Studies of Colonial Water-birds in the Kingston Area by the Canadian Wildlife Service, 1972 – 2007
D.V. Chip Weseloh1 and J. Laird Shutt2
- Canadian Wildlife Service, Environment Canada, 4905 Dufferin St. Toronto, ON
M3H 5T4 (chip.weseloh@ec.gc.ca) - Science and Technology Branch, Environment Canada,
National Wildlife Research Centre, Carleton University, 1125 Colonel By Dr., Ottawa, ON K1A 0H3
Introduction
The Canadian Wildlife Service (CWS), part of Environment Canada, has been conducting studies of colonial water-birds in the Kingston area for more than 35 years, since 1972. In that year, two Environment Canada biologists, Michael Gilbertson and Iola Price, began studies of contaminant levels and their effects in Herring Gulls and Black-crowned Night-Herons, respectively, on islands offshore from Kingston (details below). This work was precipitated by earlier U.S. studies, which had found major contaminant-related problems in fish-eating birds on the Great Lakes (Keith 1966, Ludwig and Tomoff 1966, Anderson and Hickey 1969), and there was concern that a similar situation might exist in Canada.
Four years later, in 1976, after the above study was well underway, another major study of colonial water-birds, which again included the Kingston area, was initiated independently by two ornithologists. Hans Blokpoel, with the CWS, began his work on Lake Ontario, and Bill Scharf, working with the U.S. Fish and Wildlife Service and the U.S. Army Corps of Engineers, tackled the entire U.S. side of the Lakes. The goal of their work was to census all the colonial water-birds, i.e. gulls, terns, herons, cormorants, which nested on their respective sides of the Great Lakes. It is an amazing coincidence and good fortune that such a huge project, which, to be done properly should include extensive planning and cooperation on both sides of the Great Lakes, would start simultaneously by two researchers, without knowledge of the other (details below).
The purpose of this Appendix is to describe and discuss how these two major projects on colonial water-birds, for which Kingston and eastern Lake Ontario were and are such important areas, have continued to this day. We also summarize preliminary results of new studies which have evolved from both of the original studies. We present this material, taken from scientific publications and our unpublished field studies, in a non-technical vein, keeping with the nature of this book.
1. Contaminants in Herring Gull Eggs from Snake Island (Kingston) and the Great Lakes, 1972-2007.
In 1972, after a two year study of the biology and contaminant conditions in Common Terns in Hamilton Harbour (Gilbertson and Reynolds 1972), and knowing of other contaminant conditions on Lakes Superior and Nipigon (Postupalsky 1971), Michael Gilbertson began an extensive study of several colonial water-bird species on islands off Kingston and elsewhere in the Great Lakes. His purpose was to document the extent of chemical contamination in fish-eating birds, in terms of species and geography (Gilbertson 1975). He visited Pigeon, Snake, Salmon, Black Ant and the Brother islands in the Kingston area. He found that the concentrations of pollutants in Herring Gull eggs were among the highest ever recorded for that species and that there was extensive embryonic mortality among gull chicks, resulting in the gull’s overall breeding failure (Gilbertson 1974, 1975; Gilbertson et al. 1976). Also in 1972, Iola Price, who was aware of the contaminant effects that the Kingston Field Naturalists (KFN) had documented earlier on Pigeon Island (Edwards 1963, 1964, 1970), began a study there of contaminant concentrations, eggshell thickness and lowered reproductive success in Black-crowned Night-Herons and found similar results (Price 1977).
After the 1972 field season, it became clear that of the several species of colonial fish-eating birds which nested on the Great Lakes, the Herring Gull was the single best indicator species of contaminant conditions in colonial, fish-eating water-birds on the Great Lakes (Gilman et al. 1977; M. Gilbertson, pers. comm.). Its chief advantage was that it was the only colonial water-bird species which was also a year around resident on the Great Lakes. That meant all of the contaminants in its eggs would have come from the Great Lakes and not from some southern wintering area. In 1973, Gilbertson moved his centre of activity to Scotch Bonnet Island (Gilbertson and Hale 1974a, b), off the southwest shore of Prince Edward County, where again he found elevated contaminant concentrations in gull eggs and extensive reproductive failure. In that same year, Stan Teeple (1977) continued Gilbertson’s work on West Brother Island and, again, found the same situation. Obviously, the problems were wide-spread, involved several species and were repetitive from year to year.
Based on these extensive “preliminary” studies, first in the upper Great Lakes, then in western Lake Ontario, and finally in the Kingston area, a formal Great Lakes Herring Gull Egg Contaminants Monitoring Program was launched in 1974. Its goal was to track contaminant levels and their effects in Herring Gulls throughout the Great Lakes. Additional biologists were hired by CWS, namely Glen Fox in 1974 and Andrew Gilman in 1975. Eight sites, including West Brother Island, were designated as Annual Monitor Colonies: one at Toronto Harbour, in the west end of Lake Ontario, and two in each of the remaining Canadian Great Lakes. In 1974 and 1975, research activity continued in the Kingston area; Herring Gull eggs were collected from West Brother Island for contaminant analysis (Gilman et al. 1977) and behavioural observations were made on incubating Herring Gulls on Snake Island (Fox et al. 1978). Unfortunately, Gilbertson moved to a different job at the end of the ’74 field season. In 1976, realizing there were too few Herring Gulls nesting on West Brother Island for it to remain a viable collection site, the main Herring Gull sampling site in the Kingston area was switched to Snake Island, where it has remained to this day.
In 1978, Weseloh took over as project leader and during 1979 and 1980, seven other sites from Lakes Michigan, Huron and Ontario and the Detroit, Niagara and St. Lawrence rivers were added to the program for a total of 15 (Mineau et al. 1986, Hebert et al. 1999). These sites have been maintained to the present day (2008) as the project celebrates its 35th year of data collection (see also Weseloh 2006).
In the study protocol, fresh Herring Gull eggs are collected from Snake Island, and the other sites, during late April-early May each year. One egg is taken from each of 13 completed clutches. When eggs are taken early in incubation, the female usually lays a replacement egg. Eggs are sent directly to the CWS laboratory at the National Wildlife Research Centre at Carleton University in Ottawa where they are analyzed for more than 75 different routine contaminants.
