Habitat suitability index models: Forster's tern (breeding)--Gulf and Atlantic coasts

              BIOLOGICAL REPORT 8 2 [ 10.13 1)
FEBRUARY 1987
HABITAT SUITABILITY INDEX MODELS:
FORSTER’S TERN (BREEDING)-GULF
AND ATLANTIC COASTS
Fish and Wildlife Service
U.S. Department of the Interior
Biological Report 82(10.131)
February 1937
HABITAT SUlTABILITY INDEY MODELS: FORSTER'S
TERN (BREEDING)--GULF AND ii:ANT!C COASTS
b.Y
Richard P. Martin
ana
Phil'ip J. Zwank
Louisiana Cooperative Fish and Wildlife Research Unit
Louisiana State University
Baton Rouge, LA 70803
Project Officer
Rebecca Howaru
National Wetlands Research Center
U.S. Fish and Wildlife Service
1010 Gause Boulevard
Slidell, LA 70458
Performed for
National Wetlands Research Center
Research and Development
Fish and Wildlife Service
U.S. Department of the Interior
Washington, DC 20240
This report may be cited as:
Martin, R.P., and P.J. Zwank. 1987. Habitat suitability index models:
Forster's tern (breeding)--Gulf and Atlantic coasts. U.S. i-lsh Wildl. Serv.
Biol. Rep. 82(iO.131). 21 pp.
PREFACE
The habitat suitability index (HSI) model in this report on the Forster's
tern is intended for use in the U.S. Fish and Wildlife Service's (1980)
habitat evaluation procedures for impact assessment and habitat management.
The model was developed from a review and synthesis of existing information
and is scaled to produce an index of habitat suitability between 0 (unsuitable
habitat) and 1 (optimally suitable habitat). Assumptions used to develop the
model and guidelines for model applications, including methods for measuring
model variables, are described.
This model is a hypothesis of species habitat relationships, not a
statement of proven cause and effect. The model has not been field-tested.
For this reason, the U.S. Fish and Wildlife Service encourages model users to
convey comments and suggestions that may help increase the utility and
effectiveness of this habitat-based approach to fish and wildlife management.
Please send any comments or suggestions you may have to the following address.
lnformation Transfer Specialist
National Wetlands Research Center
U.S. Fish and Wildlife Service
1010 Gause Boulevard
Slidell, LA 70458
iii
CONTENTS
Page
PREFACE ...........................................................
ACKNOWLEDGMENTS ...................................................
INTRODUCTION ......................................................
Distribution ....................................................
Life History Overview ...........................................
SPECIFIC HABITAT REQUIREMENTS .....................................
food and Foraging Habitat .......................................
Nesting Cover ...................................................
Water ...........................................................
Special Considerations ..........................................
HABITAT SUITABILITY INDEX (HSI) MODEL .............................
Model Applicability .............................................
Model Description ...............................................
Suitability Index (SIj Graphs for Model Variables ...............
HSI Determination ...............................................
Field Use of Models .............................................
Interpreting Model Outputs ......................................
REFERENCES ........................................................
APPENDIX ..........................................................
Definitions of SI Model Classes .................................
iii
vi
1
1
2
3
;:
5"
6
6
1;
13
15
16
::
21
V
ACRNQWLEDGMENTS
The ~lr~nat-ive and habitat suitability index model for the Forster's tern
was reviewed by the following individuals who have some knowledge of the
nesting requirements of this species: J. Burger, Rutgers-The State
University, New Brunswick, New Jersey; J. Parnell, University of North
Carolina, Wilmington; and G. Peterson, Center for Wetland Resources, Louisiana
State University, Baton Rouge. Evaluation of model structure and functional
relationships were provided by personnel in the U.S. Fish and Wildlife Service
Ecological Services Offices at Lafayette, Louisiana, Corpus Christi, Texas,
and Annapolis, Maryland, and by R. Hamilton and D. Pashley, School of
Forestry, Wildlife and Fisheries, Louisiana State University, Baton Rouge.
