Amphibians are highly adapted for life in wetland habitats. They form a major component of wetland faunas, and being both prey and predator, they are important in ecosystem functioning. Wetlands provide aquatic habitats that amphibians require for breeding, development, foraging, hibernation and refuge, and they form an interface with essential adjacent upland habitat. The size and type of wetlands as well as their spatial configuration and local structural characteristics are important features of these habitats for amphibian use. Because of their dependence on water, use of both aquatic and terrestrial habitat, permeable skin, and other biological characteristics, amphibians are considered to be excellent indicators of ecosystem health. Amphibians have a tremendous diversity of natural history characteristics and species differ in their patterns of habitat and microhabitat selection. Over 30 species of amphibians occur in wetlands within the Great Lakes Basin and an increasing trend in species richness exists from north to south across the region. Since European settlement, this region has lost over 50% of its wetlands. Loss rates of coastal and inland wetlands exceed 90% in some areas. Many restoration efforts are underway across the region but losses still exceed gains. No species have been extirpated from the entire basin but numerous local extirpations have occurred. However, nearly half of the species are officially designated as being of conservation concern somewhere in the basin. A more realistic estimate suggests that at least 2/3 of species are of concern. Habitat loss is reported as the primary cause of decline for 60% of species and habitat degradation by pollution is cited for 43% of the fauna. Considering the extent of wetland loss across the basin it seems reasonable to assume a similar magnitude of amphibian population loss. The current conservation status of amphibians indicates that Great Lakes wetlands are unhealthy ecosystems.
Amphibians have been a major faunal component of wetland habitats for at least 350 million years. Arguably, no other class of animal is so highly evolved to live in wetlands. As the origin of their name implies (Gk amphi- both, bios- life), amphibians lead dual lives. Larvae of most species are aquatic while adults tend to be terrestrial. However, much variation exists among species in life cycles and habitat requirements with some being entirely aquatic and others being completely terrestrial. Because of their complex life cycles most amphibians use both terrestrial and aquatic habitats to some degree (Wilbur, 1980, 1984). The importance of amphibians is often overlooked in both habitats. They often reach high densities and play pivotal roles in food webs by acting as both prey and predator (Stebbins and Cohen, 1997; Mitsch and Gosselink, 2000). Larvae tend to be herbivorous (anurans [frogs and toads]) or carnivorous (salamanders) while adults are solely carnivorous (Wilbur, 1980, 1984). Feeding by tadpoles can reduce algal population abundance and alter community structure (Seale, 1980; Holomuzki and Hemphill, 1996; Kupferberg, 1997; Loman, 2001). Predation by adult frogs may impact arthropod communities similarly (Johnson and Christiansen, 1976; Vogt, 1981). Salamanders or newts can function as keystone predators and dominate food webs (Morin, 1981; Strohmeier et al., 1989; Holomuzki et al., 1994). In terrestrial habitats such as the eastern deciduous forest, biomass of salamanders alone can exceed that of all birds and be equivalent to small mammals (Burton and Likens, 1975).
Despite the wide variation in natural histories that have evolved among amphibians, all species are still highly dependent on water, even most terrestrial species must return to wetlands at some point in their lifecycle. This strong association with their primordial habitat persists for at least two important reasons; their skin is permeable, and their reproduction is dependent on water. It is no surprise that wetlands provide essential habitat for most amphibians to reproduce, forage, or hibernate and that amphibians form a conspicuous part, and often dominate wetland faunas. Because of their physiology, natural history, and lower vagility compared to other vertebrates, amphibians are considered to be excellent indicators of ecosystem health (Vitt et al., 1991; Vershinin, 1995).
From an evolutionary perspective, amphibians seem to be outstandingly successful. They have dominated wetland faunas for 350 million years, witnessed the demise of the dinosaurs, and they survived the last ice age without species extinctions in North America (Holman, 1995). However, in recent years accumulating evidence indicates that amphibians are in serious decline on a global scale (Barinaga, 1990; Alford and Richards, 1999, Houlahan et al., 2000). Factors that have been implicated are largely anthropogenic and include: habitat loss and fragmentation, pollution, acidic precipitation, ozone depletion and increased UV-b, climate change, species introductions, over-harvesting, and disease (see Alford and Richards, 1999 for review). The consensus among herpetologists is that habitat loss and fragmentation are the most important factors globally, but declines may have multiple causes and differ among populations.
Natural habitats in the Great Lakes Region of North America have been strongly altered by a long history of human land use and high human population density. Relative to pre-European settlement, wetlands have been severely reduced in both quantity and quality (Mitsch and Gosselink, 2000). Because of their biological characteristics and dependence on wetlands, amphibians were chosen as indicators of ecosystem health in Great Lakes wetlands (Weeber, 2000). My objective is to provide a general review of the association between amphibians and wetland habitat with a focus on the Great Lakes. Reviews can never be complete and are limited by the experience and expertise of the reviewer. My review is undoubtedly biased by my own experience and any omissions of relevant information were not intended.
To assemble information, I searched electronic literature sources (Web of Science, Bibliomania, Biosis) and government internet sites. I also requested information on surveys, monitoring and research from various government agencies of each state or provincial jurisdiction (Minnesota, Wisconsin, Michigan, Illinois, Indiana, Ohio, Pennsylvania, New York, Ontario) within the Great Lakes Basin. To determine the list of amphibian species, their habitat requirements, and information on declines, I summarized information from many sources including field guides, regional accounts, government sources, COSEWIC reports, and published literature. There are surprisingly few publications dealing directly with wetlands as amphibian habitat with a Great Lakes focus. I used Great Lakes examples when possible, but referred to the general amphibian literature otherwise.
