (Table 1).
Enclosure data most accurately reflected the actual survival of chicks on the CDF.

Under the assumption of a 1:1 sex ratio, this value was divided by 2 so to reflect use of
females only in the projection. Although chicks were followed for all three years in this
study only estimates from 1997 were included because estimates for other years had such
high rates of censoring that survival estimates generated from those numbers would be

unreliable. Values in the literature (DiConstanzo 1980, Penning 1993) indicate that a

30

Table 1. Population projection calculations used to predict the long-term survival of
common terns on the Confined Disposal Facility (CDF) in Saginaw Bay, Michigan.

 

 

 

Age Class Year 0 Year 1 Year 2
0 x Number of mm3 n20b
Female Chicks in
1997
1 I100 * 0.523 n“ = n,0 * 0.523 n,, = n.0 * 0.523
2 r10l "' 0.523 nl2 = no, * 0.523 n22 = nll "‘ 0.523
3 n02 * 0.523 n,, = n02 * 0.523 n23 = n,2 * 0.523
4 x° . n03 * 0.92 . n,3 * 0.92
5 x * 0.92I . r104 * 0.92 . n” * 0.92
6 x * 0.922 . n05 * 0.92 . n15 * 0.92
7 x * 0.923 . n,6 * 0.92 . n,6 * 0.92
8 x * 0.924 . n07 * 0.92 . nl7 * 0.92
9 x * 0.925 . 1103* 0.92 . n", * 0.92
10 x * 0.926 n”0 = 1109 * 0.92 r12,0 = n19 * 0.92
N XAgeOto Age 10 £Age0to Age 10 2Age0to Age 10
A NI /N0 N2 / Nl

 

10 _
a €120: * SPEggs * SPChickS * x Clutch Size

10

b EnLI * SPEggs * SPChicks * J? Clutch Size
x=4

c where x = (3E number of breeding females from 1995 - 1997) / (1 + 0.92‘ + 0.922 +
0.923 + 0.924 + 0.925 + 0.926)

31

minimum of 14.3% of nonbreeders must return to breed to maintain a stable population
and because no values are available for nonbreeder survival in the 1 - 3 year old age
classes, this was the most representative value of nonbreeder survival. Therefore, age 1
values were generated by multiplying the age 0 values by the cube root ofO. 143, or 0.523
(Table 1). Age 2 values were generated by multiplying age 1 values by 0.523 while age 3
values were generated by multiplying age 2 values by 0.523 (Table 1). Thus, for
nonbreeders (from age 1 - 3), a 14.3% survivorship was modeled (Table 1). Age 4 - 10
year old values were generated from the mean number of greatest active nests from 1995
- 1997 which represented the mean number of females present in the colony. To
determine the number of individuals in each of these age classes, the following formula

wuuw¢
x (1 +0. 92’ +0. 92’ +0. 923 +0. 924+0. 925 +0. 92‘) = mean number of greatest active nests;

solving for x, where x was the number of individuals in age class 4. The number of
individuals in age classes 5 - 10 was found by multiplying x by powers of 0.92 since the
most accurate estimate of adult survival is 92% (DiConstanzo 1980 and Penning 1993;
Table 1).

For all other years in the population projection, age 0 values were calculated by
summing the number of individuals in age classes 4 - 10 from the previous year and
multiplying that value by the period survival rate for eggs, the period survival rate for

chicks, and mean clutch size. This yielded the number of young fledged for the entire

32

colony (Table 1). Multiplying the period survival rate of eggs, period survival rate of
chicks, and the mean clutch sizes yields the age-specific birth rate or m,. The 1 - 3 year
old age classes were obtained by multiplying the number of individuals in the O - 2 age
classes from the previous year by 0.523 (Table 1) representing the age-specific survival
rate, or p,, for all birds in the 1 - 3 year old age classes. Values for age 4 - 10 years old
were taken by multiplying the values for the 3 - 9 year olds from the previous year by
0.92 (Table 1) representing the age-specific survival rate, or p,, for all birds in the 4 - 10
year-old age classes.

Population projections were generated using all possible combinations of egg
survival, chick survival, and mean clutch size to predict the population size of common
terns on the CDF over time. Projections were generated for a minimum of 11 years to
allow for one complete generation cycle. To determine if the population was increasing,
decreasing, or staying constant, the finite rate of increase, or A, was calculated by
dividing the total population size at time t + 1 (e.g., N) by the total population size at
time t (e.g., No; Table 1). A population was considered increasing if A > 1, decreasing if

A <1, and constant ifA =1.

RESULTS

General Observations

Nests were observed in 1997 from 28 May - 1 August on the CDF (Figure 7). No
nests were observed on Duck Island as the island was completely submerged for the
research period. Nest were observed in 1996 from 22 May - 3 August on the CDF
(Figure 8) and 28 May - 11 July on Duck Island. Due to weather and rising water levels
in the Saginaw Bay, Duck Island was heavily inundated with waves and standing water
throughout the observation period in 1996. Common terns deserted Duck Island on 11
July. Researchers visited the site occasionally after that time, however, no birds were
observed using Duck Island afier 11 July. Observations at the CDF ended in 1996 and
1997 when nests were no longer present in the colony and researchers were unable to
hand capture chicks. Observations in 1995 were made from 25 May - 15 July on the CDF
and 24 May - 25 June on Duck Island. No observations were made after this time at
either site as encroaching vegetation limited visibility of nest sites. This problem was
addressed in 1996 with the use of observation blinds at the CDF. Nest checks on the
CDF continued until 7 August 1995 (Figure 9).

Common terns began nesting in 1997 on the CDF on 28 May (Figure 7).
Common terns began nesting on the CDF (Figure 8) and Duck Island in 1996 on 22 May

33

34

.33 388% Hammad—2
Sam Bacmwam 5 058 onam :33me 3:980 05 no 8.ch 98 £30 .8533 53 5588 we con—8:2 .n .wE

 

53.0 ...... _

 

$20 ...i .3
sun

MMMNWWNWWMmmmmmmwmmmmmmmmmm

 

D

JaqumN

 

 

 

35

.82 588:... 53:22
55 Ban—mam 5 £000 tom—moan 38¢me 35.080 05 8 9820 23 .mwwo .8888 EB .8888 .8 898:2 .m .mE

 

 

 

Jeqmnu

 

 

 

 

36

.32 8883
camp—22 Sam 333% E €900 558m 38¢me 35.800 05 8 ammo 28 8888 ES 8888 mo 89:82 .m .me

 

—8¢u g...fl

 

38

 

 

JeqwnN

 

 

 

. 8x.

37

and 28 May, respectively, and in 1995 on 25 May (Figure 9) and 17 May, respectively.
Chicks were first observed on the CDF in 1997 on 14 June (Figure 7). Similarly, chicks
were first observed on the sites in 1996 on 8 June (CDF; Figure 8) and 27 June (Duck
Island) and in 1995 on 30 June (CDF) and 22 June (Duck Island). The greatest number of
active nests on a given day indicating the number of breeding pairs at a colony, varied
from 204 nests in 1997 to 239 nests in 1996 on the CDF and zero nests in 1997 to 136
nests in 1995 on Duck Island (Table 2). In 1996 and 1997 the low number of nests at
Duck Island was due to the high incidence of wave inundation and complete submersion
of the island by rising water levels. Mean clutch size of common terns on the CDF varied
from 2.04 eggs per pair in 1997 (enclosure 1) to 2.61 eggs per pair in 1995 (Table 3). Of
those nests that successfully hatched chicks in 1997, the mean number of young fledged
per pair per enclosure ranged from 0.80 to 1.25 (Table 4). Only values for enclosures
used in 1997 were used to determine the number of young fledged per pair due to the
high rate of censoring in other years and outside of the enclosures in 1997.

Using data collected from enclosures in 1997, of those nests that supported 3 egg
clutches, almost 85% successfully hatched at least one young with 48.1% of 3 egg
clutches hatching 2 eggs (Table 5). Of 3 egg clutches that successfully hatched all 3
eggs, 75% successfully fledged at least 2 young while those that hatched 2 eggs
successfiilly fledged at least 1 young approximately 66% of the time (Table 5). Of 3 egg
clutches that hatched only 1 young, 83% successfully fledged one young (Table 5).
Eighty percent of 2 egg clutches successfully hatched at least one young with 50% of

these clutches hatching both eggs (Table 5). Of 2 egg clutches that successfully hatched

38

Table 2. Greatest number of active common tern nests on a given day, suggesting the
number of breeding pairs, on Duck Island and the Saginaw Bay Confined Disposal
Facility (CDF) in Saginaw Bay of eastern Michigan, summers 1995 - 1997.

 

 

Year CDF Duck Island
1995 206 136
1996 239 31
1997 204 0

 

39

Table 3. Mean clutch size (number of eggs per nest) of common tern nests on the

Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan,
summers 1995 - 1997.

 

 

Year i Clutch Size
1995 2.61
1996 2.22
1997 enclosure 1 (n = 22) 2.04
1997 enclosure 2 (n = 27) 2.09
1997 enclosure 3 (n = 27) 2.19
1997 enclosures combined 2.11
1997 outside enclosure 2.24
1997 entire colony 2.20

i across 1995 — 1997 2.34

40

Table 4. Number of common tern young fledged per pair (out of those that successfiilly
hatched) on the Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan
summer 1997. Only values obtained fi'om enclosures in 1997 are reported due to the high
rate of censoring in 1995 and 1996 and outside of the enclosures in 1997.

 

 

# of Pairs that Successfully i # Fledged
Hatched at Least One Young # Fledged per Pair
Enclosure 1 15 12 0.80
Enclosure 2 16 20 1.25
Enclosure 3 16 15 0.94

All Enclosures 47 47 1.00

 

41

Table 5. Number of common tern young surviving to hatching and/or fledging per clutch
(%) on the Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan,
summer 1997. Only values from enclosures used in 1997 are presented.

 

 

 

 

 

# from hatching
# from clutch surviving to
Clutch Size surviving to hatching Hatch fledging Fledge
3 (n = 27) 3 4 (14.8) 3 1 (25.0)
2 2 (50.0)
1 0( 0.0)
0 1 (25.0)
2 13 (48.1)* 2 3 (25.0)
1 5 (41.7)
0 4 (33.3)
1 6 (22.2) 1 5 (83.3)
0 1 (16.7)
0 4 (14.8)
2 (n = 20) 2 10 (50.0) 2 2 (20.0)
1 5 (50.0)
0 3 (30.0)
1 6 (30.0) 1 2 (33.3)
0 4 (66.7)
0 4 (20.0)
1 (n = 19) 1 2 (10.5) 1 1 (50.0)
0 1 (50.0)
0 17 (89.5)
Total 66 65

 

* one individual censored from hatching to fledging.

42

both eggs, 70% successfully fledged at least one young while those nests that hatched
only one young successfully fledged one young only 33% of the time (Table 5). Of one -
egg clutches, only approximately 11% successfully hatched one young with 50% of those
nests successfully fledging that young (Table 5).

Four and 36 chicks were banded in 1996 and 1995, respectively, on Duck Island.
No chicks were banded on Duck Island in 1997. Three-hundred-fifly-nine, 264, and 166
chicks were banded on the CDF in 1997, 1996 and 1995, respectively. Of the chicks
banded in 1997 and 1996, 98.4% and 72.3%, respectively, were banded when they were 0
- 5 days old due to lack of or slow (in relation to the observer) movement of the chicks
which aided in capture (Table 6). No birds >21 days old were handed in 1996 or 1997.

One 'chick with a severely deformed bill was banded by observers at the CDF in
early July 1996. The bird was not collected for analysis as procedures for collections had
not been previously discussed with the USFWS. In 1997 a 22-day old chick was
recaptured (originally banded at approximately 1 day old) and euthanized due to
abnormal behavior. The bird was able to stand upright but as it attempted to step forward
its’ neck would twist to the right side of the body and down so that the top of the head
was on the ground with the bill of the bird pointing forward. Additionally, the bird would
pull its’ right wing from the side of the body over its’ head. The bird was immediately
sent to the National Wildlife Health Laboratory in Madison, Wisconsin. Laboratory
results indicated the bird was infected with Pasteurella multocida, or avian cholera.

