Lake Huron has the largest collection of freshwater islands in the world. These islands are a significant contributor to the biodiversity of the region. In this paper, we give preliminary results of a project that assembles mapping of over 23,000 islands and island groups and provides the most comprehensive biodiversity assessment of islands in Lake Huron to date. The number, extent and configuration of many islands, particularly small, low-lying systems, is very dynamic depending on lake-levels. Islands in Lake Huron can be divided into three general groups: 1) limestone and dolostone islands associated with and surrounding Manitoulin and Drummond Islands and the Bruce Peninsula, 2) dense archipelagos of small nearshore Precambrian Shield islands in eastern Georgian Bay and the North Channel and, 3) small groups of low-erodible islands in Saginaw Bay.
All three of these island groups are important for supporting colonial nesting waterbirds, endemic species and communities, and migratory birds. Lake Huron islands have been somewhat buffered from anthropogenic change due to their isolation and therefore support a rich and diverse sets of species and communities. Primary threats to island communities include development and invasive species. Threats are generally greater in many of the southern island regions where fewer islands are protected. Results from this project can be used to set priorities for conservation of key sites with high biodiversity values and conservation urgency.
Of the five Great Lakes, Lake Huron embodies the biodiversity of the basin more than any other. Stretching over 350 km from Ontario's “Carolinian” zone at the mouth of the St. Clair River north to the boreal transition forests along the North Channel, Lake Huron includes a multitude of climatic, geological and biogeographical zones. This second largest of the Great Lakes, Lake Huron contains more shoreline than any other of the Great Lakes. Its deep bays, meandering shoreline and thousands of islands represent the longest freshwater coast in the world. These coasts are critical for biological diversity, and many of the Great Lakes basin's endemic, disjunct and globally rare species occur near the coast. In addition to species, the coasts of Lake Huron harbour many vegetation community types that, while poorly documented, are known to be very rare and range restricted. Eastern Georgian Bay for example contains a minimum of 19 globally rare vegetation communities (Nature Conservancy of Canada, 2006).
Islands are an integral part of the biophysical character of Lake Huron. From the intricate archipelagos of eastern Georgian Bay, to the low-lying, erodible islands of Saginaw Bay, these islands harbour many of the unique biodiversity features characteristic of Lake Huron. Due to their isolation, strong coastal influence and unique geology, these islands support ecological communities that are found nowhere else in the world. In addition, these islands support ecosystem functions and phenomena important for maintaining the biodiversity of the Great Lakes region. Collectively, Lake Huron islands are well known to be important sites for nesting colonial waterbirds (Wires and Cuthbert, 2001), for species and communities endemic to or largely limited to the Great Lakes (Henson and Brodribb, 2005; TNC, 1999), and for disjunct species and communities, especially those from western North America (Guire and Voss, 1963) and the Atlantic coastal plain (Jalava et al., 2005). Other specialized habitats that contribute to biodiversity include key stopover sites for migratory landbirds and for high quality shoals for fish spawning (Ewert et al., 2004).
The islands of Lake Huron are relatively young. Many were part of, or connected to, the mainland following the last period of glaciation when water levels were lower. However, during the Lake Nipissing stage (approximately 4000 years ago), isostatic rebound and changes in outflows caused water levels to rise about 8 m above present-day levels before receding again (Karrow and Calkin 1985). Islands with lower relief were submerged during this period, and then emerged as water levels receded to present day levels. In most regions of Lake Huron, islands are still slowly emerging as the land continues to rebound from glaciations, and, in recent years, due to lower water levels. Colonization of most islands by mainland flora and fauna has therefore occurred in the last few thousand years. Many offshore islands harbour plant and animal communities that are different from the mainland due to their isolation and unique disturbance regimes. For example, islands around Manitoulin have very unstable plant communities due to changing lake levels, limited immigration and colonial nesting waterbirds (Morton and Venn, 2000); and many islands do not have natural populations of White-tailed Deer (Odocoileus virginianus) which results in the development of vegetation community structures, such as a high density of Canada yew (Taxus canadensis), that are very rare on the mainland (Ewert et al., 2004). Although islands are rarely free of introduced species, some islands have relatively low numbers of introduced species and thus provide excellent examples of communities that are characteristic of the Great Lakes region.