Special thanks go to D. Hewitt for typing the draft document. Funding
for the model development and publication was provided by the U.S. Fish and
Wildlife Service. Patrick Lynch illustrated the cover.
vi
FORSTER'S TERN (Sterna forsteri)
INTRODUCTION
Distribution
The nesting range of Forster's terns hosts three allopatric breeding
populations. The first and most important breeding area, in terms of the
number of nesting pairs, includes the western gulf coast from the
Louisiana-Mississippi border to northern Tamaulipas, Mexico (American
Ornithologists' Union [AOU] 1983). In addition, small numbers of Forster's
terns have nested in Mobile County, Alabama (Imhof 1976). Although this
species has not been recorded nesting in Mississippi (J. Jackson, Mississippi
State University, Starkville; pers. comm.), it is observed in the coastal
regions of that State every summer, and several thousand nest in adjacent
Louisiana (Portnoy 1977; Clapp et al. 1983). The two largest colonies of
Forster's terns documented in the literature were both in Louisiana: one of
2,750 pairs in Lake Borgne on the Louisiana-Mississippi border and one of
2,263 pairs in Calcasieu Lake (Portnoy 1977).
The second nesting region extends from Long Island, New York, south along
the Atlantic coast to Cape Lookout, North Carolina (Erwin 1979; Portnoy et al.
1981; Parnell and McCrimmon 1984). About 6,000 birds breed in this region
annually (Erwin 1979; Portnoy et al. 1981). Louisiana is the most important
nesting area for Forster's terns along the Atlantic or Gulf of Mexico coasts
in terms of total number of breeding birds and mean colony size (Table 1).
The third and largest breeding area extends from southeastern British
Columbia east to south-central Manitoba, south to central California, northern
Utah and Colorado, northwestern Kansas, and northern Iowa, Illinois and
Indiana (AOU 1983). The number of nesting birds iin this area has not been
documented, but the largest concentrations of nesting birds are apparently in
the San Francisco Bay region and Manitoba (Gerrard and Whitfield 1971; Gill
and Mewaldt 1979).
Some northward and westward dispersal of Forster's terns is apparent
prior to the southward migration. The southward migration is primarily
through interior North America (AOU 1983). Forster's terns winter along
coastal areas from central California south to Costa Rica on the Pacific
coast, along the Atlantic-gulf coast from Virginia to Costa Rica, and in the
Bahamas and Greater Antilles (AOU 1983). The greatest concentrations of
wintering terns occur along the Texas and Louisiana coasts, and coastal
southern California (Rubega et al. 1984; Drennan et al. 1985).
1
Table 1. Recent distribution of Forster's tern colonies along the Atlantic
and gulf coasts of the United States.
State Colonies Pairs References
New Jersey 6 463
Maryland 7 520
Virginia 21 1,095
North Carolina 13 931
Alabama a a
Louisiana 34 9,608
Texas 64 1,773
Total 145 14,390
Erwin 1979
Erwin 1979
Erwin 1979
Parnell and McCrimmon 1984
Imhof 1976
Portnoy 1977
Texas Colonial
Waterbird Society 1982
a A few pairs may nest annually but adequate data are not available. Oberholser (1974) and Blacklock et al. (1978, in Clapp et al. 1983)
indicated that the number of Forster's terns nesting on the Texas coast has
declined since the mid-1940's because of the development of the larger
offshore islands and possibly because of bioaccumulation of DDT and PCB.
Recent investigators, however, have postulated that the number of nesting
terns has been stable (Texas Colonial Waterbird Society 1982). This
maintenance of population size may be due to increased availability of
suitable nesting areas on dredged spoil islands.1 Because of the limited
availability of quantitative historical data, the trend for nesting Forster's
terns in Louisiana cannot be determined.