Forty-nine dead chicks were recovered on the CDF in 1996 and 65 were recovered

in 1997. Of these, 38.8% and 18.5% were 0 - 1 day old in 1996 and 1997, respectively

43

Table 6. Number of common tern chicks banded in each age class (aged according to
criteria presented by Nisbet and Drury (1972)) on the Saginaw Bay Confined Disposal
Facility (CDF) in Saginaw Bay of eastern Michigan, summers 1996 - 1997.

 

 

Age ,

(days) 1996 1997
0-1 115 281
2-5 76 43
6-9 19 15
8-12 36 13
12- 15 9 4
15- 18 8 2
17-23 0 1
21+ 0 0
Unknown 1 0

 

Total 264 359

 

44

(Table 7). The greatest number of dead chicks (26.2%; Table 7) in 1997 were in the 2A
plumage class (Nisbet and Drury 1972), or 2 - 5 days old. Chicks 15 - 18 days old were
the next age class most ofien recovered as dead (22.4%) in 1996 (Table 7). Of those
chicks banded in the 0 - 1 day old age group on the CDF, 62 and 91 in 1996 and 1997,
respectively, were never recaptured again (Table 7). Over 56% and 40% in 1996 and
1997, respectively, of the chicks banded in the age 0 - 5 day old were never recaptured
after 5 days old.

Of the 49 dead chicks recovered in 1996 only 5 had visible signs of injury. One
had signs of pecking on the head while the other 4 had openings on the body (i.e., under
the right wing, belly, upper right leg) where the skin and feathers were removed exposing
muscle tissue or organs. In most instances chicks appeared healthy and had no visible
signs of injury. No 2 recovered dead chicks could be assigned to the same nest or
parental unit in 1996. Of the 65 dead chicks recovered in 1997, 12 had visible signs of
injury ranging from broken necks to pokes in the head to abrasions along the back. Only

6 nests lost two chicks, with 3 nests losing both chicks on the same day.

Predation

Two and four infra-red cameras with time-lapse recorders were Operational on the
CDF beginning on 28 May 1996 and 3 June 1997. Cameras were removed from the site
on 18 July 1996 due to improper functioning of the recording unit. Problems were later
attributed to a lack of adequate voltage in the power unit. Cameras were operational until

25 July in 1997. While cameras were operational, no predation was recorded in either

45

Table 7. Number of dead common tern chicks recovered and last age when known alive
in each age class (aged according to criteria presented by Nisbet and Drury (1972)) on the

Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan,
summers 1996 - 1997.

 

 

 

1996 1997
Age (days) Recovered Last Age Known Recovered Last Age Known
Alive Alive
0 - 1 19 62 12 91
2 - 5 4 68 17 55
6 - 9 2 33 11 46
8 - 12 2 44 13 26
12 - 15 4 10 7 22
15 - 18 11 7 1 14
17 - 23 4 5 1 40
21 + 3 0 3 65
Unknown 0 0 0 0

 

Total 49 229 65 359

 

46

year. It should be noted that when cameras were operational they produced high quality
footage of nest sites. That is, the quality of the footage was sharp enough to detect slight
movements of adults (i.e., eye blinking) as well as the presence of eggs and/or chicks.
While the nest was the main focus of each frame, background activities (i.e., other birds
moving in and out of the area) could also be seen.

Trail monitors were used beginning 10 July 1996 and 5 June 1997 at the CDF.
Both cameras continued to function through the end of the field season. Because cameras
were placed at both ends of the CDF colony, information collected only indicates
potential predators throughout the colony. No predation attempts were recorded. Species
recorded on 35 mm film include herring gull, common grackle (Quiscalus quiscula),
common tern, European starling (Stumus vulgaris), spotted sandpiper (Actitis macularia),
and Brewer’s blackbird (Euphagus cyanocephalus).

Although not recorded on film, observers noted the presence of garter snakes
within the CDF colony in 1996 and 1997. In 1997 a garter snake was observed preying
upon a 2 day-old common tern chick. Additionally, garter snakes were observed, almost
weekly, with bulges in their bodies suggesting they had recently fed, possibly on an egg
or chick.

While no mammalian predators were observed within the CDF colony in 1996
and 1997, seat was found within the colony, suggesting the presence of a fox and coyote
or feral dog. Scat contained small bone fi'agments from an avian species suggesting
predation by these individuals on common tern chicks. Additionally, in each of 1996 and

1997 evidence of caching of dead chicks was recorded on 2 separate instances, a behavior

47

commonly attributed to foxes. Observers witnessed 2 separate instances when dead
chicks were found in piles of 5 - 10 chicks per pile in each of 1996 and 1997.

Other observations suggest that predation by black-crowned night herons also
may have impacted the CDF colony. In 1995 and 1996 when black-crowned night herons
were present in large numbers (L. Williams, pers. commun.), several nocturnal predation
events at the colony were evidenced with the loss of up to a quarter of the nests within the
colony in a given night. In 1997, however, when black-crowned night heron numbers

were very low on the CDF, few nocturnal predation events were detected.

Site Characteristics

Vegetation measurements were made at the CDF in June 1996 and 1997. No
measurements were made at the CDF in 1995. Additionally, no measurements were
made at Duck Island in any year because perpetual flooding of the island did not permit
sampling. In 1997, common terns nested significantly closer (P = 0.003) to a neighboring
common tern nest than in 1996 (Table 8). Additionally, percent grass canopy cover and
percent bare ground were significantly less (P = 0.001 and P = 0.002, respectively) at
common tern nesting sites in 1996 than in 1997 while percent forb canopy was
significantly greater (P = 0.009) in 1997 (Table 8). Although not significantly different,
common terns nested closer to the water’s edge in 1996 than in 1997 (Table 8).
Additionally, although not significantly different, nests in 1997 had greater percent live

canopy and less percent total canopy, percent dead canopy, and percent litter cover than

in 1996 (Table 8).

48

Table 8. Mean (SD) vegetation characteristics at common tern nests of combined
successful and unsuccessful hatching on the Saginaw Bay Confined Disposal Facility
(CDF) in Saginaw Bay of eastern Michigan, in summers 1996 and 1997.

 

 

 

1996 1997
Variable (n = 50) (n = 95)
Closest Nest (cm)* 76.8 (45.9) 56.3 (15.9)
Water’s Edge (m) 12.8 (0.9) 13.1 (0.6)
% Total Canopy 62.6 (21.8) 61.4 (14.8)
% Live Canopy 59.1 (22.6) 60.5 (15.9)
% Dead Canopy 2.7 (6.1) 0.4 (2.0)
% Grass Canopy* 16.1 (19.7) 27.7 (21.9)
% Forb Canopy* 44.1 (24.1) 33.4 (17.4)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
% Bare Ground“ 26.3 (20.3) 33.1 (14.3)
% Litter Cover 10.1 (14.8) 5.6 (7.4)

 

* Significantly different between years (Mann-Whitney U test adjusted with Bonferroni,
P< 0.0125).

49

Comparison of vegetation variables and distance to closest nest and water’s edge
between nests that hatched and those that did not yielded no significant differences in
1997 (Table 9). Successful nests were further from the water’s edge and had greater
percentages of forb canopy cover and bare ground than unsuccessful nests (Table 9).

In 1996, those nests that successfully hatched young were significantly (P =
0.009) closer to the water’s edge than those that did not hatch young (Table 10).
Although not significantly different, successfiil nesters were further from a neighboring
nest than unsuccessful nesters and had greater percentages of forb canopy cover and litter
cover (Table 10).

Successful hatchers were significantly closer (P = 0.006) to neighboring nest in
1997 than 1996 (Table 11) while no significant differences were found between
unsuccessful hatchers between 1996 and 1997 (Table 12). Although not significantly
different, successful hatchers were further from the water’s edge and had greater
percentage of total canopy cover, live canopy cover, grass canopy cover, and bare ground
in 1997 than 1996 (Table 11). Unsuccessful hatchers in 1997 were further, although not
significantly, from the water’s edge and had greater percentages of live canopy cover,
grass canopy cover, and bare ground than unsuccessful hatchers in 1996 (Table 12).

No significant differences were found between those nests that successfully
fledged young and those that did not successfully fledge young in 1997 (Table 13).
However, those nests that did not successfully fledge young were closer to a neighbor and
the water’s edge and had greater percentage of total canopy cover, dead canopy cover,

grass canopy cover, and bare ground (Table 13).

50

Table 9. Mean (SD) vegetation characteristics at common tern nests of successful and

unsuccessful hatching on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw
Bay of eastern Michigan, in summer 1997.

 

 

Successful Hatch Unsuccessful Hatch

Variable (n = 61) (n = 34)
Closest Nest (cm) 55.2 (14.0) 58.4 (18.7)
Water’s Edge (m) 13.1 (0.6) 13.1 (0.7)
% Total Canopy 60.3 (14.6) 63.2 (13.6)
% Live Canopy 60.0 (15.7) 61.3 (16.7)
% Dead Canopy 0.3 (1.8) 0.6 (2.4)
% Grass Canopy 26.2 (20.2) 30.3 (24.9)
% Forb Canopy 33.8 (17.5) 32.6 (17.5)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
% Bare Ground 33.8 (14.5) 31.9 (14.1)
% Litter Cover 5.1 (8.0) 5.1 (6.1)

 

51

Table 10. Mean (SD) vegetation characteristics at common tern nests of successful and
unsuccessful hatching on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw
Bay of eastern Michigan, in summer 1996.

 

 

Successful Hatch Unsuccessful Hatch
Variable (n = 12) (n = 38)
Closest Nest (cm) 86.6 (66.2) 73.7 (37.4)
Water’s Edge (m)* 12.1 (1.4) 13.1 (0.5)
% Total Canopy 57.1 (30.1) 64.3 (18.6)
% Live Canopy 55.0 (28.1) 60.4 (20.8)
% Dead Canopy 2.1 (7.2) 2.9 (5.8)
% Grass Canopy 10.8 (12.9) 17.8 (21.3)
% Forb Canopy 44.2 (24.8) 44.1 (24.2)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
% Bare Ground 23.3 (26.8) 27.2 (18.1)
% Litter Cover 12.1 (16.2) 9.5 (14.6)

 

* Significantly different between years (Mann-Whitney U test adjusted with Bonferroni,

P < 0.0125).

52

Table 11. Mean (SD) vegetation characteristics at common tern nests of successful
hatching on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of
eastern Michigan, in summers of 1996 and 1997.

 

 

1996 1997

Variable (n = 12) (n = 61)
Closest Nest (cm)* 86.6 (66.1) 55.2 (14.0)
Water’s Edge (m) 12.1 (1.4) 13.1 (0.6)
% Total Canopy 57.1 (30.1) 60.3 (14.6)
% Live Canopy 55.0 (28.1) 60.0 (15.7)
% Dead Canopy 2.1 (7.2) 0.3 (1.8)
% Grass Canopy 10.8 (12.9) 26.2 (20.2)
% Forb Canopy 44.2 (24.8) 33.8 (17.5)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
% Bare Ground 23.3 (26.8) 33.8 (14.5)
% Litter Cover 12.1 (16.2) 5.1 (8.0)

 

* Significantly different between years (Mann-Whitney U test adjusted with Bonferroni,

P < 0.0125).

53

Table 12. Mean (SD) vegetation characteristics at common tern nests of unsuccessful
hatching on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of
eastern Michigan, in summers of 1996 and 1997.

 

 

1996 1997

Variable (n = 38) (n = 34)
Closest Nest (cm) 73.7 (37.4) 58.4 (18.7)
Water’s Edge (m) 13.1 (0.5) 13.1 (0.7)
% Total Canopy 64.3 (18.6) 63.2 (13.6)
% Live Canopy 60.4 (20.8) 61.3 (16.7)
% Dead Canopy 2.9 (5.8) 0.6 (2.4)
% Grass Canopy 17.8 (21.3) 30.3 (24.9)
% Forb Canopy 44.1 (24.2) 32.6 (17.5)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
%Bare Ground 27.2 (18.1) 31.9 (14.1)
% Litter Cover 9.5 (14.6) 5.1 (6.1)

 

54

Table 13. Mean (SD) vegetation characteristics at common tern nests of successful and
unsuccessful fledging (at least one young from a nest fledged) on the Saginaw Bay
Confined Disposal Facility (CDF) in Saginaw Bay of eastern Michigan, in summer 1997.