While the importance of islands to biodiversity in the Great Lake basin has long been recognized, the entire collection of islands has not fully examined. Here we present preliminary results of a project to assess island biodiversity in the Great Lake region, focusing on Lake Huron islands in both Ontario and the United States. While our analysis has been completed for Ontario, due to limitations in spatial data, this analysis has not yet been finalized for the United States portion.
The initial step in this project was the creation of the first bi-national spatial database and map of Lake Huron islands from national, provincial and state digital mapping. This database includes large islands, small islands, rocks and reefs. In Ontario, detailed spatial information was readily available. For United States islands, some of the spatial data for existing mapped islands were very coarse, and it was thus necessary to re-digitize these islands to better capture their true size and shoreline configuration. Digital navigation information was added to the island mapping database to include rocks and reefs. Names were then assigned to all of the islands based on existing maps, geographical naming databases and navigational charts.
Once the base mapping for islands was completed, different analytical methods were used for the Ontario and United States islands. A more comprehensive and accurate island data layer in Ontario, and complete datasets of island attributes allowed for an automated analysis with Geographical Information Systems (GIS) relatively early on in the project. These results were then supplemented with detailed information from the literature. In the US, the need to update spatial information on islands and assemble data from a variety of state sources delayed a detailed automated analysis. For islands in the US, basic island information extracted from the data (e.g. number, area) was supplemented with a literature review. Fortunately most of the US islands in Lake Huron have detailed studies available that could be used to assess biodiversity and conservation needs. The following provides a more detailed summary of the approaches used in this study in Ontario and the United States.
In order to create manageable units for the analysis, islands were grouped and analyzed based on the Great Lakes coastal environment (Owens, 1979) in which they occur. This was done to facilitate the comparison and ranking of islands that shared a similar geography and general physical characteristics. Coastal environments are based on relief, geology, fetch, wave exposure, ice conditions, and availability and transport of sediment. This classification system divided some larger islands (e.g. Manitoulin) into different zones to reflect distinctive coastal characteristics.
Within each coastal environment islands were divided into two groups for analysis: 1) large islands (approximately 7–14 ha, depending on size range of islands in the coastal environment) and 2) island complexes. Large islands were maintained as a single unit of analysis, while small islands were assembled into “island complexes” using an automated process. This complexing was done because portions of the Great Lakes (such as eastern Georgian Bay) contain thousands of islands, many of which are very small, share similar characteristics, and often function as a unified landscape unit. Groups of small islands within each coastal environment were assembled into island complexes based on proximity (within 200 m of each other and without any intervening land) and similar geology. The resulting analysis was then done on the island complex, rather than small individual islands.
All large islands and island complexes were then attributed with data on biodiversity, land use and land protection, and assigned scores in three general categories: 1) biodiversity, 2) potential threats, and 3) existing conservation progress. The criteria from Ewert et al. (2004) were modified and used as a basis to build a scoring method that could use an automatic approach to assigning scores to the islands. The biodiversity score included a combination of 28 scoring criteria that were applied to each large island and island complex. Biodiversity criteria included measures for biological diversity, physical diversity, size and distinctiveness (Table 1). Islands were assigned a total biodiversity score based on all criteria, and the islands were then ranked within each coastal environment based on this score. Islands were also ranked based on score for select biodiversity criteria or groups of criteria that were selected by the working group as especially important for assessing the significance of islands. These included: colonial nesting waterbirds, species of global conservation concern, isolation and physical diversity (as a surrogate for biological diversity).