Life History Overview
Forster's terns along the Atlantic and gulf coasts tend to nest in
single-species colonies of 10 to several hundred birds in saline or brackish
marshes on near-shore islands (Erwin 1979; Portnoy et al. 1981). Only a small
proportion of the gulf coast population nests in mainland marshes: 3.1% of
the colonies and 0.5% of the breeding birds in Louisiana, and 8.5% of the
colonies and 2.8% of the breeding,birds in Texas. In inland areas, floating
mats of aquatic vegetation ("pop-ups") appear to provide important nesting
2 substrate (P. Yakupsak, Lacassine National Wildlife Refuge, Louisiana; pers.
comm.).
The majority of adult terns arrive at the nesting areas during April
(Erwin 1979). Along the Atlantic and gulf coasts they usually lay eggs in
April and May, and most nesting is completed by the end of July (Kopman 1907,
1908; Bent 1910; Oberholser 1938; Portnoy 1977; Erwin 1979; Parnell and
soots 1979). Chaney et al. (1978) calculated the mean clutch size of this
species to be about 2 eggs per nest for 92 nests examined along the Texas
coast in 1977. Few published data are available on the incubation period of
Forster's terns along the gulf coast, but Chaney et al. (1978) calculated the
period to be 23 days for 6 nests examined. The mean length of time from the
initiation of hatching to emergence is about 2.8 days for each egg (McNicholl
1983). Fledging apparently occurs at approximately 40 days, but varies with
prey abundance (Clapp et al. 1983).
Data on the extent of renesting by Forster's terns are not available for
the gulf coast colonies, but Parnell and Soots (1979) and J. Burger
(Rutgers-The State University, New Brunswick, New Jersey; pers. comm) have
recorded renesting after storm damage in North Carolina and New Jersey.
Bergman et al. (1970) postulated that some renesting may occur in the interior
population if the original nest is destroyed.
SPECIFIC HABITAT REQUIREMENTS
Food and Foraging Habitat
Forster's terns tend to forage in protected shallow bays and lagoons near
the nesting colony (Oberholser 1974; Baltz et al. 1979; Chapman 1984). When
foraging over open water away from the coastal marsh, they concentrate in the
zone between the foreshore and the first bar (Chapman 1984). Forster's terns
selected water less than 1 m deep during two foraging ecology studies in
coastal California (Salt and Willard 1971; Baltz et al. 1979).
Few data on prey size and type are available; however, the diet
apparently consists primarily of small fish, crustaceans, and aquatic insects
(McAtee and Beal 1912; Salt and Willard 1971; Baltz et al. 1979). Menhaden
(Brevoortia tyrannus) and silvery anchovies (Engraulis eurystole) were
important prey items in the late fall and winter in birds collected from
coastal South Carolina, Georgia, and Florida (McAtee and Beal 1912). Baltz et
al. (1979) determined that juvenile shiner perch (C mato aster a
northern anchovy (Engraulis mordax) were the predo+m inant $%?%t!da?:
stomachs of birds collected at Elkhorn Slough, Monterey County, California.
Prey size ranges from 1 to 10 cm with a mean of about 1.5 cm (Baltz et al.
1979; Salt and Willard 1971).
Forster's terns are apparently not limited by foraging habitat or prey
availability. They nest in or near brackish
important nursery areas for many marine fishes
dependable source of appropriate-sized prey.
3
and salt marshes, which are
and crustaceans and provide a
Nesting Cover
Formerly, nesting colonies were established on natural marshy islands
that were generally within 2 km of the mainland (Kopman 1907, 1908; Job
1908; Bent 1910). The development of artificial islands by deposition of
dredged spoil has created additional suitable nesting areas (Portnoy 1977;
Chaney et al. 1978; Texas Colonial Waterbird Society 1982). Almost all of the
recent colonies in Louisiana were located on natural islands: only 0.8% of
the breeding birds were found on dredged spoil islands in 1976 (Portnoy 1977).