The data from these eggs permit a rigorous calculation of both geographic and time trends in contaminant levels in Herring Gull eggs (Pekarik and Weseloh 1998, Weseloh et al. 2006). A time analysis of the Snake Island data shows that there has been a dramatic decline in the concentrations of most measured contaminants since 1974. The seven compounds illustrated (Figure 1) have declined from 68% to 99% from 1974 to 2005. In other words, on average, concentrations of the contaminants in Herring Gull eggs from Snake Island have declined 92%, i.e. if concentrations were 100 parts per million (i.e. 100 parts in 106)in 1974, they are 8 parts per million today. Although this is a dramatic decline, the gull eggs are not yet contaminant-free by any means. However, they are much “cleaner”, much less contaminated, than they were 35 years ago.
Figure 1. Percent decline in concentrations of several contaminants in Herring Gull eggs from Snake Island, 1974-2005 (except for 2,3,7,8-TCDD (“dioxin”), which was first measured in 1981 and last measured in 2003).
Time trend data from Snake Island eggs also have been used to examine how food webs are changing in Lake Ontario and how these changes impact contaminant trends in Herring Gulls (Hebert et al. 2006, Hebert and Weseloh 2006). In more recent years at Snake Island, Herring Gulls appear to be consuming more prey from lower trophic levels than previously, which would decrease their exposure to persistent contaminants, such as PCBs. Thus, some of the decrease being seen in their contaminant levels comes from this change in diet rather than strictly from declining environmental levels,
Examples of the year to year fluctuations in contaminant levels, i.e. the time trend, of two widespread contaminants, DDE and PCBs, in gull eggs from Snake Island are shown in Figure 2 (Pekarik and Weseloh 1998, CWS unpublished data). DDE, the common breakdown product of DDT, declined at a constant rate from 1974, after the widespread use of this pesticide was banned, until 2002. There is a much more rapid decline from 2002 to 2004 but this simply may be the result of the low value in 2004 and a final assessment on the current rate should await data from more recent years. The decline in PCBs, on the other hand, has slowed in recent years, starting in 1995. These data indicate that some contaminants (DDE) have not yet begun to level off, i.e. no background level has yet been reached and/or current levels are well above background concentrations. For others, like PCBs, the rate at which the concentrations are declining has slowed down; they are not declining as quickly as they were in the past. This suggests they may be approaching a background level in the environment. Alternatively, the birds may have encountered a PCB source they were not previously using. This could be mediated by re-suspended sediments…or they could be feeding on a food source with higher PCB levels than previously. Annual data from Snake Island for these compounds and others for 1974 to 2002 are available in Weseloh et al. (2003) and Jermyn-Gee et al. (2005) and references therein.
Data from a single site, like Snake Island, when collected for several years, produce the time trends illustrated above (Figures 1 and 2). When data from several sites are collected in the same year(s), geographical patterns or areas of high and low contaminant concentrations can be identified. Data for all 15 sites across the Great Lakes for the years 1998-2002 were averaged for six of the above contaminants. The geographic patterns for mirex and dieldrin are shown in Figure 3. Eggs from Snake Island ranked as the 4th and 11th most contaminated for those two compounds, respectively (Weseloh et al. 2006). These results demonstrate that different contaminants can be more, or less, concentrated in different parts of the Great Lakes. An area that is elevated in mirex may not be elevated in dieldrin, and vice versa. This type of analysis identifies “hot spots” for various compounds and is very useful in pinpointing where remedial actions may be most needed. As mentioned, this project continues to this day and eggs are collected from Snake Island annually.
Figure 2. Plot of concentration against year to illustrate the time trends of two contaminants showing a mostly constant rate of decline (DDE, but see text) and a slower rate of decline in recent years (PCBs).
Figure 3. Geographical patterns in the concentration of two contaminants among the 15 Herring Gull monitoring sites. Eggs from Snake Island (striped bars) were the 4th most contaminated for mirex and the 11th most contaminated for dieldrin. Sites are arranged from Lake Superior to the St. Lawrence River. Individual site names are shown in the dieldrin graph with the corresponding lake names shown in the mirex graph.
2. The Great Lakes Bi-National Colonial Water-bird Survey: What species and how many colonial water-birds breed in the Kingston area and on the Great Lakes?
Decadal Surveys
The Canadian portion of this survey was originally intended to be a one-time occurrence when it was started by Hans Blokpoel in 1976. Blokpoel’s motivation in initiating the Canadian survey was two-fold; he wanted a broad lake-wide understanding of the population dynamics of the Ring-billed Gull and the Common Tern. The gull was becoming very abundant and usurping much of the habitat of the tern, whose numbers were declining dramatically. But also, in light of the contaminants studies mentioned above, which had now been underway for 5 years, he wanted to better document possible effects of contaminants at the population level (H. Blokpoel, pers. comm.). Ironically, on the U.S. side of the Great Lakes, Bill Scharf initiated the same kind of study in the same year, 1976. Scharf’s interest came from the combined inquires of the U.S. Fish and Wildlife Service and the U.S. Army Corp of Engineers, who wanted to know the species and numbers of colonial water-birds nesting on the U.S. side of the Great Lakes, but especially on dredge-spoil islands, which were being created by the Corp (B. Scharf, pers. comm.).
The idea of a Great Lakes-wide survey of colonially-nesting water-birds was not new. Jim Ludwig, and the Ludwig banding team, had surveyed nearly all water-bird colonies of Lake Huron, northern Lake Michigan, and south eastern Lake Superior from 1960 to 1962. They visited 180 colonies of four species: Herring and Ring-billed Gulls, and Common and Caspian Terns, and tallied approximately 72,000 nests (Ludwig 1962). Ludwig’s work set the wheels in motion for the more extensive survey by Scharf, and perhaps for Blokpoel as well. After the results of their surveys were tabulated (it had taken five years to complete the Canadian portion, which had been done on a one lake per year basis), Blokpoel and Scharf realized the enormity and the value of the project they had completed (Blokpoel 1977, Scharf et al. 1978, Weseloh et al. 1986). Nearly 900 active nesting sites had been visited and the number of nests tallied (of the major species) was: 920 cormorants, 75,015 Herring Gulls, 406,962 Ring-billed Gulls, 3,908 Caspian Terns and 10,775 Common Terns for a total of 497,580, or just under a half a million nests (=pairs). Among the Lake Ontario islands in the Kingston area, there were 6 sites, comprising 15 colonies of 6 species with over 30,000 nests (Table 1).