The analysis of threats used 17 scoring criteria to assess the existing human uses and potential vulnerability of islands. The threats analysis included an assessment of direct threats, such as boat launches, anchorages, residences, cottages, building density, invasive species, pits and quarries, lighthouses, distance to mining claims, road density and percent of the island occupied by cropland. Islands were assigned a threat score based on the total scoring of all threat criteria, and ranked based on threat scores within the coastal environment they occur.
Existing conservation and land protection policy was also assessed for each large island and island complex. Parks, protected areas, conservation lands and existing recognition of biodiversity values (e.g. Important Bird Areas) were categorized into four groups to reflect the general type of associated conservation. Existing conservation progress scores did not directly contribute to the biodiversity or threat scores, but the proportion of these conservation lands on each island and island complex were assessed to provide further insight into island values and identify potential conservation gaps and needs.
Highest scoring islands for biodiversity and threats within each coastal environment were identified based on the natural breaks (Jenks) method provided in ArcGIS software (Environmental Systems Research Institute Inc. 2002). Along with the protection gap analysis, potential priority islands and island complexes for conservation were then identified.
Biodiversity and conservation values of the US islands were based on a review of existing information and by updating basic spatial information. Many of these US islands have been inventoried by Michigan Natural Features Inventory and others (see Penskar et al., 2002; Penskar et al., 2000; Soule, 1993). Priorities for colonial nesting waterbirds have also been identified (Wires and Cuthbert, 2001) and basic island metrics have been developed for the shoreline units of Lake Huron (based on Reid et al., 2001). The US islands were assessed based on these information sources, the Great Lakes ecoregion plan in the United States (The Nature Conservancy, 1999), data on endangered, threatened, and special concern species available from the Michigan Natural Features Inventory, information from the Michigan chapter of The Nature Conservancy and Michigan Nature Association, and other published and unpublished descriptions of biodiversity. The protected status of islands in the Michigan portion of Lake Huron was determined from a newly developed data layer, CARL (Conservation and Recreational Lands database) and a review of threats to islands in Lake Huron derived from expert opinion and other sources (e.g. Vigmostad, 1999; Wires and Cuthbert, 2001).
Scores comparable to those developed in Ontario have not yet been developed as the databases needed to derive these scores were just becoming available as this paper was written. We anticipate being able to score islands in Michigan similar to those in Ontario shortly.
Biodiversity and threats
This project mapped over 23,000 island polygons within Lake Huron (Figure 1). It also created the first comprehensive dataset of islands within the United States portion of Lake Huron, and increased the number of documented US islands from 200 (Soule, 1999) to almost 600. Lake Huron islands range in size from less than a few square metres to Manitoulin Island at 2,766 km2. The vast majority of these islands occur within the northern and eastern portion of Georgian Bay, while large areas in the southern part of Lake Huron have very few islands (Table 2). Within all the Great Lakes, almost 75% of islands occur in Lake Huron. This is the largest collection of freshwater islands in the world.
The islands of Lake Huron are highly variable, but can be generally grouped into three major groups, based on their underlying geology and substrate, with each of these groups having different biodiversity features.
The largest group of islands in Lake Huron, and also in the entire Great Lakes region, occurs in northern and eastern Georgian Bay. These islands are on igneous rock of the Precambrian shield and are characteristically low, small, and found clustered within dense nearshore archipelagos. The most isolated islands include Gull and Papoose Island. The islands of northern and eastern Georgian Bay are the most significant in Lake Huron for disjunct flora from the arctic and western cordillera, Atlantic coastal plain plant communities, as well as northern populations of many reptiles and amphibians, including the globally rare eastern foxsnake (Elaphe gloydi) and eastern massasauga (Sistrurus catenatus). Over 70% of the entire coast of northern and eastern Georgian Bay occurs on islands. Biodiversity tends to be richer in the south (greater number of rare species), which corresponds to the areas of highest threat, including higher densities of roads and buildings and a greater number of marinas. Islands in the North Channel are the least threatened in Lake Huron. This is due to their distance from large population centres and, with 42% of the islands in parks and conservation reserves, a high level of land protection. Relatively lower biodiversity scores for species and communities for islands in the in the northeastern section of Georgian Bay may reflect a lack of field studies.