Artificial islands are apparently more important as colony sites in Texas
because of the lack of undisturbed natural islands; 55.6% of the colonies
examined along the Texas coast during 1973-80 were located on artificial
islands (Texas Colonial Waterbird Society 1982). Dredged spoil islands in
saline waters along the upper gulf coast are initially vegetated by smooth
cordgrass (Spartina alterniflora), and sufficient vegetative cover for the
establishment of nesting colonies of Forster's terns should develop within 2
to 5 years (Chaney et al-.
Forster's terns in
marshy islands vegetated
(saltmeadow cordgrass)
Oberholser 1974; Portnoy
1978; Lewis and Lewis 1978).
coastal colonies characteristically nest on small
with 5. alterniflora in saline marsh and 5. patens
in brackish marsh (Bent 1921; Oberholser 1938;
1977; Parnell and Soots 1979; Texas Colonial Water-bird
Society 1982; Clapp et al. 1983; G. Peterson, Center for Wetland
Resources, Louisiana State University, Baton Rouge, pers. comm.).
The availability of wind- and wave-accumulated mats of vegetation or wood
(wrack) within Spartina spp. marsh is apparently one of the most important
factors influencing the establishment of tern colonies, although Forster's
terns may occasionally nest in marshes that lack wrack. Wrack is usually
deposited above the mean high-tide line by storm tides and wind during the
winter, and provides quality nesting substrate the following spring. The
presence of wrack was noted by most authors describing the substrate of
Forster's tern nests in the eastern and southern United States (Job 1908; Bent
1921; Oberholser 1938; Chaney et al. 1978; Parnell and Soots 1979; Texas
Colonial Waterbird Society 1982; G. Peterson, pers. comm.). Because wrack
normally accumulates parallel to the island shore, tern colonies tend to be
long and narrow. Wrack consisting of aquatic vegetation such as Zostera
marina is usually only 5 to 10 cm in depth, while wrack composed of coarse
material, such as S. alterniflora culms, regularly accumulates to 0.5 m (J.
Parnell, University of North Carolina; pers. comm.). Apparently, the
important factor determining colony establishment is a quantity of wrack
sufficient to completely cover the underlying marsh vegetation, thus creating
an open ridge within the marsh suitable for nest placement; Forster's terns
will not utilize wrack deposits that have a large amount of marsh vegetation
penetrating the mat (J. Parnell, pers. comm.). Wide wrack deposits are
probably superior to narrow deposits because nests are less readily located by
predators using the elevated ridge as a travel lane. Bleached wrack is
particularly suitable as a nes.t substrate as it provides good camouflage for
tern eqqs and chicks (Burner and Lesser 1978). Because tern nests are
generally shallow platform;
elevation afforded by nesting
potential wave and storm tide
(Bent 1921; Chancy-et al. 1978), the increased
on wrack protects the nest and contents from
damage, which is apparently the primary source
4
of nest failure in Texas and North Carolina (Chaney et al. 1978; Parnell and
soots 1979). Occasionally, terns will establish small nesting colonies on
sand or shell beaches if wrack deposits are present and vegetative cover is
available nearby (Oberholser 1938; Parnell and Soots 1979). Common terns that
used wrack as a nest substrate had greater nesting
nested on the ground (Burger and Lesser 1978).
success than pairs that
Vegetative density does not appear to be important except that ground
cover must be sufficient to trap a dense mat of wrack for a nesting substrate
and to provide some protective cover for young chicks (Burger and Lesser 1978;
J. Parnell, pers. comm.).
Water
The physiological water requirement of Forster's terns
within the saline and brackish habitats surrounding the nest
is probably met
need not be considered as a factor of habitat suitability (We1
ing colonies and
ty 1982).