In the years following the completion of the first survey, Blokpoel and Scharf discussed repeating the Great Lakes-wide census. If one complete census allowed researchers to establish population levels of all species and pinpoint their geographical distributions, think what another survey, ten or a dozen years later, would do for establishing the beginning of a time trend. According to Scharf (pers. comm.), it was Blokpoel who then convinced the powers that be of the merits of a second survey, and in 1989 it was begun. A more coordinated approach was taken with this survey; Lakes Superior, Michigan and Huron were surveyed for island-nesting colonial water-birds in 1989, the lower lakes were surveyed in 1990 and, in 1991, the marsh-nesting terns were surveyed. Once the 2nd survey was completed, in 1991, the idea of an ongoing survey every 10 or so years, i.e. a decadal survey, was firmly established with current and future researchers and government officials on both sides of the border. Subsequent surveys have been carried out; the 3rd Decadal Survey was begun in 1997 (Morris et al. 2003, Weseloh et al. 2003, Cuthbert et al. 2006) and the 4th was begun in 2007.
The results of the surveys of the breeding populations of colonial water-birds on Lake Ontario islands in the Kingston area for the three decadal surveys which have been completed (1976, 1990, 1997/98), and for less extensive censuses in 2006 and 2007, are summarized in Table 1. The number of nests (=breeding pairs) is reported for eight species. The Cattle Egret no longer breeds on any of the Lake Ontario islands in the Kingston area and probably will not return any time soon. The Common Tern is a very rare breeder; in four of the five above surveys there were fewer than five pairs in any one year. However, given that Ring-billed Gulls have vacated Pigeon Island, an historical Common Tern nesting area and a single pair of terns has nested there each of the last two years, that species might slowly re-colonize the island. Four species have increased substantially since the first survey: Double-crested Cormorant, Herring Gull, Ring-billed Gull and Caspian Tern. The gull numbers are now approximately double what they were in 1976 and cormorant numbers have increased 77-fold. Caspian Tern numbers have increased 40-fold but, unfortunately, that increase has been at Little Galloo Island at the expense of Pigeon Island. Most Caspian Terns abandoned Pigeon Island in 2001, when nesting was reduced to only six pairs; none has been recorded nesting there since (Weseloh & Shutt, pers. obs.). The Great Black-backed Gull increased substantially from 1976 to 2000, from zero to 24 nests, but declined dramatically from 2000 to 2007, i.e. down to a single nest. This has been due, perhaps solely, to the effects of type E botulism (see below). Lastly, numbers of night-heron nests have declined by more than 90% since 1976. On four of five islands, where night-herons have either abandoned the site or where their numbers have declined, cormorants started nesting on those islands after the night-herons had been nesting there and took over or destroyed the dogwood and/or lilac bushes where the night-herons had been nesting.
These losses of individual species from Pigeon Island point to an overall loss in species diversity there. When the KFN visited Pigeon Island from 1961-1970, there were usually five colonial water-bird species nesting there (Edwards 1963, 1964, 1970). In 1976, there were still five species there but in 2007, there were only 3, including one which was not there in 1976. So in 31 years, three species have abandoned Pigeon Island: Ring-billed Gulls, Caspian Tern and Cattle Egret. Also, during that time, the Great Black-backed Gull has come and gone as a nesting bird there. The reason for this decline in species numbers is not fully known. The number of Caspian Tern nests started to decline in the mid-1990s and may have done so as a result of the increasing number of nesting GBBGs, a known predator. On 17 May 1995, 22 dead Caspian Terns were found in Great Black-backed Gull territories on Pigeon Island (M. Patrikeev, pers. comm.), suggesting that the black-backs were preying upon them. The departure of the terns and Ring-billed Gulls (who abandoned the island about the same time as the terns), may also have been influenced by the shear abundance of cormorants nesting on the island. Cormorants steal nest material from other species with such determination and relentlessness that their victims often experience nest failure. They are known to have displaced Herring Gulls and Black-crowned Night-Herons from a nesting site in Hamilton (Somers et al. 2007).
In spite of the fluctuating number of nesting colonial water-bird species on Pigeon Island over the years, the numerical make up among the major species nesting across the eastern basin has been amazingly constant. In 1976, the numbers of four species, Herring and Ring-billed Gull, cormorant and Caspian Tern, made up 99.3% of the colonial water-bird species nesting on the Ontario islands of the Kingston area. In 2007, that percentage was 99.6%.
In terms of number of nesting sites, i.e. locations (usually islands) where any species of colonial water-bird nests, there were only four sites in 1976; that grew to nine in 2007 (Table 1). In terms of the number of colonies, i.e. the number of specific site-species combinations, e.g. Herring Gulls and cormorants at Pigeon Island counts as two colonies, there were 12 in 1976 and 24 in 2007. Thus, what we have seen on Lake Ontario islands in the Kingston area over the last 31 years has been a tremendous increase in the number of breeding colonial water-birds, in the number of sites where they breed and the overall number of water-bird colonies.
Table 1. Numbers of nests (=pairs) of colonially-nesting water-bird species in the Lake Ontario waters of the Kingston area during the three Decadal Surveys (1976/77, 1990 and 1997-2000) and in 2006 and 2007. This table includes all sites where a given species was recorded during at least one of the census periods.
One of the most important offshoots of the Decadal Surveys was the realization that one survey every 10 years was not frequent enough to capture rapid changes in any of the colonial water-bird populations. More frequent surveys would have better documented the decline of Ring-billed Gulls and Caspian Terns and the rapid increase in cormorant nests between the first and second surveys. Based on the early results of the Blokpoel surveys (Blokpoel 1977, Blokpoel and McKeating 1978) and on the dramatic increase observed in cormorant nests on a trip to Little Galloo Island in 1978, Weseloh decided to institute annual counts of nests at all cormorant colonies in Lake Ontario, as well as in the rest of the Canadian Great Lakes (see below). At that time, this amounted to approximately 14 colonies. Thus, this annual survey of cormorant nests, which in 2007 included over 40 colonies in the lower Lakes and which is now accomplished with the help of many cooperators, also had its beginning as CWS work in the Kingston area.