 

 

Successful Fledge Unsuccessful Fledge
Variable (n = 21) (n = 74)
Closest Nest (cm) 56.5 (14.9) 56.3 (16.2)
Water’s Edge (m) 13.2 (0.7) 13.1 (0.6)
% Total Canopy 61.0 (15.5) 61.5 (14.8)
% Live Canopy 61.0 (15.5) 60.3 (16.2)
% Dead Canopy 0.0 (0.0) 0.5 (2.3)
% Grass Canopy 25.0 (21.6) 28.4 (22.1)
% Forb Canopy 36.0 (18.2) 32.7 (17.3)
% Woody Canopy 0.0 (0.0) 0.0 (0.0)
% Bare Ground 33.0 (14.9) 33.1 (14.3)
% Litter Cover 6.0 (6.8) 5.5 (7.6)

 

55

Assessing Reproductive Success and Determining Survival Estimates

Daily survival estimates for common tern eggs fi'om initiation of incubation to
hatching on the CDF varied from 99.5% in 1997 to 95.8% in 1996 (Table 14). Duck
Island daily egg survival estimates varied from 0% in 1997 to 93.7% in 1995 (Table 15).
Period survival rates (L = 24) from egg laying to hatching for the CDF varied from 35.7%
to 88.8% in 1996 and 1997, respectively (Table 14). On Duck Island, period survival
estimates (L = 24) for eggs varied from 0% in 1997 to 21.1% in 1995 (Table 15). It
should be noted that censoring at Duck Island was 21.4% and 32.7% in 1996 and 1995,
respectively. Therefore, this data set does not meet the assumption of allowable
censoring (< 10%; Mayfield 1961) and care should be taken in interpretation of the
results at this site. Censoring of eggs from initiation of incubation to hatching was < 10%
for all years on the CDF.

Due to complete submersion of Duck Island no young were produced from this
site in 1997. Only 4 young were known to hatch from the Duck Island site in 1996. All
birds were 2 days old at banding. When the site was checked 2 days later, it was
completely covered in water and chicks were not old enough to fledge. Also, chicks
likely did not make it to dry land as the closest land mass to Duck Island is > 400 m
away. Thus, daily and period survival estimates for chicks on Duck Island in 1996 were
0%. Daily and period survival (L = 28) rates for chicks on the CDF in 1996 were 98.2%
and 59.8% (Table 16), however, censoring of chicks at this site was > 10% and care
should be taken when interpreting the results. Daily survival estimates for chicks from

hatching to fledging for the 3 enclosures erected in 1997 varied from 98.3% to 99.8%

56

Table 14. Daily (SD) and period (8..) survival rates (variance; Mayfield 1961) of common
tern eggs on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of

eastern Michigan, summers 1995 - 1997.

 

 

Year 39 Sp
1995 0.9644 (<0.0001) 0.4189 (<0.0001)
1996 0.9580 ( 0.0007) 0.3571 ( 0.0004)
1997 enclosure 1 0.9798 (<0.0001) 0.6123 ( 0.0129)

1997 enclosure 2

1997 enclosure 3

1997 enclosures combined
1997 outside enclosure
1997 entire colony

x across 1995 -1997

0.9577 ( 0.0001)
0.9951 (<0.0001)
0.9798 (<0.0001)
0.9806 (<0.0001)
0.9804 (<0.0001)
0.9676 (<0.0001)

0.3582( 0.0113)
0.8887( 0.0055)
0.6123( 0.0043)
0.6249( 0.0011)
0.6224( 0.0009)
0.4661( 0.0023)

57

Table 15. Daily (SD) and period (8,.) survival rates (variance; Mayfield 1961) of common
tern eggs on Duck Island in Saginaw Bay of eastern Michigan, summers 1995 - 1997.

 

1995 1996 1997

 

SD 0.9373 (<0.0001) 0.9294 (0.0001) 0.0000
Sp 0.2114 ( 0.0006) 0.1725 (0.0015) 0.0000

 

58

Table 16. Daily (SD) and period (8,.) survival rates (variance; Mayfield 1961) of common
tem chicks on the Saginaw Bay Confined Disposal Facility (CDF) in Saginaw Bay of

eastern Michigan, summers 1996 - 1997. No estimates of chick survival are available for
1995.

 

 

SD 8,.
1996 0.982 (<0.001) 0.598 (0.003)
1997 enclosure 1 0.998 (<0.001) 0.932 (0.004)
1997 enclosure 2 0.983 (<0.001) 0.619 (0.006)
1997 enclosure 3 0.994 (<0.001) 0.840 (0.005)
1997 enclosures combined 0.991 (<0.001) 0.755 (0.002)
1997 outside enclosure 0.983 (<0.001) 0.615 (0.002)
1997 entire colony 0.986 (<0.001) 0.676 (0.001)

i across 1996 -1997 0.984 (<0.001) 0.637 (0.002)

 

59

with period survival (L = 28 days) estimates varying from 61.9% to 93.2% (Table 16).
Daily and period survival (L = 28 days) rates for chicks outside of the enclosures on the
CDF in 1997 were 98.3% and 61.5% respectively (Table 16). However, censoring of
chicks outside of the enclosures was >10% and care should be taken in interpreting
results. Overall (including chicks inside and outside of the enclosure) daily and period
survival rates for chicks on the CDF in 1997 were 98.6% and 67.6%, respectively (Table
16).

Overall period survival (L = 52 days) rates from initiation of incubation to
fledging ranged from 21 .4% in 1996 to 74.7% in an enclosure in 1997 (Table 17). No

estimates of overall survival from initiation of incubation to fledging are available for

1995.

Population Projections

F ive-hundred-seventy-six population projections were generated using all possible
combinations of egg survival, chick survival, and mean clutch size. Of those, only 47
combinations of egg survival, chick survival, and mean clutch size yielded )1 2 1,
indicating a stable or increasing population (Table 18). Based on the mean values
calculated for egg survival (0.4461; Table 14), chick survival (0.637; Table 16), and
mean clutch size (2.34; Table 3) for common terns at the CDF a population projection
(Table 19) was generated yielding a A = 0.94, consistently after 12 years, suggesting that
the population is decreasing by 6% per year. Use of the mean values for egg survival,

chick survival, and mean clutch size provides the most representative scenario of

60

Table 17. Overall period (8,.) survival rates (variances; Mayfield 1961) of common terns
from initiation of incubation to fledging on the Saginaw Bay Confined Disposal Facility
(CDF) in eastern Michigan, summer 1996 - 1997.

 

 

Year ' SP
1996 0.2135 (<0.001)
1997 enclosure 1 0.5707 (<0.001)
1997 enclosure 2 0.2217 (<0.001)
1997 enclosure 3 0.7465 (<0.001)
1997 enclosures combined 0.4623 (<0.001)
1997 outside enclosure 0.3843 (<0.001)
1997 entire colony 0.4207 (<0.001)

x across 1995 -1997 0.4016 (<0.001)

 

61

Table 18. Combinations of common tern egg survival (Mayfield estimate, Mayfield
1961), chick survival (Mayfield estimate), and mean clutch size on the Confined Disposal
Facility (CDF) in Saginaw Bay of eastern Michigan used in population projections that
yielded A 2 1, indicating a stable or increasing population.

 

 

Egg Survival Chick Survival i Clutch Size )1
0.8887 0.598 2.61 1.007
0.8887 0.932 2.61 1.069
0.8887 0.619 2.61 1.011
0.8887 0.840 2.61 1.054
0.8887 0.755 2.61 1.039
0.8887 0.615 2.61 1.010
0.8887 0.676 2.61 1.023
0.8887 0.637 2.61 1.015
0.8887 0.932 2.22 1.046
0.8887 0.840 2.22 1.031
0.8887 0.755 2.22 1.061
0.8887 0.676 2.22 1.001
0.8887 0.932 2.04 1.034
0.8887 0.840 2.04 1.019
0.8887 0.755 2.04 1.005
0.8887 0.932 2.09 1.037
0.8887 0.840 2.09 1.023
0.8887 0.755 2.09 1.008
0.8887 0.932 2.19 1.044
0.8887 0.840 2.19 1.029
0.8887 0.755 2.19 1.014
0.8887 0.676 2.19 1.000

 

62

Table 18. Continued.

 

 

Egg Survival Chick Survival 35 Clutch Size )1.
0.8887 0.932 2.11 1.038
0.8887 0.840 2.11 1.024
0.8887 0.755 2.11 1.009
0.8887 0.932 2.24 1.047
0.8887 0.840 2.24 1.032
0.8887 0.755 2.24 1.017
0.8887 0.676 2.24 1.003
0.8887 0.932 2.20 1.044
0.8887 0.840 2.20 1.030
0.8887 0.932 2.20 1.015
0.8887 0.676 2.20 1.000
0.8887 0.932 2.34 1.053
0.8887 0.840 2.34 1.038
0.8887 0.755 2.34 1.023
0.8887 0.676 2.34 1.008
0.8887 0.637 2.34 1.001
0.6249 0.932 2.61 1.013
0.6249 0.619 2.61 1.005
0.6249 0.932 2.34 1.004
0.6224 0.932 2.61 1.018
0.6224 0.840 2.61 1.005
0.6224 0.932 2.34 1.004
0.6123 0.932 2.61 1.016
0.6123 0.840 2.61 1.002

 

63

Table 18. Continued.

 

Egg Survival Chick Survival i Clutch Size A

 

0.6123 0.932 2.34 1.002

 

 

36 mod 36 :ad wwd 000 00° and Nod 006 :06 «
«New: 003: mndOm 5903 3.03 m:.3.~ deom modem wmdmm mod? ova: :00? Z

 

64

3.: wow wed mod mow OEMN ofimw OEMN 05.2 05mm 052 09mm 2
N0: Nod: med mvd mvd med 05.2 053 oan 05mm 05mm 05mm 0
m9: mmd: 2.3 mmd: mmd: wmd: wmd: oodm coda oowm oo.w~ oowm w
and: 00.: mod: w0w: 2.: 2.: 3.: : 3.: 3.3. mvdm mvdm 00.9. N.
No.0 3...: 3.: owe: Q10: ::.N: ::.N: ::.N: :.N: 8.2 8.3. 8.3 c
mud Eno— :¢.N: 2.3 000: Ed: 2.: :6: 3.: :6: 36m 90mm m
m00 cod: N5: m3: ova: 00.: £6: :03 :03 :03 :03 modm v
Quw 2.0: N0: EN: 00.3 310: 5d: hv.:N end: 00: m0: and: m
wmd: $0: 30.0: 8.2 00mm mo.w~ :ONm #09:. no.3 mhdm mndm 3.6m N
5.3 N0: 3.; :Hm :.Nv wwdv 003 ONE wvdo wiwn wwdm wwém :
00% 00.3 00% wade 3.2. :08 coax 30.8: No.2: 00%: mod: flame: o

 

:: 30> 0: 30> 0 30> w 30> :1 30> o 30> n 30> v 30> m 30> m 30> : 30> o 30> 03:0
0w<

 

0023:0030 .:0 0000:3000 00.: 08: 00330003 :03 : 030:. 00m .v0m n 030 00:30 308
:03 .500 M 3338 0:030 .830 M 3333 30 was: 030000 0.30 033000000000 0008 05 300000000 00:00.65
.3wEo::>: E0030 0: .050: €203”: 30000.5 005.300 0:: :0 0030 008800 .:0 00:00.83 00:02:09: .0: 033:.