Special Considerations
Predation. Some characteristics of islands are related to their
potential to sustain terrestrial predators. Rats (Rattus spp.), raccoons
(Procyon lotor) and mink (Mustela vison) are predators of nesting terns
(Provost 1947; Austin 1948; Sprunt 1948; G. Peterson, pers. comm.). Thus, the
perceived potential
predators is
of an island to support a viable population of terrestrial
important when estimating the suitability of the site as tern
nestiny habitat. Island elevation, island size, and distance from the
mainland are probably the most important variables determining successful
predator colonization.
Islands with little relief (less than 0.5 m maximum elevation) tend to
have low vegetative diversity and are periodically flooded, so probably would
not support viable populations of terrestrial predators. As island size
increases the maximum elevation tends to increase, thereby increasing
vegetative diversity and the potential to support predators. Islands with
some topographic variability also provide refuge for predators from high
tides; however, Chaney et al. (1978) postulated that islands less than 20 ha
in size, even if they had some topographical relief and vegetative diversity,
would not support large predators indefinitely.
In Texas, the largest tern colonies were found on small islands up to 1
ha in area (Texas Colonial Waterbird Society 1982). Similar quantitative data
are not available for Louisiana, but Kopman (1907) stated that colonies tended
to be established on "small" islands and G. Peterson (pers. comm.) and S.
Cardiff (Museum of Zoology, Louisiana State University, Baton Rouge; pers.
comm.) stated that the Forster's tern colonies they examined were on islands 1
ha or less in area.
As the distance from the colony island to the mainland increases, the
probability of predator colonization decreases (McArthur and Wilson 1967). In
this model, "mainland" refers to contiguous habitat greater than 20 ha in
area.
5
Avian predators, especially herring gulls (Larus argentatus), can cause
serious egg or chick loss in tern colonies along the northern Atlantic coast.
However, the smaller laughing gull (Larus atricilla) is the common nesting gull along the southern Atlantic andgurfcoast in the United States; and
predation by this species does not appear to be an important limiting factor
of nesting terns (Portnoy 1977; Burger and Lesser 1978).
Disturbance. Human disturbance and development of traditional nesting
areas are often cited as reasons for tern colony abandonment (Davis 1965;
Gochfeld 1974, 1976; Portnoy 1977). Development of the larger offshore
islands along the Texas coast is considered the reason Forster's terns
abandoned some traditional colony sites (Oberholser 1974). The extremely
large colony of 2,750 pairs of nesting terns in Lake Borgne in Louisiana was
deserted with no apparent fledging of young after the colony was intentionally
disturbed by humans, and some adult birds were apparently shot (Portnoy 1977).
Even if human disturbance does not cause complete colony abandonment, it can
cause the adults to leave temporarily, exposing the eggs and young to heat
stress and predation (Portnoy 1977). Only 1 of 32 colonies in Louisiana and
14' of 64 colonies in Texas were noted as being included in State, Federal, or
private refuges and thereby protected from most disturbances (Portnoy 1977;
Texas Colonial Waterbird Society 1982).
Although we could not find documented reports of Forster's tern
susceptibility to oil pollution, the potential for adverse effects is great
because this species nests and roosts at the high-tide line (Chapman 1984).
In addition, the method of feeding (aerial plunging) brings the birds into
regular contact with the water surface, Because the Louisiana and Texas
coasts are important oil production areas and because a large proportion of
the world,'s breeding and wintering Forster's terns are found there, special attention should be given to important nesting and roosting islands should a
major spill occur.
HABITAT SUITABILITY INDEX (HSI) MODEL
Model Applicability
Geographic area and season. The habitat suitability index model in this
report was developed for application within the normal breeding range of the
Forster's tern along the gulf coast of Louisiana and Texas. There are
apparently few differences in nesting habitat requirements along the Atlantic
coast; therefore, the HSI model may also be used to evaluate potential habitat
in coastal New Jersey, Maryland, Virginia, and North Carolina. The HSI model
was developed to evaluate Forster's tern habitat during the breeding season
(March through July).