Annual Surveys of Double-crested Cormorants
The Double-crested Cormorant was first found nesting in Lake Ontario on Scotch Bonnet Island where Farley Mowat noted “…about 40 old nests and some addled eggs.” in August 1939 (Baillie 1947). It was first found nesting in the eastern basin of Lake Ontario on Gull Island, New York, in 1945, with 14 nests. The second colony in the eastern basin was on Snake Island in 1949, also with 14 nests. For the next seven years, until 1956, the known number of cormorant nests in Lake Ontario’s eastern basin fluctuated between 25+ and 64. However, this period of growth was not to persist. From 1957 to 1973, a span of 17 years, the number of nests in the eastern basin declined from 22 to 0 (Price and Weseloh 1986). This decline was reflective of conditions during that time, the “pesticide era”, when eggshell thinning, embryonic mortality and failed reproductive success were common features of many colonies of colonial water-birds on the Great Lakes. In the mid-1970s, however, conditions and the situation improved markedly. DDT/DDE had been legislated out of use. A new nesting colony was found at Little Galloo Island in 1974 and by the late 1970s nesting had resumed, in earnest, on Pigeon Island. From 1975 to 1979, the number of cormorant nests on Pigeon and Little Galloo islands (the only Kingston area colonies active then) grew from 28 to 76 to 96 to 196 to 315 (Price and Weseloh 1986). This is an average annual growth rate of 83.1% and much too great to be accounted for by local annual productivity alone; immigration to those islands must have been occurring.
In 1979, with assistance from the New York State Department of Environmental Conservation (NYSDEC), CWS began annual censuses of all cormorant colonies in eastern Lake Ontario. Clearly, the worst was over; contaminant concentrations in eggs of colonial water-birds continued to decline (Pekarik and Weseloh 1998, Ryckman et al. 1998). The 1980s would see the cormorant population in the eastern basin grow from 373 nests on those two colonies to nearly 5,000 nests on five colonies, with the addition of colonies on Snake, Salmon and False Duck islands. The average annual growth rate for cormorants in the eastern basin during the 1980s was 29.6%. Cormorants resumed nesting on Snake Island in 1983 with four pairs and by 1989 there were 164 nests present. By the end of the 1990s, the eastern basin population stood at just over 8,200 nesting pairs, still on five colonies (nesting on West Brother Island replaced the short-lived colony on Salmon Island). The average annual growth rate for that decade was 4.95%. The population in the eastern basin would peak the next year (2000) at 10,407 pairs.
Since 2000, in the eastern basin of Lake Ontario, the cormorant population has declined from 10,407 nests on five colonies to 7,402 on nine colonies in 2007. New colonies have started, or resumed, on the U.S. side of the Lake at Gull and Bass islands (which receive varying levels of control, under permit, each year, e.g. egg-oiling, knocking down nests, shooting adults,), and on the Canadian side at East and Middle Brother islands, and False Duck Shoal (Gull Bar). Cormorants have abandoned False Duck (Swetman) Island; all other previous sites mentioned above are still active. During the 2000s, there have been short-lived colonies on Yorkshire Island (2002) and Grape and Nut islands (2005). As far as is known, these sites were only active for the year indicated. The reason for this decline in nest numbers and the increase in diversification of nesting sites might be the control strategies that have been going on at major colony sites in the basin.
With the return and the increase in cormorants in eastern Lake Ontario, there has been concern by the public and by island owners about impacts from these birds, e.g. the killing of trees and fouling cottages. Land owners are permitted under provincial law to protect their property from damage by wildlife (cormorants) and there have been sanctioned (permitted) culls of cormorants on the Brother Islands (2004) and West Spectacle Island (2006). There has also been concern about potential impacts from cormorants to other colonial species. This has been a major rationale for the management of cormorants on Little Galloo Island (Farquhar et al. 2003). Lastly, there has been great concern over possible cormorant impacts to fisheries. A detrimental relationship between cormorants and fisheries, however, has been more difficult to establish than have other concerns. Several apparently illegal culls or unsanctioned management actions against cormorants have taken place, often at colonies where private property or impacts to other bird species or vegetation is not a concern, and where conflict with fisheries appear to be the main issue. These are known to have occurred and been documented as follows: extensive shooting at Little Galloo (Weseloh et al. 2003) and Pigeon (Ewins and Weseloh 1994) islands, egg destruction on Little Galloo Island (Weseloh and Struger 1985), and the release of two raccoons on Pigeon Island where all eggs of cormorants and Herring and Great Black-backed Gulls were eaten or destroyed (C. Pekarik and DVW, unpubl. data). Interestingly, on 24 June 1997, a fisherman in a boat with two raccoons in cages was encountered heading out from Prince Edward Point to the False Duck Islands, to release the raccoons to harass the cormorants nesting there. When told that was not a good idea, he changed course to nearby Timber Island, where no colonial water-birds nested (L. Shutt, C. Hebert, K. Williams, pers. obs.).
In the St. Lawrence River portion of the Kingston area, an area much less studied by CWS, cormorants have also resumed nesting in recent years. Historically, cormorants were first found nesting in the upper St. Lawrence River at Black Ant Island in 1945 (Baillie 1947). Since the end of the pesticide era, they were first noted nesting in the upper portion of the River at Strachan Island, upstream from Cornwall, after a hiatus of at least 15 years (Weseloh et al. 2007). They did not start or resume nesting in the Kingston area of the River until 2001, when 18 nests were found on Blanket Island in U.S. waters just south of Black Ant Island, 23 km east of Kingston. They started nesting on Black Ant Island in 2003 and have continued on those two islands through 2007, when they totalled nearly 300 nests. Cormorants nested for one year on West Spectacle Island (2005) but were harassed there in the autumn of 2005 and culled, under permit, in the spring of 2006 and have not returned (D. Radke, pers. comm.; DVCW, pers. obs.).