65

common tern survival across time on the CDF. Using the lowest and highest values for
egg survival, chick survival, and mean clutch size, population projections were generated
that yielded A = 0.88 and A = 1.08, respectively, consistently after 12 years. Using the
lowest values of egg survival (0.3571; Table 14), chick survival (0.598; Table 16), and
mean clutch size (2.04; Table 3), a population projection indicates that the common tern
population is decreasing by 12% per year (Table 20). Under the best values for egg
survival (0.8887; Table 14), chick survival (0.932; Table 16), and mean clutch size (2.61;
Table 3) a population projection indicates that the CDF common tern colony is increasing
by 8% per year (Table 21).

If conditions remained constant across years, under the most representative
scenario, it would be only 40 years until the common tern colony on the CDF had so few
individuals that the colony would be driven to extinction. The end point for the colony to
be considered active was determined when < 1 individual was present in each of age
classes 4 - 10. Under the worst case scenario, the colony would be driven to extinction in
25 years.

Making the assumption that the mean values for egg survival, chick survival, and
mean clutch size are the most representative values for common terns on the CDF from
1995 - 1997 a A = 0.94 was obtained (Table 19). To obtain a A = 1.0, by manipulating
only egg survival and maintaining the mean values for chick survival and mean clutch
size, would require the increase of egg survival to 0.90 (Table 22). However, if chick
survival were the only variable to be manipulated while maintaining mean egg survival

and mean clutch size, a A = 1.0 would not be attainable. Increasing chick survival to 1.0

66

 

00.0 00.0 00.0 00.0 :0.0 00.0 00.0 50.0 00.0 00.0 :0.0 0
00.00 00.00: 00.00: 00.00: 00.50: 00.00: 00.000 00.500 :5.000 00.000 00.500 :0000 Z

 

:0.5 00.0 00.0 00.0 00.0 05.00 05.00 05.00 05.00 05.00 05.00 00.00 0:
00.5 0:.0 00.0 00.0 00.0 00.0 05.00 05.00 05.00 05.00 05.00 05.00 0
00.0 00.5 50.0 00.0: 00.0: 00.0: 00.0: 00.00 00.00 00.00 00.00 00.00 0
00.0 00.5 00.0 00.0 0:.:: 0:.:: 3.: 3.: 00.00 00.00 00.00 00.00 5
00.0 0:.5 5:.0 00.0 00.0: ::.0: :0: :0: ::.0: 00.00 00.00 00.00 0
05.0 05.0 05.5 00.0 00.0: 00:: 5:.0: 5:.0: 0:.0: 5:0: 00.00 00.00 0
0:.0 00.0 00.5 00.0 00.0 00.0: 00.0: :00: :00: :00: :00: 00.00 0
00.0 00.0 05.0 00.5 0:.0 00.0: 00.: 00.0: 00.0: 00.0: 00.0: 00.0: 0
05.5 :00 00.0: 00.0: 50.0: 50.5: 50.00 05.00 05.00 05.00 05.00 05.00 0
00.0: 05.0: 00.0: 00.00 05.00 :000 00.00 0000 50.00 :000 00.00 00.00 :

:0.00 00.00 0:.00 00.00 55.00 50.50 00.00 00.00 0005 00.00 00.00 05.00: 0

 

::80> 0:80> 080> 080> 530> 030> 080> 030> 080> 030> :80> 080> 08:0
00<

 

0023:0080 .00 0000:8000 00.: :00: 0000:0003 :03 : 0:030. 000 00.0 u
000 080:0 3000 008 .0000 N 32200 0:0:00 .: 500.0 n :8>:>.Sm 000 00:00 030000 030 00003 05 0000000000 00000 .0000
08030:): E0030 0: 5:000 00:08.0 800005 0000000 05 00 00000 0000000 .:0 00000.80“: 0000:0000 .00 0:03.

67

 

 

 

00.: 00.: 00. : 00. : 00. : 00. : 00.0 00.0 00.: 0:. : :00 .0
05.000: 0:.0: :: 50.000: 50.000 05.050 :0.000 00.005 00.0 :0 05.000 50.050 00.005 :0.000 2
0:.50 00.0 00.0 00.0 00.0 05.00 05.00 05.00 05.00 05.00 05.00 00.00 0:
05.00 00.00 00.0 00.0 00.0 00.0 05.00 05.00 05.00 05.00 05.00 05.00 0
50.00 00.00 00.00 00.0: 00.0: 00.0: 00.0: 00.00 00.00 00.00 00.00 00.00 0
:0.00 00.50 00.00 05.50 0:.:: 0:.:: 0:.:: 0:.:: 00.00 00.00 00.00 00.00 5
:0.00 00.00 00.00 00.00 :0.:0 ::.0: ::.0: ::.0: ::.0: 00.00 00.00 00.00 0
0:.:0 00.00 00.00 00.00 00.05 00.00 5:.0: 5:.0: 0:.0: 5:.0: 00.00 00.00 0
50.00 05.00 00.00 00.:0 00.50 00.00 00.:0 :0.0: :00: :00: :00: 00.00 0
00.00 05.00 00.00 50.00 00.00 00.:0 00.00 50.00 00.0: 00.0: 00.0: 00.0: 0
05.00: 50.: :: 00.00: 00.00 00.05 00.50 :0.00 00.0: 50.50: 05.00 05.00 05.00 0
00.500 00.000 00.0:0 0:.000 00.55: 00.00: 00.00: 50.00: 05.0:0 05.000 00.00 00.00 :
00.0:0 0: .050 05.000 00.000 05.000 05.000 00.050 0:.0:0 00.000 00.0:0 00.000 05.00: 0
::000> 0:000> 030> 030> 530> 030> 0000> 0.00; 030> 030> 28> 0000> 000:0
00<

 

0000200000 .00 000000000 000 080 08200000 00.0 : 030B 000 .:0.0
H 00:0 080:0 0.008 000 .0000 M 003300 0:0:00 .50000 n :0>:E0m 000 00:00 0080000 0000 0000 05 0000000000 00000.85
00020:): E00000 0: 5:050 00:00”: 0000005 0000000 05 00 £08 0000000 .00 000000000 0000:0000 .:0 0309:.

68

Table 22. Manipulation of common tern egg survival (SP), chick survival (Sp), and mean
clutch size to generate A = 1.0. Mean values generated for egg survival, chick survival,
and mean clutch size were used as the basis of the manipulations.

 

 

Sp Egg SP Chick 3? Clutch Size A
sP Egg 0.9000 0.637 2.34 1.00
8,; Chick 0.4661 1.000 2.34 0.97

i Clutch Size 0.4461 0.637 4.50 1.00

69

would only yield a A = 0.97 (Table 22). Thus either egg survival, mean clutch size, or
both would also have to be increased to generate a A = 1.0. Mean clutch size would need

to increase to 4.5 eggs per nest, while maintaining mean egg and chick survival, to obtain

a A = 1.0 (Table 22).

DISCUSSION

Results from this study indicate that the common tern population in Saginaw Bay
of eastern Michigan is declining. Major threats to survival include wave inundation
(Duck Island) and predation (CDF). While other factors such as contaminants and site
characteristics (e. g., distance to neighboring nest, vegetation cover at nests) may impact
survival of common tern eggs and chicks, threats to survival due to these factors were not
readily obvious and further investigations of these impacts may be needed.

General Observations

Duck Island

The primary threat to common tern reproductive success on Duck Island was high
water levels. As Duck Island sits relatively low in the water (Figure 3), it is highly
susceptible to wave inundation and complete submersion. High water years will continue
to pose a threat to the reproductive success of the Duck Island colony. Waters of the
Great Lakes are known to fluctuate, with high water levels reported every 8 - 15 years
(Project Management Group 1989). Changes in water levels occur as a result of long-
term shifts in precipitation, inflow and outflow, and evaporation patterns, and short-term
changes in wind intensity and direction and atmospheric pressure (Project Management
Group 1989). However, although cycles in Great Lakes water levels are known, the

70

71

recent impacts of El Nino may make these cycles less predictable. Although El Ninos
generally appear every 2 - 7 years, since the early 1990's they have appeared almost
yearly (Nash 1997). El Ninos are responsible for major month-to-month variations in
climate and depending on the impacts to the jet stream can result in higher snow- and
rainfall in the northern sections of the United States, including the Great Lakes region
(Nash 1997), thus elevating lake levels. If the current trends in El Nino continue, this site
may be lost for breeding for several years to come. Because patterns in high water levels
can be tracked, it may be possible to circumvent the problem of no nesting at this site
during these years by taking steps that may prevent loss of the entire colony’s
reproductive effort in these years. Suggestions for increasing reproductive success during
high water years include the use of nesting platforms that are discussed in detail in
Chapter II of this document.

Confined Disposal Facility

The phenology of egg laying, hatching, and fledging was relatively consistent
from 1995 - 1997 with chicks hatching approximately 2 weeks later on the CDF in 1995
than in other years. While the amount of precipitation and temperature have been linked
to the timing of initial nesting (Becker et al. 1985, Burger and Gochfeld 1991), it is not
known whether weather is linked to chick hatching. In 1995, when chicks hatched
approximately 2 weeks later than in 1996 or 1997, temperatures were actually warmer
than normal and warmer than in 1996 and 1997, and precipitation at this site was normal
(NCAA 1995, 1996, 1997), therefore, these factors may not have influenced chick

hatching.

72

Researchers were able to band considerably more chicks on the CDF in 1997 than
in other years for several reasons. First, researchers improved their handling of birds and
captured birds earlier (0 - 5 days old) when they were less mobile Which increased the
number of birds banded. Second, more chicks were likely banded in 1997 due to the
increased survival of eggs which in turn increased the number of eggs that hatched (Table
14). The increased survival of eggs in 1997 may be attributed to a lower predation rate,
which is discussed below. Finally, because enclosures were used in 1997, chicks were
contained in an area which permitted researchers to more easily capture chicks

throughout the hatchling stage.

Predation

The literature indicates that chicks are most susceptible to death within 5 days of
hatching (Langham 1972, LeCroy and Collins 1972, Matteson 1988). Data collected in
this study support this finding with approximately 50% of chick deaths or last age when
known alive occurring at S 5 days old (Table 7) in 1996 and 1997. Although chicks are
semi-precocial, most stay in or near the nest for 1 - 3 days after hatching. Thus, the loss
of these chicks is likely due to predation rather than an inability of researchers to relocate
chicks.

Although not captured on film, it is suggested that predation of eggs and chicks
(Table 7) may be the primary threat to the CDF common tern colony. A predation event
by a garter snake of a common tern chick was witnessed in the colony. Although only

one removal of a chick by a garter snake was observed, it is highly probable that this was

73

not an isolated incident. Additionally, many eggs that were lost from the colony were
completely removed, that is, no egg shell fragments were left behind. The removal of
complete eggs is consistent with removal by snakes. Garter snakes have been
documented removing eggs from other common tern colonies in northern Michigan (F.
Cuthbert, unpubl. data). However, it is recognized that snakes are not the only predator
that remove entire eggs from a nest. Other potential predators that remove entire eggs
include black-crowned night herons (Collins 1970, Hunter and Morris 1976) and skunks
(Austin 1945).

Analysis of scat found within the colony also suggests predation by a fox and/or
coyote. Although coyotes have not been reported in the literature as potential predators
of cormnon tern eggs or chicks, foxes have been previously suspected in the removal of
eggs and chicks at colonies in locations along Lake Michigan (Scharf and Shugart 1991),
Minnesota (McKearnan and Cuthbert 1989), Lake Ontario (Courtney and Blokpoel
1983), and the Atlantic coast (Palmer 1941). Additionally, on four separate occasions
cached piles of dead chicks were found within the colony, a behavior attributed to foxes
(Hazard 1982). Other potential predators of common terns that may cache prey items
include mink (Courtney and Blokpoel 1983), although this species was not identified on
the CDF.