Cover types. Forster's terns are extremely selective in their choice of
nesting habitat, with a vast majority of the colonies established in saline or
brackish marsh habitats classified as "estuarine intertidal persistent
emergent" by Cowardin et al. (1979).
6 Minimum habitat area. Minimum habitat area is defined as the minimum
amount of contiguous habitat that fulfills all life requisites for a species;
however, the minimum habitat area principle is not readily applicable to
colonial nesting species that travel to a food source. Published data on the
minimum area required by Forster's terns were not found. If quantitative data
on minimum area were collected in the future and the size of the site under
evaluation is less than the minimum, the HSI for this species would be zero.
Verification level. The output of this HSI model is an index between 0
and 1.0 that is believed to reflect habitat potential for Forster's terns.
Earlier drafts of this model were reviewea by J. Burger, Department of
Biology, Rutgers-The State University, New Brunswick, New Jersey; J. Parnell,
Department of Biological Sciences, University of North Carolina, Wilmington;
G. Peterson, Center for Wetland Resources, Louisiana State University, Baton
Rouge. Personnel in the U.S. Fish and Wildlife Service Ecological Services
Offices at Lafayette, Louisiana; Corpus Christi, Texas; and Annapolis,
Maryland, also reviewed the model. This model has not been field-tested.
Model Description
Overview. Habitat suitability for Forster's terns is dependent on
several factors. Five factors that influence nesting habitat quality are
incorporated in this model: two variables that describe cover, two variables
describing the physical characteristics of the nesting island, and the amount
of human disturbance in or near the study area. The relationships of habitat
variables to the nesting HSI value are illustrated in Figure 1.
Although waves and storm tides can cause serious nest loss, they are
environmental factors and not a component of the habitat per se, so are not
included in this model. An additional factor that may be important when
assessing habitat quality, although quantitative data are lacking, is the
influence of tidal amplitude on colony establishment and success. Wrack
deposition on low marshy islands would be reduced in areas exposed to great
fluctuations in tidal range. In those areas, wrack would tend to accumulate
on islands with some topographic relief. Colonies established in areas of
large tidal amplitude would also have an increased potential for colony damage
by waves and storm tides. In North Carolina, Forster's terns only nest in
Pamlico and Cove Sounds, which are exposed to relatively little tidal
fluctuation, indicating that colony site selection may be influenced by tidal
amplitude (J. Parnell, pers. comm.).
The availability of water and prey is assumed to be non-limiting within
the breeding range of Forster's terns; therefore, those variables are not
considered in the model.
The HSI model uses the known habitat requirements of Forster's terns to
assess the potential suitability of a given site to support a nesting colony.
The assumptions for each variable and the associated primary data sources are
presented in Table 2. The following section explains the logic used in the
formulation of the HSI equation.
7
Habitat variable Component
"1
Percent of island covered
with S. alterniflora or
S_. paTens
V2 Wrack quality
-Nesting cover
"3 Island size
"4 Distance of island from Island characteristics HSI
mainland or other island
=- 20 ha in area
cn
"5 Disturbance level Disturbance
Figure 1. Relationship of habitat variables to the habitat suitability index (HSI) for
nesting Forster's terns.
Table 2. Variable sources and assumptions for Forster's tern suitability
indices.
Variable Source Assumption
v1
5
V3
Oberholser 1938
Chaney et al. 1978
Burger and Lesser 1978
Marsh vegetation (Spartina
alterniflora or 2. patens)
is necessary for wrack
accumulation, nest support,
and chick cover.
Oberholser 1938
Portnoy 1977
Texas Co1 onial Water-bird
Society 1982
Parnell and Soots 1979
J. Burger, pers. comm.
Optimal sites contain an
abundance of wrack for nest
substrate.
Chaney et al. 1978
J. Parnell, pers. comm.
G. Peterson, pers. comm.
The relative protection from
wave damage afforded a colony
and the potential of an
island to support terrestrial
predators are dependent on
island size.