In summary, in 2007 in the lake Ontario waters of the Kingston area, there were 7,285 cormorant nests on 7 islands and another 295 nests on 2 islands in the upper St. Lawrence River. This growth from 315 nests in 1979 to 7,575 nests in 2007 represents an average annual growth rate of approximately 12.0%
3. Nocturnal Roosting on Kingston Area Islands by Caspian and Common Terns
Roosting
Roosting is a relatively little studied aspect of the behaviour of all birds. By definition, roosting is sleeping; a roost is the location where birds sleep and going “to roost” is the process of flying to where they will sleep. However, sleeping is only one of the activities that birds do at their roost (Brooke and Birkhead 1991). During the post-breeding period, they spend a great deal of time preening, some young birds may still be begging (and getting) food from adults, they may squabble and interact with other individuals in a variety of ways. Most colonial water-birds roost communally, i.e. in large groups, just like the way they nest. During the breeding season, the roost and the breeding site are often one and the same, although some gulls and terns are known to desert their clutches during egg laying and early incubation and roost away from their breeding colony early in the season. During the spring, summer and autumn, in our area, roosting can occupy from 25% to 50% of the daily activity of a colonial water-bird, so it occupies a considerable portion of their time.
The Importance of Islands for Colonial Water-birds
Islands are extremely important habitat for colonial water-birds. Their most important function is probably to provide an area free of terrestrial predators. These islands then become suitable for ground- and tree-nesting birds to use for breeding habitat (Blokpoel and Scharf 1999). Hundreds of gulls, terns and cormorants have used many of the islands in eastern Lake Ontario for breeding over the years. This kind of use is easy to document and the species accounts in the main body of this book abound with nest counts from various islands and years.
A less obvious function of islands is to serve as a safe resting and stopover location during spring and autumn migration, outside of the breeding season. This kind of use is easy to document for migrant species that do not nest in the Kingston area. For example, if one sees a flock of Brant on Snake Island, they are obviously using it as a stopover site. However, it is much more difficult to determine the status of an island for species which are both summer residents and migrants through the same area, e.g. how does one identify or separate locally breeding cormorants and terns on islands from migrant individuals of the same species who are using the island as a stopover location? Failure to identify or recognize this cohort of birds underestimates the value of that island/site.
It is natural to assume that colonial water-bird species which breed on a given island in the spring and early summer will also roost there in late summer and autumn. For example, Herring Gulls and cormorants breed on Snake and Pigeon islands and on False Duck Shoal; they also roost on these islands during the summer and autumn. However, Caspian Terns do not breed on any of the three above islands, yet they, too, roost on them during the summer and autumn. Likewise, Common Terns do not breed on Snake or Salmon islands but they, too, roost on both of them. So, some species, or some individuals of some species, will vacate their breeding sites in mid-summer and go roost, and perhaps take up temporary residence elsewhere. What causes some birds – or some species – to remain at a breeding site after breeding is completed and continue to roost there while others abandon an apparently safe site – safe enough to breed – in favour of roosting at another location? Unfortunately, we do not know the answer to that question.
Identifying Local Resident and Migrant Colonial Water-birds at Roosts
Observations of colour-banded cormorants on Snake Island have shown that upwards of 50% of the birds observed there in the early spring (April 2006, DVCW unpublished data) did not breed there, nor did they frequent the island during the breeding season. This suggests that they were migrants or at least non-resident birds, who were using Snake Island as a stopover site. Using the terns from the above paragraph as a further example, neither species has bred on Snake or Salmon islands in recent memory. However, for at least the last 7 to 10 years, they have been observed to roost on both of those islands from mid-summer onwards, sometimes numbering in the hundreds. For example, on 17 September 2001, from a blind, Weseloh observed approximately 50 Caspian Terns and 450 Common Terns come in to roost on Snake Island. It is not known where any of the Caspian Terns had bred or were raised; none was colour-banded. They could have come from any of three or four Lake Ontario colonies or farther north. However, five of the Common Terns were banded, each with different coloured leg band. Three of them had been banded by Dr. Lee Harper, a tern researcher from Massena, NY, one each in 1998, 1999 and 2000 all in Lake St. Lawrence (roughly between Morrisburg and Ingleside, Ontario), approximately 140 km down the St. Lawrence River. At least one of the others had been banded by Mark LaBarr, a tern researcher from Huntington, VT, on Lake Champlain. Obviously, some of these birds were “regional” breeders, as there are few Common Terns breeding in the Kingston area. The others may or may not have been, but all were using Snake Island as a stopover site. They had all come to the Kingston area from a reasonable distance away. These islands, which provide safe overnight roosting sites, are probably very important in maintaining local populations of those birds in the post-breeding season. If they were not there, and no other islands or sites were available for roosting, the birds would probably move right on through the area and not be present late into the season.
It would be useful to evaluate all the Lake Ontario islands in the Kingston area for their use or potential as roosting sites for (colonial) water-birds. This would help identify key sites which these birds use other than their breeding colonies. To date, this has been done for only a few islands; Snake and Salmon islands are noted roosting sites which attract colonial water-birds in addition to those species which nest there. They are known to serve as stopover sites and/or roosting sites for migrant cormorants, Common and Caspian Terns, and various waterfowl and shorebird species, e.g. nine Whimbrel came in to roost on Salmon Island at 30 minutes after sunset on 17 September 1998 (DVCW, pers. obs.). Salmon Island, but not Snake Island, is a roosting site for Great Black-backed Gulls, a species which does not breed there. Pigeon Island, on the other hand, is not noted to attract any additional colonial water-bird species during the post-breeding season. Neither Caspian nor Common Terns are known to roost there. Why there is such a difference among which species are attracted to these three islands remains unclear. Offshore distance may be a factor in the case of Pigeon Island.
A Night on Snake Island
Confirming nocturnal roosting, especially with Common and Caspian Terns is not easy and most birders probably seldom have such an opportunity. In blinds we have erected on Snake and Salmon islands, we often stay overnight on the islands to observe this roosting behaviour. Below, we provide a short description of a typical late summer evening on Snake Island as cormorants, gulls and the two tern species come in to roost (DVCW, unpublished).
I arrive at Snake Island at 3:45 PM, moor my 4.3 m (14 ft) aluminum boat on shore and enter the blind at 4:00 PM, 3.5 hours before official sunset. Cormorants, Herring Gulls and a few Ring-billed Gulls flushed from the island on my arrival. The gulls start returning to the island within minutes of my entering the blind; the cormorants take 20 or 30 minutes to start returning. The numbers of all three species increase as the afternoon wanes and sunset approaches and as birds return, presumably from foraging. Cormorants return mostly in small groups of 1-4 but also in larger flocks of 20-30+ and their numbers build up quickly, once the first bird ventures onto the island. Gulls, for the most part, return singly and the increase in their number is more subtle. By 4:45 PM there are about 200 cormorants perched on rocks at the east sometimes the south ends of the island and about 40 Herring Gulls walking around the interior of the island. Both species usually fly directly to the island where they settle either right on the island or in the nearby water and swim up to it.