Although herring gulls and ring-billed gulls also have been suspected predators of
common tern eggs and chicks, observation of interactions with these species often did not
illicit defense responses from common terns at the CDF. Herring gull chicks and adults

were observed on several occasions walking through the common tern colony and rarely,

74

if ever, caused a defense response from nesting adults. On more than one occasion
common tern adults remained on their nests as herring gulls passed through the colony
without apparently acknowledging the presence of these intruders. Thus, it may be that
herring gulls and ring-billed gulls on the CDF pose more of a threat in the form of
competition for nesting space (Figure 5), as suggested by Courtney and Blokpoel (1980)
and Morris and Hunter (1976), than they do as predators of common tern eggs and chicks.

Of the other common avian predators of common tern eggs and chicks, including
ruddy tumstones (Farraway et al. 1986), black-crowned night herons (Collins 1970,
Morris and Hunter 1976), and great horned owls (Courtney and Blokpoel 1983), only
black-crowned night herons were observed in the colony vicinity. Anecdotal evidence
suggests that black-crowned night herons may have impacted the survival of common
tern eggs and young. As the occurrence of black-crowned night heron declined on the
CDF, overall common tern egg and chick survival increased from 1995 - 1997 (Tables 14
and 16, respectively). However, this is purely anecdotal and more research on potential
impacts to common tern survival from black-crowned night herons needs to be
conducted.

One advantage of coloniality is that it allows for synchronized nesting (i.e., all
eggs hatch at approximately the same time). Synchronized nesting produces a sudden
abundance of eggs and chicks that exceeds the daily needs of the local predators, thus
each individuals chances of being preyed upon are decreased (Lack 1968). While
synchronized nesting was evident on the CDF colony (Figures 7 - 9) it is questionable

whether or not this advantage was fully realized at this site. It may be possible that given

75

the small size of the colony in relation to the overall size of the CDF, and the number of
potential predators that it can support, that it is unlikely that common terns could produce
enough eggs and chicks to sufficiently swamp the needs of the predators for an extended
period of time. Thus, impacts due to predation at this site may remain high.

While it is possible that eggs and chicks may have been lost due to factors other
than predation, their impacts were not readily obvious. For example, contaminants may
have impacted survival of eggs and chicks, however, evidence consistent with
contaminant effects (e. g., cross bills, thin egg shells; Gilbertson et al. 1991, Lundholm
1987) was only evident when one chick was found with a crossbill, suggesting this is an
unlikely impact to survival. However, to completely rule out impacts due to
contaminants, further research needs to be completed.

Although avian cholera was reported for one bird in the colony, it is not known
what impact it had on other common terns. Due to the acute nature of the disease, few
sick birds can be seen during an outbreak (Friend 1987). However, once cholera is
present, death follows relatively immediately with birds generally dying off in large
numbers in a short time period. Although other young birds were lost at this site a large
die-off did not take place. Rather, deaths of many of these birds could likely be attributed
to predation (Table 7). However, to completely rule out the presence of avian cholera,
necropsies and virology investigations would need to be performed. Although avian
cholera was isolated in the dead bird, cholera is known to occur in many forms (K.
Miller, Nat. Wildl. Health Center, Madison, Wisconsin, pers. commun.), and this strain

may have been less aggressive than strains that can be linked to large die-offs of other

76

avian species.

While infra-red cameras produced high quality pictures of nest sites, no predation
attempts were captured on film. Although this may be interpreted as failure of this
technique in assessing predation at this site, there were many factors that may have
affected the overall success of this technique. First, given the structure and location of
the CDF, it was often difficult to run cameras on consecutive days. Often times, wave
and weather conditions did not permit transport of batteries, which powered the system,
out to the site under safe conditions. This left cameras unattended for extended periods of
time which resulted in either draining of the battery or running out of VHS tape. If
procedures could be developed to safely transport batteries and tapes to the CDF on a
daily basis, it is possible that cameras could run more regularly, thus potentially capturing
predation events on tape.

One drawback of this technique, however, is the small field of view of the camera.
During daylight hours the nest of interest and adjacent nests could be seen easily.
However, during the evening hours, only the nest of interest was fully illuminated. The
use of a camera which supports a stronger infi'a-red beam of light may address this
problem. It may be possible that predation events at nearby nests were missed during the
evening hours as camera optics were not powerful enough to illuminate the surrounding
area.

While trail monitors proved useful in identifying species that entered either end of
the colony, they were only useful in identifying potential predators to the site. A

technique such as described by Danielson et al. (1996) may prove beneficial in recording

77

predators at the nest site. They developed an inexpensive camera system that works
similar to the trail monitors but is triggered to take a picture on 35 mm film when there is
motion within a nest (i.e., an egg is removed). However, this system works only on a
single nest at one time and may not prove adequate in assessing predation of colonial

nesters (i.e., would require several camera systems to obtain a sample of nests within the

colony).

Site Characteristics

While vegetation variables fluctuated between years and within a year between
successful and unsuccessful nests, no apparent significant changes in vegetation variables
in relation to success or failure of nests are evident on the CDF. Comparisons of all nests
between 1997 and 1996 yielded a significant difference with common terns nesting closer
to a neighboring nest in 1997. It is unclear why nests were closer together in 1997 than
1996 as, the greatest number of nests in 1997 was less than in 1996 (Table 2), and the
overall size (length and width) of the colony did not change between years. This may be
an artifact of a small sample size in 1996.

Significantly higher percentage of grass canopy and lower percentage of forb
canopy in 1997 (Table 8) was likely due to the annual succession of vegetation at this
site. Basu et a1. (1978) stated that vegetation on a legume-dominated site undergoes
successional changes, eventually becoming grass dominated which may account for the
apparent shift from forb to grass canopy cover. It is unclear why percent bare ground

increased from 1996 to 1997. In general, the amount of bare ground decreases over time,

78

on an undisturbed site, due to an increase in the accumulation of dead and dying matter.
However, the apparent increase in bare ground may be an artifact of a small sample size
in 1996 or may be linked to the acidic nature of fecal matter produced by the birds that
can kill vegetation at the nest site (McBrayer et a1. 1995).

No apparent changes in vegetation variables in relation to success or failure of
nests are evident on the CDF. However, it should be noted that the percent of vegetation
cover at this site in both years at nests that successful and unsuccessful hatched and
fledged young was higher (approximately 60%; Tables 8 - 13) than the 10 - 30%
vegetation cover reported in the literature (Blokpoel et a1. 1978, Burger and Gochfeld
1991, Soots and Parnell 1975) as the preferred amount of cover at common tern nest sites.
It is possible that if the vegetation cover at this site were less, survival of eggs and chicks
would be greater than that calculated for 1995 - 1997. Matteson (1988) reported that
higher percentages of vegetation cover caused lower reproductive success and/or greater
nest abandonment of common terns in Green Bay, Wisconsin. Additionally, Palmer
(1941) indicated that while some amount of vegetation is necessary to provide refuge
from predators and thermal exposure, too much may be detrimental to survival.
Researchers on Lime Island, Michigan, were able to follow common tern chicks for
longer periods of time due to chicks becoming entangled in the vegetation as they tried to
escape, thus making them more vulnerable to predation (Millenbah and Cook 1997).
Vegetation cover at this site averaged 69.3%. Therefore, researchers are planning to
investigate whether manipulation of substrate at this site to decrease vegetation cover will

increase the overall survival of common tern eggs and chicks. A similar investigation

79

may be warranted at the CDF colony.

Successful hatchers nested significantly closer to the water’s edge than
unsuccessful hatchers in 1996 (Table 10). Austin (1948) noted that land predators, such
as the common rat, start predation attempts at one edge of the colony, working their way
to the other side, with predation occurring more fi'equently on the outer edges of the
colonies than in the middle or edges closest to the water, which may support why

successful hatchers nested closer to the water’s edge.

Assessing Reproductive Success and Determining Survival Estimates

Estimates of common tern egg survival from initiation of incubation to hatching
fluctuated from 1995 - 1997. Higher survival rates in 1997 (Table 14) may be attributed
to less frequent loss of eggs to predators. Throughout 1997 researchers noted an apparent
lack of predation of common tern eggs. Although eggs were lost due to nest
abandonment and human error, the majority of losses could be assigned to some type of
predation activity, with most nests experiencing complete loss of eggs or obvious
destruction of eggs (i.e., holes poked in eggs). In 1995 and 1996, several predation
events were evident with a loss of up to a quarter of the nests in one night. In 1997, this
type of event was not seen. Although only anecdotal, it may be possible that the increase
in survival of common tern eggs in 1997 may be related to the decrease in black-crowned
night herons on the CDF. However, further research is needed to determine the link
between black—crowned night heron occurrence and egg loss, if one truly exists.

Use of enclosures in 1997 likely did not contribute to reducing survival of

80

common tern eggs because although nests are contained within an area, eggs are not
mobile, thus having them enclosed does not permit them to be captured more frequently
or more easily. It is difficult to compare these estimates with survival estimates in the
literature as most survival estimates are in relation to adult annual survival and sub-adult
survival (Austin 1948, Austin and Austin 1956, DiConstanzo 1980, Nisbet 1978, Penning
1993).

Chick survival on the CDF was higher in 1997 than 1996 (Table 16) and may be
attributed to the use of enclosures, however, this is difficult to verify. While enclosures
may have artificially inflated survival estimates for chicks by protecting them from
predators, it is difficult to assign an increase in survivorship specifically to the use of
enclosures. Because censoring of chicks in 1996 and outside of the enclosures in 1997
was high (> 10%) estimates of chick survival at these time periods and for these locations
may be unreliable. Thus, comparisons with enclosure data may be inappropriate.

Survival of chicks was higher than survival of eggs for common terns on the CDF
from 1995 - 1997 (Tables 14 and 16). One reason for the increased survival of chicks
may be attributed to chick mobility. Eggs are not mobile so chicks may be less
susceptible to predation. Additionally, parental aggression at nests was markedly more
noticeable after hatching, which is consistent with the parental investment theory (Trivers

1974), thus parental defense of chicks may have helped reduce impacts due to predation.

81

Population Projections

Information synthesized in population projection models of the common tern
colony on the CDF suggest that the population is declining. This is consistent with
reports of other common tern populations in the Great Lakes (Courtney and Blokpoel
1983, Morris et al. 1980, Shugart and Scharf 1983). As values for sub-adult survival and
adult survival were taken from the literature, it may be possible that these projections
overestimated or underestimated the actual trends in overall population numbers. To
obtain the most accurate reflection of population trends on the CDF, values for adult and
sub-adult survival for birds at this site would be needed. Obtaining these values would
require, at a minimum, several years of intensive mark-recapture. However, the values
generated from the population projection models presented here provide a general view of
the dynamics of this population and, along with observations, offer a starting point for
understanding the threats and impacts to the continued existence of this species. One
drawback of the population projections presented here is that they were developed
assuming constant conditions (e. g., weather, predation, competition) across many years
which does not accurately reflect the changing nature of the environment. A more
intensive modeling project would need to be initiated to account for all possible
variations in population projection variables.

Manipulations of variables (mean clutch size, egg survival, and chick survival)
assist in determining where the primary impact to survival lies. These manipulations
suggest that both egg and chick survival need to be improved to increase population

numbers at this site (Table 22). Manipulating only egg or chick survival requires the

82

survivorship of these variables to be unrealistically high. Chick survival would have to
be greater than 100%, if egg survival were kept constant, to achieve a A _>_ 1. This is not
biologically possible. Egg survival would need to equal or exceed 90% and this rate of
survival would be uncharacteristic of a ground-nesting shorebird. Additionally, egg
survival appears to be the more limiting factor (Tables 5 and 22) in determining whether
the population will increase or not, thus both variables need to be manipulated. While
clutch size can also be manipulated, mean clutch size would need to be 4.5 eggs per
clutch (Table 22), if egg and chick survival were kept constant, for A 2 1. As common
terns rarely have clutches larger than 3 eggs, this option is unrealistic. Other options for
manipulating clutch size are presented below.