% McArthur and Wilson 1967 The potential for predator
Portnoy 1977 colonization decreases as the
Texas Colonial Waterbird island becomes increasingly
Society 1982 isolated from the mainland.
v5 Davis 1965
Gochfeld 1974
Gochfeld 1976
Portnoy 1977
Human disturbance is
detrimental to tern colony
establishment and nesting
success.
Cover component. Vegetative cover is an important component of Forster's
tern nesting habitat. Coastal nesting colonies are normally located in stands
of S. alterniflora in saline marsh and, occasionally, 5. patens in brackish
march. Spartina spp. are important components of the nesting habitat because
they trap and are a major component of wrack and provide escape cover for
chicks.
sites.
Decumbent patches of SpartiFa spp. also provide secondary nesting
This model assumes that optimal nesting habitat is found on low,
periodically flooded saltmarsh islands vegetated with near-monotypic stands of
5. alternifiora with canopy cover of at least 25% (VI).
_
9
Islands with some elevational relief have increased vegetative diversity
when compared with low, regularly flooded islands (Lewis and Lewis 1978).
Because vertebrate species diversity is correlated with increased spatial
heterogeneity and plant species diversity (MacArthur and MacArthur 1961), the
potential for an island to sustain a viable population of one or more
predators increases with an increase in elevation. Islands with a maximum
elevation less than 0.5 m above mean high tide are normally only vegetated
with Spartina spp. and do not support woody. vegetation. Because vegetative
composition is dependent on island topography, the cover variable, VI, in-cludes
an elevational component.
The presence of wrack (V2) is very important for colony establishment.
Quantifying this variable is difficult, so it should be measured on a relative
scale. Optimum habitat contains extensive wrack deposits that completely
cover the underlying marsh vegetation and provide a substrate that elevates
the nest above mean high tide. The importance of annual replenishment of
wrack is supported by the observation that common terns that used wrack as a
nest substrate generally did not use wrack from previous years due to
compaction and decomposition (Burger and Lesser 1978).
Island characteristics. Two variables--size and distance from the
mainland--are used to quantify the potential of islands to support tern
predators and provide protection from wave damage to the colony. As the size
of the island (V3) increases, the potential to support a viable predator
population increases, primarily because of increased topographic diversity.
Islands greater than 70 ha are relatively unsuitable for the establishment of
nesting colonies because they potentially support large predators. The
minimum island size required to support Rattus spp. is not known, although
these predators probably do not persist on small, periodically flooded, marshy
islands. However, rice rats (Oryzomys palustrus) are widely distributed in
marsh habitats along the Atlantic and gulf coasts (Wolfe 1982), and although
they can be an important predator of marsh wren (Cistothorus palustris) eggs
(Kale 1965), we could not find any published data documenting rice rat
predation of tern eggs. It is assumed that the SI score for island size
is low (0.2) for islands up to 0.1 ha in size, is equal to 1.0 between 0.1 and
1.0 ha, and decreases to 0.1 at greater than 20 ha. The suitability of very
small islands (i.e., less than 0.1 ha) is low due to the high probability of a
colony being damaged by waves: colonies established deep within a marsh on
larger islands are protected by the wave-damping effect of the vegetation.
The island isolation variable (V 4) is derived from the island
biogeography theories of MacArthur and Wilson (1967); as the distance from
the mainland increases, the potential for successful predator colonization
decreases. Conversely, increasing insularity exposes the nesting colony to
potential severe wave and tidal damage, which may impose an upper limit on the
distance of high quality island study sites from the mainland. Optimum
nesting islands are assumed to be separated from the mainland by 1 to 3 km of
water sufficiently deep to create an effective predator barrier (greater than
0.5 m deep at mean low tide).
Disturbance. Forster's terns are sensitive to human disturbance of the
nesting colony (V,). It is assumed that commercial or recreational boating
10