The roost arrival behaviour of Common and Caspian Terns is much different from that of gulls or cormorants. Little is usually seen or heard of the presence of either species until around official sunset, often not until after sunset; only rarely will an occasional individual be heard calling or seen before this. After sunset, the distinctive call of Caspian Terns and their circling of the island become more frequent. Individuals begin to settle directly on the island, usually on a slightly elevated portion or ridge. Again, once the first individuals settle on the island, others follow quickly. Their raucous calling dominates the sounds during the post sunset-twilight period and the calling often continues all through the night.
Compared to the other colonial water-birds present, Common Terns are much more delayed but also tentative, flighty and synchronous in their arrival at their roost, at least during the post-breeding season. It will almost be dark when they usually arrive and they arrive en masse. A flock of 40, 50 even up to 100 birds will seem to come out of nowhere; all of a sudden they are flying around the island. They circle about wildly first out over the water, then over the island, then out over the water again. They may be joined by another flock and the whole mass circles about again. Then, on a low swoop over a low-lying portion of the island, where water has inundated small boulders, rocks and/or gravel, they will alight in a flutter of wings. They are somewhat vocal on arriving and settling and may also continue calling during much of the night. Ducks, including Mallards, Common Mergansers, American Wigeon and Gadwall, and geese and shorebirds also start arriving just prior to dusk; many others often arrive overnight.
In the morning, Common Terns usually begin to leave the island, i.e. their roost, very early, sometimes even before the first hint of light which is nearly one hour before official sunrise. Departure by Caspian Terns is usually not far behind. Both species have certainly departed before sunrise. It is difficult to tell when gulls leave the island as they are usually scattered throughout the island and do not usually congregate in a flock on the island (roost) as the other species do.
In terms of morning roost dispersal, cormorants are the last to depart from the Snake Island roost site. Their morning departure is the most delayed. As first light progresses and sunrise approaches, most cormorants are still in their overnight standing positions. Many are preening; some are still sleeping with their heads under their wings. Soon, a few individuals will take to the water and a few others will fly from the island directly. Over the next 30 minutes, or so, more and more individuals will follow this pattern until there is a steady stream of cormorants departing from the island. It is only well after sunrise, sometimes as much as an hour, when virtually all the birds have finally departed.
4. Roosting on Navigational Buoys in Kingston Harbour by Black and Common Terns and Purple Martins
It may seem a bit unusual to include the Black Tern in a survey of colonial water-birds which use the eastern basin of Lake Ontario, but it is included for good reason. Black Terns were accidentally discovered roosting on lighted navigational buoys in the Kingston Harbour one evening by Weseloh. Preliminary investigations showed, in fact, that both, Black and Common terns, as well as Purple Martins, roosted on several of the lighted navigational buoys in Kingston Harbour. Hence, a nocturnal survey route consisting of up to nine such buoys in the Harbour was established. The route consisted of the following buoys: KH2 – the red buoy west of Fort Henry; K1 – the green buoy off the Kingston Yacht Club; K3- the green buoy off the Kingston Penitentiary; K5 – the green buoy north of Snake Island; K8 – the red buoy south of Salmon Island; KN2 – the green buoy off the north end of Salmon Island; and KA1 – the green buoy off the east end of Garden Island (Figure 4). Occasionally buoys H43, east of Cedar Island, and H42, south of the Spectacle Islands were also included. The number of terns was counted and/or estimated as each buoy was approached slowly by boat and illuminated momentarily with a strong searchlight. The birds usually flushed when the buoy was illuminated but returned as soon as the light was extinguished.
In 2000, Weseloh conducted a survey of the above buoys for roosting birds on 21 evenings between 20 May and 23 September. Surveys always were initiated after total darkness and took about an hour and 20 minutes to complete. Due to weather conditions or limited time, not all buoys were visited on all surveys. Black Terns were found roosting on the buoys on all surveys during May (N=3), June (N=1) and July (N=4) and on the first nine of 11 surveys in August; they were present up to and including 27 August but were not present on 28 and 29 August or in September (see below). The average number of terns per buoy per month was 9.4, 7.0, 11.0 and 5.0, respectively; the maximum number on any night was 40, which occurred during both the 1st and 3rd weeks of July. The average number of terns per survey was 42 and the maximum number was 103 on 22 July. The total number of Black Terns recorded on the 21 surveys was 706 and the last evening on which they were observed on the buoys was 27 August.
Common Terns and, much to our surprise, Purple Martins also roosted on the lighted buoys with Black Terns. Common Terns were less numerous than Black Terns and were present from 24 June to 23 September. Peak numbers occurred on 21 and 23 August with an average of 11 birds per buoy. The average number of Common Terns per survey was 31 and the maximum number was 130. The total number of Common Terns recorded on 21 surveys was 653. Purple Martins were more numerous and were present from 21 May to 3 September. Peak numbers occurred on 27 July when 38 birds per buoy were recorded. The average number of martins per survey was 51 and the maximum number was 226. The total number of Purple Martins recorded on the 21 surveys was 1075.
The other interesting item about these three species roosting on a single buoy was their distribution on the buoy. Purple Martins were always in the upper superstructure near the top of the buoy. They never perched on the bulbous lower portion of the buoy. Black Terns were almost always on the bulbous lower part and rarely (but occasionally) on the superstructure. Common Terns roosted on both areas roughly equally. The first question one might ask is why do they roost on the buoys? Obviously, they are free from mammalian land predators when roosting on buoys; no coyotes, raccoons or skunks can get to them. But why not roost on the islands like other species do, and, in fact, like other Common Terns do? We simply do not know. From our observations, space is not a limiting factor on either Snake or Salmon islands. Purple Martins usually roost in trees or under concrete bridges (Brown 1997). We have been unable to find any references to where Black Terns normally roost at this time of the year.
Figure 4. Locations of the lighted navigational buoys (darker grey circles) which were surveyed for roosting Black Terns, Common Terns and Purple Martins.