Egg and chick survival are the most logical variables that can be manipulated in
Michigan because little impact can be made on birds in the 1 - 2 year old age class as they
are generally not present in Michigan until they reach breeding age of 3 - 4 years old.
Additionally, little evidence of threats to adult survival were observed in Saginaw Bay
suggesting that adult survival is not limiting the population size. Although it may be
possible to have some impact on clutch size, common terns in Saginaw Bay have been
fairly consistent in being able to produce average sized clutches, as reported in the
literature for this species (Ehrlich et al. 1988), and thus this variable does not appear to be
abnormal. However, data suggest that those clutches that support more eggs (2 - 3 eggs)
have a better chance of fledging more young than those nests that only support one egg
(Table 5). This is somewhat intuitive based solely on probabilities. One option for

increasing fledging success may be to transfer eggs from larger clutches to smaller

83

clutches. That is, remove one egg from a 3 egg clutch and place it with a one egg clutch.
Data from this study suggest that clutches that support 2 eggs have the greatest success in
fledging young from a nest (Table 5). However, this suggestion may be a moot point
given that in 3 egg clutches the third egg is generally smaller and has a reduced likelihood
of survival (N isbet 1973). A more feasible option may be the complete removal of a
completed clutch from a successful nesting pair to a pair that produces only a one egg
clutch. As common terns are known to renest when initial nesting is unsuccessful, this
option may be more appropriate. However, further investigations into the feasibility of
this option are warranted.

The population projections generated in this study do not provide a positive
outlook for common terns nesting on the CDF. Additionally, previous data suggest that
the population on the CDF has decreased considerably from the 1980's when the number
of breeding pairs exceeded 600 (Michigan Natural Features Inventory (MNFI) 1996).
Although the common tern population has persisted on the CDF for almost 20 years, it is
apparent that the population is decreasing and it appears that it will continue to follow
this trend. One obvious concern, therefore, for this population relates to the concept of
minimum viable population (MVP). Defined, MVP is the minimum conditions necessary
for the long-term persistence and adaptation of a population in a given area. Has the CDF
common tern colony reached its MVP or is it actually below its MVP threshold level? If
the population is below its MVP, measures need to be implemented immediately to
increase the population as management under the current regime will not be sufficient to

support the population at a A 2 1. Thus, under the current regime, if the population is

84

below its MVP it would be impossible to increase egg and chick survival to a level that
would permit the population to increase on its own without intervention through some
additional form of intensive management. Data from this study intimate that it may be
possible that the CDF colony has reached or dropped below its MVP (Table 22),
however, further investigations and data collection would be needed to determine the

accuracy of this statement.

CHAPTER II
USE OF NESTING PLATFORMS TO INCREASE COMMON TERN
REPRODUCTIVE SUCCESS

INTRODUCTION

As common tern numbers have declined in recent years due to loss of habitat,
competition with gulls, and predation, a variety of steps have been taken to promote the
continued existence of the species. In some locations throughout Canada, management
efforts aimed at increasing the reproductive success of common terns include reducing
vegetation cover and halting succession (Dunlop et al. 1991, Morris et al. 1992),
excluding ring-billed gulls from common tern nesting areas using monofilarnent line
(Blokpoel and Tessier 1983, Blokpoel at al. 1997), and posting areas to deter visitation by
human intruders (Dunlop et al. 1991, Morris et al. 1992). However, in some instances
these efforts have proved less than successful. Thus, researchers continue to investigate
alternatives for increasing the reproductive success of common terns.

One alternative that has proved to be somewhat successful at increasing common
tern reproductive success is the use of nesting platforms. Nesting platforms or rafts have
been used in a variety of locations in the United States (Stricker 1995), Canada (Dunlop

et al. 1991), and Europe (Norman 1987, Hoeger 1988) and have been described for

85

86

common terns (Eades 1970, Turrian 1980, Norman 1987, Hoeger 1988, Dunlop et a1.
1991, Stricker 1995), Forster’s terns (Sternafosteri; Techlow 1983, D. Best, USFWS,
pers. commun.), and caspian terns (Lampman et al. 1996). Dunlop et al. (1991) found
that common tern nesting platforms provided suitable nesting sites that were not
susceptible to erosion, not occupied by ring-billed gulls, and were isolated from ground
predators and human disturbance. The platforms were built to rise with increasing water
levels, thus protecting the nests from flooding. Platforms have been deemed successful in
providing quality nest sites for common terns where otherwise quality habitat is lacking
(Norman 1987).

Michigan common terns have historically nested on islands in Wildfowl State
Game Area in Saginaw Bay that are highly susceptible to flooding by rising water levels
and wave inundation. Islands in this area that are known to support or have supported
nesting common terns include Defoe, Lone Tree, and Duck Islands (MNFI 1996; Figure
10). Lone Tree Island once supported up to 2,000 breeding pairs but the last record of
nests at this site occurred in 1977 when 25 nests were observed (MNFI 1996). Defoe
Island has supported only a small number of nests at any given time with the last
occupancy of common terns also recorded in 1977. Duck Island, however, had been
observed to support approximately 75 - 100 breeding pairs as recently as 1995 (D. Best
pers. commun., K. Millenbah pers. observ.). The threat of flooding due to high water
levels and wave inundation coupled with the recent presence of common terns make
Duck Island an ideal site for testing the effectiveness of nesting platforms in increasing

the reproductive success of the common tern. Placement of platforms at this site may

87

Sand Point
North
Island
[x Heistermann
Island
CU
' s:
Duck O
Island 0 a) BAY PORT
(’J
Maisou
Island \9
a}?
Defoe 90“
Island
Lone Tree w
Island
0
0“"
SEBEWAING

Fig 10 Location of Duck Island and other islands of Wildfowl Bay State Game Area in
Saginaw Bay, Michigan.

88

prove especially beneficial during high water years and/or during storm events. As Duck
Island represents one of the last natural nesting sites for common terns in Saginaw Bay,

evaluating the effectiveness of these platforms may prove beneficial in promoting the

long-term existence of the species.

OBJECTIVES

Specific objectives of this part of the study were to:

1) determine the effectiveness of nesting platforms in increasing common
tern reproductive success; and

2) make recommendations for managing common terns for increased

reproductive success on nesting platforms.

89

METHODS

Two nesting platforms were erected near Duck Island (Figure 11) in summer 1995
and 1997 with 3 platforms erected in 1996. No platforms were installed near the CDF as
water depths were not suitable for anchoring platforms. Platforms were anchored
approximately 5 - 10 m from Duck Island on the northeast side of the island to provide
additional shelter to the platforms from prevailing winds and wave action.

Platforms resembled those used by Dunlop et al. (1991) and Stricker (1995).

Each platform consisted of 3 sections (each section = 1.2 m X 2.4 m) held together by 2
side connector panels (each panel = 3.7 m long; Figure 12). Sections were made of one
sheet of 1.2 m X 2.4 m X 2 cm treated plywood and 5 - 8 sheets of1.2 m X 2.4 m X 2.5
cm polystyrene insulating styrofoarn. Connector panels consisted of a 5.1 cm X 10.2 cm
X 3.7 m piece of lumber with 8 - 10 evenly spaced 15 cm X 25 - 33 cm kleats (Figure
12). Because of the size and weight of each platform component, an entire platform
could be transported in the back of a pick-up truck or on a small boat (approximately 5.2
m long) with relative ease. Additionally, one platform can be easily installed by 2
individuals.

Platforms were constructed in a systematic manner to ensure that adjacent
components matched during actual assembly at a common tern nest site. Initially, one 5.1

90

91

      

Nesting
Platforms

Fig. 11. Duck Island in Saginaw Bay, Michigan and placement of common tern nesting
platforms in summer 1995 - 1997.

92

ado—Ema 93 33883 b80388 8832a 3033 nouabmsa .32 - 33 5883. camEoE
Sum Bufimam E 982 :25 Ho: com: snobs—a acumen E8 88an one he 8388852 own—5:835 .2 .wE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

93

cm X 10.2 cm X 2.4 m (end piece) piece of lumber was permanently attached to one end
of each of 2 of the 3 pieces of plywood. The end piece was attached to the plywood with
the 5.1 cm side down to provide a 10.2 cm wall on the end of the plywood. The end
piece was attached using 8 pieces of 1.0 cm threaded rod evenly spaced along the 2.4 m
side of the plywood (Figure 12). The threaded rod was cut approximately 5.1 cm longer
than the thickness of the plywood and the end piece (total length = 17.2 cm). A nut and
washer were added to either end of the threaded rod to hold the end piece and plywood
together. Next, 5 wooden kleats (15.2 cm X 24.8 cm X 2.5 cm) were evenly spaced along
the 2.4 m side of the end piece and nailed in place. Kleats were nailed in place with
galvanized nails with the 15.2 cm edge of the kleat flush with the top edge of the end
piece (Figure 12).

Next, one section with the end piece attachment (end section) was turned upside
down with the end piece end to the outside. It should be noted that platforms were
completely constructed upside down due to the ease of construction in this position.
Before proceeding, 12.7 cm were cut from the ends of both 5.1 cm X 10.2 cm X 3.7 m
pieces of lumber (side connector). One side connector was clamped to the underside of
the platform with one end butted up to the previously attached end piece. Again the 5.1
cm side was placed down on the plywood to allow for a 10.2 cm retaining wall. Ensuring.
that the 5.1 cm X 10.2 cm X 3.7 m was flush with the edge of the plywood, 3 - 4 evenly
spaced 1.0 cm holes were drilled through the plywood and side connector piece. A 1.0
cm X 17.8 cm piece of threaded rod was then inserted into each hole and held in place by

a nut and washer on either side end. The same process was repeated for the second side

94

connector piece. It should be noted that the side connector pieces described above should
not be permanently attached to the plywood as they serve as the final step in securing all
3 platform sections together. This allows the platform sections to be secured once
installed but also allows for disassembly and easy transport of platforms when not in use.

The middle section of plywood was then laid across both side connector pieces
adjacent to the first section with the 2.4 m sides together. Again, the side connectors
were clamped to the plywood, holes drilled, and threaded rod added as previously
described. Finally, the second end section (plywood with the attached end piece) was
added similar to the first 2 sections. When complete, the side facing the floor supported a
10.2 cm barrier completely surrounding the outer edge of the plywood.

Next, 5 - 8 sheets of 1.2 m X 2.4 m X 2.5 cm polystyrene were added to each
section and served as the floatation device. Eight sheets of polystyrene were used in
1995, however, 5 sheets were deemed to provide adequate floatation for platforms in
1996 and 1997. Polystyrene was held in place on each plywood piece by 0.6 cm threaded
rod. The threaded rod was cut approximately 5.1 cm longer than the thickness of the
polystyrene and plywood to allow for a washer and nut on either end of the rod. To
prevent nuts and washers from wearing through the polystyrene, a 5.1 cm X 5.1 cm
wooden washer also was used. A hole was counter sunk in the wooden washer to allow
the nut and washer to be protected from getting pushed or stripped off during transport or
once the platform was installed. It is advised to add and secure polystyrene to one section
before continuing as side edges of polystyrene may need to be shaved slightly with a

hacksaw to allow a proper fit between platform sections.

95

After the polystyrene was added and secured, 10 wooden kleats were added to
each side connector as described previously. Next, the platform was disassembled and
reassembled right side up on the ground. Prior to disassembling the platform, each
section and connector panel was marked to ensure pieces matched properly when
reassembling. Before attaching side connector panels, one additional modification was
made to each side connector panel and the threaded rod securing it to the plywood. Each
piece of threaded rod received an additional nut and washer on the plywood side opposite
the side with the polystyrene. This ensured stability of the threaded rod and held the
threaded rod in place during transport and storage of the platforms. To accommodate the
additional nut and washer, holes were counter sunk on the side of the connector panel that
meets the plywood to allow for clearance of the nut and washer. The connector panels
were then placed over the threaded rod on both sides of the platform and secured with a
nut and washer (Figure 13).

To secure anchors, 2.5 cm holes where drilled through the plywood at each comer
of the platform inside of the side connector panels. In 1995, platforms were anchored
using 3.7 m of polypropylene rope connected at each corner to 4 - 15 cm X 30 cm
concrete blocks (16 per platform). In 1996 and 1997, anchors consisted of 3.7 m of wire
rope connected to a 2 m piece of rerod (4 per platform). Wire mm was attached to the
platform using u—clamps. To attach the wire rope to the rerod, a hole was drilled near the
top of the rerod, an eyebolt secured through the hole, and the wire rope threaded through
the eyebolt. The rope was held in place using u-clamps. Each piece of rebar was

subsequently driven approximately 1.0 m into the bay floor.