The northern coast of Lake Huron is characterized by a discontinuous archipelago of resistant limestone and dolostone islands that extends from the scattered islands off the southern Bruce Peninsula west to Manitoulin Island, St. Joseph Island, the Les Cheneaux islands and the Mackinac and Bois Blanc islands. These islands are especially important for a concentration of globally rare species and communities endemic to the Great Lakes, ranging from species such as dwarf lake iris (Iris lacustris), lakeside daisy (Hymenoxys herbacea) and Pitcher's thistle (Cirsium pitcheri) to communities such as Great Lakes alkaline cobble/gravel shore, limestone bedrock lakeshore and wooded dune and swales. Lake Huron northern islands are particularly noteworthy for the concentration of species and communities associated with limestone and dolostone bedrock and include Manitoulin Island, Drummond Island, and smaller islands associated with these two large islands. Some of the highest quality alvars and best sites for lakeside daisy and dwarf lake iris in the world are found on these islands and island complexes (Reschke et al., 1999). These islands are the highest scoring in Lake Huron for biodiversity values. Within the Ontario portion of the Great Lakes, they are second only to islands in the Western Lake Erie basin. Larger islands with road access (e.g. Manitoulin, Great La Cloche) have some of the highest threats scores in Lakek Huron.
Finally, the small group of Lake Huron islands in Saginaw Bay, Michigan supports vast areas of Great Lakes marshes that, together with the adjacent mainland, provide cover and food resources for large numbers of migrating waterfowl and shorebirds. Nearshore waters around these islands support spawning and nursery areas for many species of fish. Some beach ridges also support distinctive prairie and savanna communities.
While some parts of Lake Huron have very few islands, those that occur are important for biodiversity conservation. The Ontario portion of Lake Huron between the southern boundary of the Bruce Peninsula and the St. Clair River has only 39 islands, but these islands have one of the highest mean biodiversity scores in Lake Huron. Biodiversity values include colonial nesting waterbirds (Chantry Island in particular), coastal wetlands and fish habitat.
While the level and types of threats vary greatly on the islands of Lake Huron, two significant threats dominate: 1) incompatible development, especially in the Les Cheneaux, southeastern Georgian Bay region and along the Bruce Peninsula and Manitoulin Island, which results in habitat loss, including fragmentation, and degradation of natural processes in and nearshore waters, and 2) the spread of invasive species, particularly in Saginaw Bay where islands under public ownership are being invaded by non-native animal and plant species such as common reed (Phragmites australis), zebra mussel (Dreissena polymorpha) and common carp (Cyprinus carpio) that may alter ecological and trophic-level dynamics. Other threats to some islands include modification of ecological communities due to overbrowsing by introduced deer, and the potential effects of climate change. Threats related to recreation, shoreline hardening, alteration of substrate in nearshore waters due to dredging, and contaminants may also have consequences for the biota and processes that maintain the islands in Lake Huron.
In general, large, nearshore islands with road and ferry access are the most threatened and have the least amount of protected areas (e.g. Manitoulin Island). Several large privately-owned islands that are isolated have very low threat scores, such as Great Duck, Cockburn and Fitzwilliam. Some larger islands that are protected have relatively high threat scores (e.g. Flowerpot Island and Beausoleil Island) due to existing recreational infrastructure.
There is great variation in the conservation status of Lake Huron islands. Generally, islands in the southern areas tend to have less protection and greater competing land uses than regions in the north. For example, while almost 50% of the islands of eastern and northern Georgian Bay are within regulated protected areas, almost none of the islands in the East Christian Island Peninsula and Nottawasaga Bay region to the south are protected.