5. Colour-banding studies of Double-crested Cormorants in eastern Lake Ontario
Since 2000, CWS and Wildlife Services of the U.S. Department of Agriculture (Tommy King and Jennifer Chastant, Starkville, MS) have colour-banded over 3,000 cormorant chicks on Pigeon, Snake and West Brother islands and False Duck Shoal. They have also maintained an elevated bird blind on Pigeon and Snake islands, as well as occasionally on Salmon Island and False Duck Shoal, to watch for and identify colour-banded cormorants. This work is part of a larger project, extending from Lake-of-the-Woods to New York Harbour that has colour-banded over 24,000 cormorants. The main goal of this work has been to document inter-colony movements of cormorants and to assess various aspects of their population dynamics, e.g. age at first breeding and age-specific mortality. Below we provide a summary and comparison of the origin of colour-banded cormorants observed on those three islands during one year.
In 2006, extensive observations were made on Snake and Pigeon islands and False Duck Shoal (Gull Bar) to identify colour-banded cormorants which bred or occurred on those islands. In total, 253 different individually colour-banded cormorants (banded in 2005 or earlier) were identified: 131 on Snake Island, 74 on Pigeon Island and 48 on False Duck Shoal. At least 97 (74.1%) of the colour-banded cormorants observed on Snake Island had been banded there as young, flightless chicks suggesting a high degree of natal site fidelity. On Pigeon Island, 13 (17.6%) of the colour-banded birds observed there had been banded there. False Duck Shoal was a new colony and no birds had been banded there yet. The locations of other sightings are summarized in Table 2.
Table 2. Percent of sightings of individually colour-banded cormorants made at three sites in 2006. It is read as follows: 16.8% of the colour-banded cormorants observed at Snake Island were banded at Other Eastern Basin colonies.
The pattern that emerges from these data is that cormorants from Snake Island, which is located close to the mainland, in the least exposed area and among several other islands, showed a much greater level of fidelity to their natal site than did the birds from Pigeon Island. Snake Island also had the lowest per cent of cormorants and the fewest number that had been banded outside the eastern basin.
At Pigeon Island, which is located in much more open water and considerably farther offshore in the open lake, there was a much smaller per cent of birds which showed fidelity to the island and a greater proportion of cormorants which had been banded outside eastern Lake Ontario than was seen at Snake Island. It also had a small number of cormorants that had been banded from the more westerly Great Lakes.
False Duck Shoal is located in an exposed portion of Lake Ontario and is the most westerly located of the three islands where we made observation. At that site, we observed a roughly comparable per cent of colour-banded cormorants (20.8-29.2%) from 1) other colonies in the eastern basin of Lake Ontario, 2) other Lake Ontario colonies, 3) the Oneida Lake/Lake Champlain/Lakes St. Francis and St. Pierre areas of the St. Lawrence River and 4) other Great Lakes sites and James Bay (Table 2). Again, with such an exposed offshore site, cormorants which may have been “looking” for a site at which to nest, would probably encounter False Duck Shoal or Pigeon Island before they would encounter Snake Island. Also, as a new colony, it may have been easier for birds from elsewhere to secure a territory there than it would have been on Snake or Pigeon islands.
In trying to explain Pigeon Island’s greater rate of occurrence of cormorants from outside eastern Lake Ontario, the density of nesting cormorants and the amount of available space for additional nesting could be factors. In 2006, Pigeon Island had 1,606 cormorant nests for a density across the island of 0.13 cormorant nests per sq m (the area of Pigeon Island is 1.2 ha). Snake Island had 617 nests for an overall density of 0.21 cormorant nests per sq m (the area of Snake Island is 0.3 ha). So, cormorants were nesting less densely on Pigeon Island; a new or foreign bird might be able to establish a territory within a sub-colony more easily on Pigeon Island than on Snake Island. On both islands, the area actually occupied by cormorant nests, i.e. the various sub-colonies, was relatively small. From area measurements taken on a known number of nests in 2001 (J. Duffe, unpublished data), we estimated that the area occupied by cormorant nests in 2006 was 36% of Pigeon Island and 28% of Snake Island, i.e. 64% and 72% of the two islands, respectively, were still available for nesting. So, there was considerable space available on both islands. We do not have an area measurement for False Duck Island.
An incident which happened on Pigeon Island in 2002 has to be considered as a possible contributing factor to the low per cent of natal colony fidelity shown by colour banded cormorants at this site in 2006. Between 21 and 30 May 2002, person or persons unknown deposited two adult raccoons on Pigeon Island. When found on 30 May, the raccoons had broken, killed or consumed virtually all eggs and chicks in the approximately 1100 nests of cormorants and Herring and black-backed gulls that were on the island. There is no way of assessing the impact of this incident on the fidelity (in 2006) of cormorants born on Pigeon Island. However, during 2003 and 2004, the number of cormorant nests on Pigeon was less than 50% of what it was in 2001 or 2002. Clearly, many of the birds which bred on Pigeon Island in 2002 did not return during, at least, the two subsequent years. In 2005 and 2006, the number of nests slightly exceeded the number present in 2002, but an unknown number of Pigeon Island-born birds may have already permanently left the island.
6. Great Black-backed Gulls on Pigeon Island and the movements of two satellite tagged breeding adults
Eastern Lake Ontario holds a unique position for the Great Black-backed Gull (GBBG) on the Great Lakes; until recently it was, without question, the centre of their breeding distribution on the Lakes. The first breeding record for the GBBG on the Great Lakes was at Little Haystack Island, on the west side of the Bruce Peninsula, in Lake Huron in 1954 (Krug 1956). The second record was on Lake Ontario at Gull Island at Presqu’ile Provincial Park and the third location record was for Pigeon Island in 1971. From 1954 until 1981, GBBGs only nested sporadically on the Great Lakes and then only as single pairs on scattered islands, mostly on Lake Ontario (Angehrn et al. 1979, CWS unpublished data). In 1981, the residency of GBBGs on Lake Ontario was given a significant boost when four pairs were found nesting on Little Galloo Island (Weseloh 1984); in that year there was also a single pair on Pigeon Island. Multiple nesting pairs continued on Little Galloo for the next 20+ years and the number of nesting pairs on Pigeon Island also increased. In 1987, Pigeon Island had its first multiple pairs when two nests were found. By 1990, there were five pairs nesting there and this increased to 10 in 1995 and 16 pairs in 2000 (CWS published data).