96

1.0 cm
Threaded Rod

 
 
 
   
   

0.6 cm
Threaded Rod

     

/

7 //

 

Fig. 13. Diagrammatic representation of one common tern nesting platform used near
Duck Island in Saginaw Bay, Michigan summer 1995 - 1997. Illustration depicts
partially disassembled platform.

97

Platforms were covered to a depth of 5 cm with coarse gravel in 1995 and fine
sand in 1996 and 1997. Each substrate provided adequate drainage of the platform.
Seams were covered with window screening to prevent chicks and substrate from falling
through gaps. An attempt was made to install rafts prior to the arrival of terns to the
nesting site, but after ring-billed gulls had established territories. All platforms were
removed at the end of the breeding season. Observations of common terns using
platforms were made as described in Chapter I.

To assist other researchers in construction of similar nesting platforms, a list of
supplies, including quantities and costs, is provided in Table 23. All prices and quantities

are given for one platform.

98

Table 23. Supplies, including prices and quantities, for building one common tern
nesting platform used near Duck Island in Saginaw Bay, Michigan in summer 1995 -
1997. Prices are given in 1997 dollars.

 

Unit Price Total Price

 

Supply Quantity ($) ($)
1.2 m X 2.4 m X 1.0 cm treated plywood 3 25.90 77.70
Polystyrene insulating stryrofoam 15 8.43 126.45
5.1 cm X 10.2 cm X 3.7 m 2 3.99 7.98
5.1 cm X 10.2 cm X 2.4 m 2 2.29 4.58
0.6 cm nut 36 0.04 1.44
0.6 cm washer 36 0.03 1.08
5.1 cm X 5.1 cm wooden washer 18 Scrap ---
1.0 cm nut 104 0.06 6.24
1.0 cm washer 104 0.05 5.20
2 m rerod 4 5.00 20.00
3.7 m length wire rope 4 3.00 12.00
U—Clamp 8 0.39 3.12
Eyebolt 4 0.46 1 .84
0.6 cm threaded rod (length = 17.8 cm) 18 0.19 3.42
1.0 cm threaded rod (length = 17.2 cm) 40 0.36 14.40
Kleat 20 Scrap ---
Galvanized nails 100 0.07 7.00
Substrate (430,000 cm3) 1 --- 45.00
Window screening (0.5 m X 2.4 m) 2 0.20 0.40

 

TOTAL 337.85

 

RESULTS AND DISCUSSION

Two platforms were installed at Duck Island on 22 May 1995 and 12 May 1997.
Three nesting platforms were placed at the Duck Island colony on 8 May 1996.
Platforms were removed in 1995 on 7 August. On 15 May 1996 and 20 May 1997 all
platforms were missing from the area. The shorelines of Heistennann and Maisou Islands
(Figure 8) were searched for remnants of the platforms but no remains were found.
Platforms were believed to be lost to theft in 1996 and 1997 despite the presence of
identification tags. No additional platforms were built in 1996 or 1997.

Evidence of one common tern nest scrape was found on one platform on 31 May
1995. Throughout the summer, platforms required repeated repositioning as wave action
moved platforms several meters from the original location. Common terns and ring-
billed gulls used platforms primarily as loafing sites. Lack of use of platforms for nesting
by common terns in 1995 may be attributed to the inadequate anchoring of platforms with
concrete blocks and resulting instability. To prevent excessive movement of platforms,
concrete blocks were replaced with rerod that was driven into the bay floor in 1996 and
1997.

Although little data were collected on the effectiveness of using nesting platforms
to increase reproductive success of common terns in Saginaw Bay, other researchers have

99

100

had varying degrees of success (Eades 1970, Tunian 1980, Norman 1987, Hoeger 1988,
Dunlop et al. 1991, Sticker 1995) with similar nesting platforms. The most limiting
factor when using platforms at this site is the semi-remote location which makes it
difficult to monitor platforms on a regular basis thus ensuring that they are not removed
or taken without prior authorization of the researcher. Prior to further attempts at using
nesting platforms in this area, procedures need to be developed that ensure platforms are
not removed without consent of the researcher. It is recommended, however, that once
this problem is dealt with that future attempts to assess the effectiveness of nesting

platforms be initiated.

SUMMARY AND RECOMMENDATIONS

Results from this study indicate that the common tern population in Saginaw Bay
of eastern Michigan is declining. Threats to survival include wave inundation (Duck
Island) and predation (CDF). To reverse this declining trend it will be necessary to
increase the survival of common tern eggs and chicks, the primary variables that can be
impacted in Michigan. While other factors such as contaminants and site characteristics
(e.g., distance to neighboring nest, vegetation cover at nests) may impact survival of
common tern eggs and chicks, threats to survival due to these factors were not readily
obvious and further investigations of these impacts are needed to fully understand all the
potential impacts to the survival of these colonies. Although little data were collected on
the effectiveness of using nesting platforms to increase reproductive success of common
terns in Saginaw Bay they may prove useful if logistical details can be corrected.

As the common tern population in Saginaw Bay continues to decline immediate
steps need to be taken to halt this trend and to increase survival of eggs and chicks.
Predator control activities may be warranted on the CDF as a threat to survival at this site
appears to be predation by a number of different species (i.e., snakes, foxes, birds). One
option to explore may be the use of predator exclosures similar to those used by piping
plover (Charadrius melodus; F. Cuthbert, pers. commun.) researchers with the exception

101

102

plover (Charadrius melodus; F. Cuthbert, pers. commun.) researchers with the exception
of an open top. These exclosures would be identical to the enclosures used in this study,
however, their primary purpose would be to exclude predators not enclose common tern
eggs and chicks. One drawback of using this technique is that it would not exclude avian
predators, as the top of the exclosures would be open. It may be possible to fence in the
entire colony, with the subsequent removal of all terrestrial predators (i.e., snakes, foxes),
however, this does not address the potential impacts due to avian predators. Impacts due
to predation also may be addressed through an intensive predator control and removal
program on the CDF, however, a great deal of time, money, and energy would be needed
to make this a viable solution. At best it appears that efforts such as using predator
exclosures would be the most feasible solution for providing some temporary relief to the
colony from predation activities.

Options also need to be explored to increase reproductive success at sites
impacted during high water years and may include using nesting platforms. Nesting
platforms have been used for common terns in Canada and Ohio with varying degrees of
success (Dunlop et a1. 1991, Stricker 1995). Nesting platforms may be advantageous
because they are insusceptible to erosion, not occupied by ring-billed gulls, and isolated
from ground predators and human disturbance. Further investigations need to be
conducted to assess the feasibly of using these structures in Saginaw Bay including
research on site locations for placements of these structures. Another option to
investigate for low-lying nest sites may be the addition of substrate to the site to elevate

the area further above the water line. This techniques has been used in Wisconsin with

103

some success (F. Strand, Wisc. Dept. Nat. Resour., pers. commun). In addition to the
placement of substrate, a wooden frame was added around the site to hold the substrate in
place. One drawback of this technique, however, was the damage incurred to the
structure during the winter months due to freezing conditions.

Although site characteristics generally did not vary significantly between years,
and thus did not appear to alter survival from year to year, the values recorded for some
variables (i.e., cover) were greater than those reported in the literature as preferred habitat
conditions for the species. This suggests that the quality of habitat at this site may be
poor. Although no year to year differences were detected, negative impacts to survival
may have been greater than if the habitat quality more accurately reflected the values
reported in the literature as quality habitat. Thus, investigations into the quality of
habitat, specifically on the CDF, and its relationship to egg and chick survival are
necessary to account for potential impacts to survival due to poorer quality habitat.
Examples of potential impacts to survival due to poor quality habitat may include: 1) a
decrease in visibility around the nest due to encroaching vegetation which may reduce an
adults ability to identify predators thus increasing predation attempts, and 2) an increase
in vegetation cover surrounding a nest may provide too much shade thus not keeping
eggs and chicks warm resulting in death due to exposure. Manipulations to the habitat
(i.e., vegetation control and removal) may be necessary if current habitat conditions are
found to be of poor quality and resultantly negatively impact survival of common tern
eggs and chicks.

Investigations also need to be performed on the potential impacts of contaminants

104

to the CDF colony. Previous research (L. Williams, pers. commun.) has documented
elevated levels of some contaminants in Saginaw Bay. Although the objectives of this
study were not to determine the impacts due to contaminants, more research is needed to
rule this out as a factor affecting survival.

Ultimately, the best solution for ensuring the long-term survival of the Great
Lakes common tern population is to restore areas that previously supported thriving
populations of common terns. But how, where, and when this becomes a feasible
solution are up for debate. In the interim, however, it may be beneficial to investigate
alterations to existing human-made sites for increasing reproductive success and survival
of common terns. For example, researchers in Wisconsin (F. Strand, pers. commun.)
created habitat on the end of an old coal dock by filling an area (approximately 50 m2)
with fine grained sand. This area had previously supported common terns but this
modification resulted in an increase in common terns nesting on the site with a peak
number of nests reported on the site 2 years after the manipulation. A similar response
was evident on Lime Island, Michigan after substrate was added to the existing coal dock
(Millenbah and Cook 1997). It may behoove researchers to identify potential sites
throughout Michigan, particularly in Saginaw Bay, that could be manipulated to
encourage common tern nesting. Additionally, investigations into manipulations on the
Saginaw Bay CDF (e. g., vegetation control) at different locations may be warranted.
However, care should be taken when approaching this option as threats from predators
are intense at this site and may actually cause a further decline in reproductive success

than if another more suitable site (e.g., less threat due to predation) was identified.

LITERATURE CITED

LITERATURE CITED

Austin, CL. 1929. Contributions to the knowledge of the Cape Code Sterninae. Bull.
Northeastern Bird-Banding Assoc. 5:123-140.

. 1932. Further considerations to the knowledge of the Cape Cod Sterninae.
Bird-Banding 32123-139.

. 1942. The life span of the common tern. Bird-Banding 11:155-159.
. 1945. The status of the Cape Cod terns in 1944. Bird-Banding 17:21-28.

. 1948. Predation by the common rat (Rattus norvegicus) in the Cape Code
colonies of nesting common terns. Bird-Banding 14:60-65.

. 1953. The migration of the common tern (Sterna hirundo) in the western
hemisphere. Bird-Banding 24:39-55.

. and CL. Austin, Jr. 1956. Some demographic aspects of the Cape Code
population of common tern (Sterna hirundo). Bird-Banding. 27:55-66.

Basu, P.K., H.R. Jackson, and V.R. Wallen. 1978. Alfalfa decline and its cause in mixed

hay fields determined by aerial photography and ground survey. Can. J. plant Sci.
58:1041-1048.

Barrows, W.B. 1912. Michigan bird life. Mich. Agric. College Spec. Bull'., E. Lansing.
l4pp.

Becker, P.H., P. Finck and A. Anlauf. 1985. Rainfall preceding egg-laying- a factor of
breeding success in common terns (Sterna hirundo). Oecologia 65:431-436.

Bent, AC. 1921. Life histories of North American gulls and terns. Smithsonian
Institute Bull. 113: 1-337.

105

106

Blokpoel, H., P.M. Catling, and GT. Haymes. 1978. Relationship between nest sites of

common terns and vegetation on the eastern headland, Toronto outer harbor. Can.
J. Zool. 56:2061-2067.

, and WC. Scharf. 1991. Status and conservation of seabirds nesting in the Great
Lakes of North America. ICBP Tech. Publ. No.11. 26pp.

. and GD. Tessier. 1983. Monofilarnent lines exclude ring-billed gulls fi'om
traditional nesting areas. Pp. 15-19 in W.B. Jackson and B. Jackson Dodd, eds.