Many islands in the southern Lake Huron portion of Michigan are protected. Most islands in Saginaw Bay, the southernmost islands in the US portion of Lake Huron, are under State of Michigan or US government ownership. In addition, many islands of the Thunder Bay region, near Alpena, Michigan, are protected as part of the Michigan Islands National Wildlife Refuge or by the Michigan Nature Association. In the northern Lake Huron portion of Michigan, a smaller proportion of islands (or parts of islands) are under public or non-governmental ownership. This reflects the much larger number of islands in that region. Round Island, near Mackinac Island, is a federally designated Wilderness Area.
This study has identified priority islands for biodiversity conservation within the Ontario portion of Lake Huron. Table 3 highlights the islands with the top three total biodiversity scores for the nine Lake Huron coastal environments in Ontario. These islands and island complexes typically have higher biodiversity scores compared to other islands in the coastal environment due to diverse ecological systems (e.g. Strawberry Island), high biological diversity (e.g. Kettle Point Island complex) and distinctive features (e.g. Great Duck Island). Many islands that scored high for biological diversity also had high scores for physical diversity. As physical diversity was measured from remote data sources, some islands identified for high physical diversity, but with limited direct information on biodiversity, may be candidates for more detailed field studies (e.g. Browning Island complex).
The database generated for this project will have multiple applications in advancing the conservation of the biodiversity of Lake Huron. In addition to highlighting islands with high biodiversity values, and islands with a high threats (Figure 2) to help direct conservation efforts, the information can be used to identify and map particular aspects of island biodiversity such as globally rare species, fish habitat, and isolated islands with forest cover, which may be important emergency stop-over sites for migratory songbirds. The information can also be used by researchers to identify gaps in inventories by comparing islands with documented species and communities with islands that lack field studies, but share similar physical features such as size, shoreline types and isolation.
Protection of Lake Huron islands is progressing but much more needs to be done. Remaining protection is needed to focus not only on the species and communities of concern, but also the ecological processes needed to maintain these islands. Islands need to be integrated into both regional and local conservation and land use planning to recognize the distinctive needs and high importance of these unique systems.
Islands are a key component of Lake Huron's biodiversity and form a significant portion of the lake's coastal ecosystems. The islands are extremely diverse in terms of size, isolation, geology and biology. The three major groups of islands described in this paper are defined largely by geology and, to some degree, by latitude. These islands support distinctive plant and animal communities whose expression has been shaped coastal processes and isolation. The result has been the development of a rich and diverse set of species and communities on Lake Huron islands which have been somewhat buffered from anthropogenic change due to their isolation. Consequently, many islands remain undeveloped and in good condition ecologically compared to the mainland. This is especially true for islands south of the Precambrian Shield.
Relative to the Lake Huron mainland shoreline, islands are disproportionately valuable as colonial nesting waterbird sites, as sites that support Great Lakes endemic flora, fauna and communities characteristic of limestone and dolomite, and for concentrations of communities such as Great Lakes marsh in Saginaw Bay. And from a Great Lakes-wide perspective, the Lake Huron islands, as the world's largest collection of freshwater islands, also emerge as globally significant sites.
Funding for this project has been provided by the US Environmental Protection Agency – Great Lakes Program Office and the Canada-Ontario Agreement. Information for this paper was created through the Bi-nation Framework for the Conservation of Great Lakes Island project. In addition to the authors, this project team includes: M. Seymour (US Fish and Wildlife Service), R. Greenwood (EPA), K. Vigmostad (IJC), L. Wires (University of Minnesota), F. Cuthbert (University of Minnesota). Team members for the Ontario portion of the project includes: W. Bakowsky (NHIC), B. Crins (Ontario Parks), J. Mackenzie (NHIC) and M. McMurtry (NHIC). GIS and technical support for this project has been provided by G. White (The Nature Conservancy of Canada), J. Slatts (The Nature Conservancy), T. Krahn (Provincial Geomatics Service Centre, OMNR). The authors also wish to thank two anonymous reviewers.