This increase in nests of Great Black-Backed Gulls in eastern Lake Ontario had piqued our interest to know more about their annual movements. It was well known that GBBGs wintered on the GLs and Angehrn et al. (1979) had shown that banded GBBGs recovered in Ontario, mostly on Lake Ontario, had come from the lower St. Lawrence River as well as the mid-Atlantic coast. Our question was, where did locally breeding GBBGs from Lake Ontario go in the winter and when did they return to Pigeon Island? With that question in mind, Richard Joos, Jamie Reid and Weseloh trapped two breeding adult GBBGs, from different nests, on Pigeon Island on 21 May 2002 and fitted each bird with a satellite transmitter. By analyzing the data acquired from those transmitters, we were able to determine the birds’ general pattern of annual movement. The two GBBGs remained near their breeding colony from April until August. During this time, they consistently flew (40-45) km northwest to the Bay of Quite where they spent time at a dumpsite near Deseronto, presumably feeding, and time on the nearby water, presumably loafing. Interestingly, they also made some of these trips at night, which adds new insight into the ecology of these birds. After breeding, both birds travelled south to the area of Rochester, NY, where they stayed until late December. They then travelled to Niagara Falls for a week or two before continuing southwest to Lake Erie, near Cleveland, OH, where they spent the remainder of the winter. Both birds occupied similar overlapping ranges during that winter. The birds began heading back to their Pigeon Island breeding colony in early March and arrived there in early April.
7. Botulism in colonial water-birds at Snake, Salmon and Pigeon Islands, 2004-2006
Avian botulism is a paralytic disease which occurs when birds consume a neurotoxin produced by Clostridium botulinum, a gram-positive bacterium commonly found in soils and lake and wetland sediments. Fish-eating birds may be exposed to the toxin directly or through the consumption of contaminated fish, invertebrates or by scavenging dead, botulinum-exposed fish or birds. There are seven distinct botulism neurotoxins (Brand et al., 1983), two of which (types C and E) have caused significant mortality in birds. Type E botulism was first reported in birds in 1963 and 1964 by Herman (1964) and Kaufman and Fay (1964), respectively, when Common Loons and several gull species were recovered from the shores of south-eastern Lake Michigan. First reports of a major die-off involving colonial water-birds associated with type E botulism in the eastern basin of Lake Ontario did not occur until August of 2004
Systematic surveys of dead colonial birds on the islands in eastern Lake Ontario (including the Kingston area) were initiated in the fall of 2004 as a result of reports of large numbers of dead cormorants and gulls washing up on beaches and shorelines in Prince Edward County. Six islands in the eastern basin of Lake Ontario were visited every ten days from 1 September to 30 October in 2004 and every ten days from 1 July to 30 October in 2005 and 2006.Three of the survey islands (Scotch Bonnet, False Duck and False Duck Shoal) are located off Prince Edward County while the remaining three islands (Pigeon, Snake and Salmon) are much closer to Kingston. During each ten day period all islands were surveyed on foot. All fully-flighted, dead birds were identified to species in 2004. However, dead Herring Gulls, Ring-billed Gulls, Great Black-backed Gulls, Common Terns, Caspian Terns and Double-crested Cormorants were also identified as either adult or young of the year in 2005 and 2006.
Over the three years of the study, a total of 4,750 dead colonial water-birds was counted on the six survey islands (Table 3), 1,267 of these were located on the three islands close to Kingston, see Table 4. On these islands, the majority (63%) of dead birds encountered were Double-crested Cormorants, followed in descending order by Herring Gulls, Ring-billed Gulls, Great Black-backed Gulls and Caspian Terns. The proportion of dead birds found on each of the three islands in the study area changed over the three years of the study. For example, in 2004 only 7% of dead birds found on the close islands were encountered on Pigeon Island. However, by 2006 that number had risen to 74% while the proportion of birds found on Salmon Island declined from 79% to 7% over the same period. These changes may, in part, reflect changes in the abundance of species known to favour particular sites for roosting and loafing. Salmon Island was a favoured roosting location for the small but expanding GBBG population in eastern Lake Ontario. In 2004, 46 of the 94 dead birds observed on the island were GBBGs. By 2006, only 9 GBBGs were found on Salmon Island, coinciding with a 90% decline in the number of breeding pairs in eastern Lake Ontario during the same period. The presence of type E botulism in the Lake Ontario food web has probably contributed significantly to the decline of this species in the Kingston area.
This study is an important complement to that carried out by the NYSDEC which is based on beached bird surveys. Surveying land-based roosting sites for mortality samples a much different group of birds than that sampled on beached bird surveys, which would include water-based roosting birds, e.g. Common Loon and several waterfowl species.
Table 3. Number of dead colonial water-birds found on six Lake Ontario islands during mortality surveys during 2004-2006.
Table 4. Number of dead colonial water-birds found on Pigeon, Snake and Salmon islands during 2004-2006.DCCO = Double-crested Cormorant, HERG = Herring Gull, RBGU = Ring-billed Gull, GBBG = Great Black-backed Gull, CATE = Caspian Tern.
Acknowledgements – We wish to thank the many colleagues, technicians, contractors and volunteers who have helped us over the years (our apologies to those whom we have forgotten here): Glenn Barrett, Larry Benner, Tania Havelka, Dave Moore, Cynthia Pekarik, Stan Teeple, Earl Walker, Kim Williams, Neil Burgess, Rob Dobos, Kate Gee, Margie Koster, Clive Hodder, Paul Mikoda, Jamie Reid, Dave Ryckman, Christine Bishop, Jen Chastant, Lee Harper, Craig Hebert, Pierre Mineau, Richard Joos, Tommy King, Don Tyerman, Ron Weir, Peter Rubens, Jim Farquhar, Irene Mazzocchi, John Haig, Doug McRae and Scott Rush. We are appreciative to Glen Fox, Michael Gilbertson, Andrew Gilman and Iola Price for historical information on the early years of CWS activities in the Kingston area. We are very grateful to Dr. Jim Day for ongoing access, all these years, to Snake and Salmon islands and to John and Meg d’Esterre for their hospitality and annual privilege to Weseloh (since 1979) of renting a cottage on Garden Island, which is only a 10 minute boat ride from Snake Island!! Dave Moore, Craig Hebert and Don McNicol reviewed various portions of the manuscript. Dave Moore also prepared the Tables and Figures.