Proc. Ninth Bird Control Seminar, Bowling Green State Univ., Bowling Green,
Ohio. '

, , and RA. Andress. 1997. Successfirl restoration of the Ice Island

common tern colony requires on-going control of ring-billed gulls. Colonial
Waterbirds 20:98-101.

Brewer, R. 1991. Pages 33-58 in R. Brewer, G.A. McPeek, and R.J. Adams, eds. The
atlas of breeding birds of Michigan. Mich State Univ. Press, East Lansing.

Burger, J. and M. Gochfeld. 1991. The common tern: its breeding biology and social
behavior. Columbia Univ. Press, New York. 413pp.

Bumess, GP, and RD. Morris. 1992. Shelters decrease gull predation on chicks at a
common tern colony. J. Field Ornithol. 63(2):186-189.

Collins, CT. 1970. The black-crowned night heron as a predator of tern chicks. Auk
87:584-5 86.

Courtney, RA. and H. Blokpoel. 1980. Food and indicators of food availability for
common terns of the lower Great Lakes. Can. J. Zool. 58:1318-1323.

, and . 1983. Distribution and numbers of common terns on the lower
Great Lakes 1900-1980: a review. Colonial Waterbirds 6:107-120.

_—

Danielson, W.R., R.M DeGraff, and T.K. Fuller. 1996. An inexpensive compact
automatic camera system for wildlife research. J. Field Ornithol. 67(3): 414-421.

Daubenmire, RF. 1959. A canopy coverage method of vegetational analysis. Northwest
Sci. 33:43-64.

DiConstanzo, J. 1980. Population dynamics of a common tern colony. J. Field Ornithol.
51:229-243.

107

Dunlop, C.L., H. Blokpoel, and S. Jarvie. 1991. Nesting rafts as a management tool for a

declining common tern (Sterna hirundo) colony. Colonial Waterbirds 14(2):116-
1 20.

Eades, R. 1970. An artificial raft as a nesting site for terns on the Dec. Seabird Report.
Henry Burt and Son Ltd., Bedford, United Kingdom. 19pp.

Ehrlich, P.R., D.S. Dobkin and D. Wheye. 1988. The birder’s handbook. Simon and
Schuster, New York. 785 pp.

Evers, DC. 1992. A guide to Michigan's endangered wildlife. Univ. Michigan Press,
Ann Arbor. 103 pp.

. 1994. Endangered and threatened wildlife of Michigan. Univ. Michigan Press,
Ann Arbor. 412 pp.

F arraway, A., K. Thomas, and H. Blokpoel. 1986. Common tern egg predation by ruddy
tumstones. Condor 88:521-522.

Friend, M. 1987. Pages 69-82 in M. Friend, ed. Field guide to wildlife diseases: general
field procedures and disease of migratory birds. US. Dept. Inter., US. Fish
Wildl. Serv. Resourc. Publ. 167. Washington, DC.

Fox, GA. 1976. Egg shell quality: its ecological and physiological significance in a
DDE-contaminated common tern population. Wilson Bull. 88:459-477.

Gilbertson, M., T. Kubiak, J. Ludwig, and G. Fox. 1991. Great Lakes embryo mortality,
edema, and deformities syndrome (glemeds) in colonial fish-eating birds:

similarity to chick-edema disease. J. Toxicol. Environ. Health 33:455-520.

, and MA. Reynolds. 1972. Hexachlorobenzene (HCB) in eggs of common terns
in Hamilton Harbour, Ontario. Bull. Environ. Contarn. Toxicol. 12:726-732.

Hatch, J .J . 1970. Predation and piracy by gulls at a temery in Maine. Auk 87:244-254.

Haymes, GT, and H. Blokpoel. 1978. Seasonal distribution and site tenacity of the
Great Lakes common tern. Bird-Banding 49(2): 142-151.

Hays, H. 1984. Common terns pirating fish on Great Gull Island. Wilson Bull. 82:99-
100.

Hazard, EB. 1982. The mammals of Minnesota. Univ. Minn. Press, Minneapolis.
280pp.

108

Hodgkins, M. 1993. 1993 Census of wildlife use of Saginaw Bay Confined Disposal
Facility, Bay County, Michigan. US. Fish and Wildl. Serv., East Lansing, MI.
13 pp.

Hoeger, S. 1988. Schwimmkarnpen: Germany’s artificial floating islands. J. Soil Water
Conserv. 4:304-306.

Hunter, RA. and R. D. Morris. 1976. Nocturnal predation by a black-crowned night
heron at a common tern colony. Auk 93:629—633.

LaBarr, MS. 1995. Vermont common tern recovery plan (draft). Nongame and natural
heritage program. Vermont Fish and Wildlife Dept. 33 pp.

Lack, D. 1968. Ecological adaptations for breeding birds. Western Print. Serv., Ltd.
Bristol, England. 409pp.

Lampman, K.P., ME. Taylor, and H. Blokpoel. 1996. Caspian terns (Sterna caspia)
breed successfully on nesting rafts. Colonial Waterbirds. 19:135-138.

Langham, N.P.E. 197 2. Chick survival in terns (Sterna spp.) with particular reference to
the common tern. J. Anim. Ecol. 385-395.

LeCroy, M., and CT. Collins. 1972. Growth and survival of roseate and common tern
chicks. Auk 89:595-611.

Ludwig, J .P. 1962. A survey of the gull and tern populations of Lakes Huron, Michigan,
and Superior. Jack-pine Warbler 40(4):104-119.

. 1991. Ring-billed gull. Pages 216-217 in R. Brewer, G.A. McPeek, and R.J.
Adams, eds. The atlas of breeding birds of Michigan. Mich State Univ. Press,
East Lansing.

Lundholm, E. 1987. Thinning of eggshells in birds by DDE: mode of action on the
eggshell gland. Comp. Biochem. Physiol. 88(1):1-22.

Lyon, WI. 1927. The fox snake feeds on eggs of the common tern. Wilson Bull.
39:186.

Matteson, SW. 1988. Wisconsin common tern recovery plan. Wisconsin Endangered
Species Report 41. Wisc. Dept. Nat. Resourc. 82pp.

Mayfield, HM. 1961. Nesting success calculated from exposure. Wilson Bull. 73:255-
261.

109
McBrayer, L.D., J .M. Arnold, T.D. White, and SW. Calhoun. 1995. Reproductive status

of common tern colonies on the Niagara Frontier. Final Report, FYS 1993 - 1995.
State Univ. of New York, Buffalo. 37pp.

McKearnan, J .E., and F.J. Cuthbert. 1989. Status and breeding of common terns in
Minnesota. Colonial Waterbirds. 12: 1 85-190.

Michigan Natural Features Inventory. 1996. Common tern database. Lansing, Michigan

Millenbah, K.F., and BS. Cook. 1997. Factors affecting the reproductive ecology of the
common tern (Sterna hirundo) on Lime Island in the eastern upper peninsula of
Michigan. Annual rep. to the Mich. Dept. Nat. Resourc. 12pp.

Morris, R.D., H. Blokpoel, and G. D. Tessier. 1992. Management efforts for the

conservation of common tern Sterna hirundo colonies in the Great Lakes: two
case histories. Biolog. Conserv. 6027-14.

, and RA. Hunter. 1976. Factors influencing desertion of colony sites by
common terns (Sterna hirundo). Can. Field-Nat. 90:137-143.

, , and J .F. McElman. 1976. Factors affecting the reproductive success of

common tern colonies on the lower Great Lakes during the summer of 1972. Can.
J. Zool. 54: 1 850-1 862.

, I.R. Kirkham, and J .W. Chardine. 1980. Management of a declining tern
colony. J. Wildl. Manage. 44:241-245.

, and DA. Wiggins. 1986. Ruddy tumstones, great horned owls, and egg loss
from common tern clutches. Wilson Bull. 98(1): 101-109.

Nash, J .M. 1997. Is it the El Nino of the century? Time. 31 :56-58.

NOAA,1995. Climatological data: Michigan. Dept. Commerce. 110(5- 7):4-5.
___,1996. Climatological data: Michigan. Dept. Commerce. 111(5- 7):4-5.
_, 1997. Climatological data: Michigan. Dept. Commerce. 112(5- 7):4-5.

Nisbet, I.C.T. 1972. Disaster year for terns. Man and Nature. Massachusetts Audubon
Soc. Bull. 12:16-21.

. 1973. Courtship feeding, egg size, and breeding success in common terns.
Nature 241:141-142.

llO

. 1978. Dependence of fledging success on egg size, parental performance and
egg composition among common and roseate terns. Ibis 120:207-215.

, and W.H. Drury. 1972. Measuring breeding success in common and roseate
terns. Bird-Banding 43297-106.

, and M.J. Welton. 1984. Seasonal variations in breeding success of common
terns: consequences of predation. Condor 86:53-60.

Norman, D. 1987. Are common terns successful at a man-made nesting site?, Ringing &
Migration 8:7-10.

Palmer, RS. 1941. A behavior study of the common tern. Proc. Boston Soc. Nat. Hist.
42: 1-1 19.

Penning, W.L. 1993. Common terns in western Lake Superior: history, management,
and population modeling. MS. Thesis, Univ. Minnesota, Minneapolis. 61pp.

, and F.J. Cuthbert. 1993. The history of colonial waterbird management in
Duluth-Superior Harbor 1937-1990. Loon 65:163-174.

Project Management Group. 1989. Living with the lakes: challenges and opportunities.
Prog. Rep. to Intern. Joint Commission. 108pp.

Scharf, WC. 1981. The significance of deteriorating man-made island habitats to
common terns and ring-billed gulls in the St. Mary’s River, Michigan. Colonial
Waterbirds 4:445-459.

. 1991. Common tern. Pages 221-222 in R. Brewer, G.A. McPeek, and R.J.
Adams, eds. The atlas of breeding birds of Michigan. Mich State Univ. Press,
East Lansing.

, and G.W. Shugart. 1985. Population sizes and status recommendation for
double-crested connorants, black-crowned night herons, caspian terns, common
terns, and Forster's terns in the Michigan Great Lakes in 1985. Rep. to the
Michigan Dept. of Nat. Resour.

, and . 1991. Gulls, terns, and corrnorants of the upper U.S. Great Lakes -
1989. Interim rep. to US. Fish Wildl. Serv. Contract No. 14-16-0009-89-006.
53 pp.

Shields, M.A. and T. W. Townsend. 1985. Nesting of Ohio’s endangered common tern.
Ohio J. Sci. 85:45-49.

111

Shugart, G.W., and WC. Scharf. 1983. Common tern in the northern Great Lakes:
current status and population trends. J. Field. Ornithol. 54(2):160-169.

Snedecor, G.W. and W.G. Cochran. 1980. Statistical Methods, Seventh Edition, The
Iowa State Univ. Press, Ames. 507pp.

Soots, RP. and J .F. Parnell. 1975. Ecological succession of breeding birds in relation to
plant success on dredge islands in North Carolina. Univ. North Carolina Sea
Grant Program Publ. UNC-5G-75-27. 91pp.

Stricker, NJ. 1995. The common tern in Ohio: nesting ecology and management. MS.
Thesis, Ohio State University. 108pp.

Techlow, AF. 1983. Forster’s terns nesting platform study. Dept. Nat. Resour. Prog.
Rep., Madison, WI

Trivers, R.L. 1974. Parent-offspring conflict. Am. Zoolo. 14(1):249-264.

Turrian, F. 1980. Notes breves et faits divers. Nos Oiseaux 35:341-343.

Weseloh, D.V., T.W. Custer, and BM. Baune. 1989. Organochlorine contaminants in
eggs of common terns from the Canadian Great Lakes, 1981. Environ. Pollution

59:141-160.

Wiggins, D.A., R.D. Morris, I.C.T. Nisbet, and T.W. Custer. 1984. Occurrence and
timing of second clutches in common terns. Auk 101:281-287.

Winterstein, SR, and K. F. Millenbah. 1996. Factors affecting the reproductive ecology
of the common tern (Sterna hirundo) in Saginaw Bay of eastern Michigan. Ann.
rep. to the US. Fish Wildl. Serv. E. Lansing, Michigan l4pp.

"Illlll'lllllllllllllll“