The Laurentian Great Lakes of North America and its drainage basin is a large and diverse ecosystem. It spans about 800 km (N-S) by 1200 km (E-W). About 18% of the world's freshwater is contained within its five major lakes (Superior, Huron, Michigan, Erie, Ontario) smaller lakes and tributaries. The basin is inhabited by 10 and 25% of the human populations of the United States and Canada respectively (Botts and Krushelnicki, 1988) and it extends from the boreal forest in the north to the eastern deciduous forest in the south and includes prairie elements in the west. Geologically, the area is also diverse including pre-Cambrian granitic rock of the Canadian Shield in the north to Devonian and Ordovician limestone bedrock in the south. A variable temperate climate exists as a result of interactions between cold dry Arctic air, and warm moist air from the Gulf of Mexico. Many wetlands occur along portions of the Great Lakes coasts or are associated with some of the system's 35,000 islands. Many inland wetlands also occur in the basin ranging from those associated with lakes and rivers to marshes, swamps, bogs, fens and ponds. The physical and structural diversity occurring in Great Lakes wetlands and their adjacent uplands provides a diverse range of habitats for amphibians.
An accurate estimate of the extent of wetlands occurring in the Great Lakes is largely unknown (SOLEC, 2000; but see Mitsch and Gosselink, 2000). Nearly 1400 coastal marshes extend over 6,129 km and cover 1,209 km2 on the United States side of the Great Lakes (Herdendorf, 1987). The extent of coastal wetlands on the Canadian side is less well known but total wetland coverage for southern Ontario alone is estimated at 9,000 km2 (Glooschenko and Grondin, 1988). Estimates of loss of wetland habitat since European settlement began are more accurate. Based on knowledge of mapped wetlands, it is estimated that over 50% of their area was lost over the past century (Glooschenko and Grondin 1988, Dahl 1990). Over two-thirds of coastal wetlands in the Great Lakes have been lost (71–80% in lower lakes) (Environment Canada, 1995; Shirose et al., 1995; Mitsch and Gosselink, 2000). Losses within some jurisdictions in the basin have been even greater: Illinois (90%) (Suloway and Hubbell, 1994); Indiana (87%), Michigan (50%), Minnesota (42%), New York (60%), Ohio (90%), Pennsylvania (56%), Wisconsin (46%) (Dahl, 1990); and southern Ontario (90%) (Snell, 1987). These losses resulted primarily from conversion of wetlands to agricultural land use but also because of urban expansion and human recreation. No studies have investigated the impact of this habitat loss on the amphibian fauna at the spatial scale of the entire Great Lakes Basin.
Amphibian fauna of the Great Lakes
Species pool and habitat requirements
At least 30 or 31 species of amphibian occur in the Great Lakes basin (Harding, 1997; Conant and Collins, 1998; Table 1). Some authors recognize Boreal Chorus Frogs as a distinct species (Pseudacris maculata) rather than a subspecies of the Western Chorus Frog (P. triseriata). This number climbs to 35 if four peripheral species are included (see Harding, 1997). Most of these species have large overlapping geographic ranges (Conant and Collins, 1998). Although 15 of these species occur in each Great Lake (or individual basins) (Table 1), a decreasing latitudinal gradient of species richness occurs from south to north: Erie (28 spp.), Ontario (22), Michigan (21), Huron (17), Superior (17). This gradient results because half of the species reach their northern range limits in the basin.
The amphibians of the Great Lakes are a young fauna. They have only recently recolonized the Great Lakes from southern refugia in the post-Pleistocene (Holman, 1995, 1998). Retreat of the Wisconsin Ice Age glaciers across the basin occurred about 15,000 to 10,000 y BP (south to north). Because of isostatic and eustatic changes, the present configuration of the Great Lakes has only existed for about 4,000 to 5,000 y. Holman (1992) developed a model for post-glacial reinvasion of Michigan by amphibians based on known ecological tolerances of species. He designated species as ‘primary invaders’(can tolerate coniferous forest or tundra), ‘secondary invaders’ (mixed wood) or ‘tertiary invaders’ (deciduous forest). Fossil records for the region support his model and suggest that most amphibians reinvaded rapidly as the ice sheet retreated (Holman, 1995, 1998 and references within). Recent research in molecular phylogeography holds promise to shed light on directions of post-glacial dispersal corridors in the Great Lakes basin (e.g., Austin et al., 2002).
Habitat requirements of amphibian species in the Great Lakes vary widely (Table 2). Amphibians occur in many types of wetlands and aquatic habitats including: ephemeral pools, seepages, creeks, streams, rivers, ponds, bogs, fens, swamps, marshes and lakes. However, most species use both aquatic and terrestrial habitats to varying degrees at some points in their life cycles. Salamanders range from totally aquatic species and occur at considerable depths in lakes or rivers (e.g., mudpuppy), to completely terrestrial species occupying upland forests (e.g., Red-backed, Northern Slimy, and Ravine Salamanders). The Eastern Newt has aquatic larvae and adults, but the juvenile eft stage uses terrestrial habitat for several years. Habitat relationships of anurans are also diverse and complex. The Bullfrog, Green Frog, and Mink Frog are often considered to be ‘permanent water’ species because of their strong affinities for aquatic sites for breeding, foraging, hibernation, and their long development times requiring tadpoles to overwinter in ponds. Whereas Spring Peepers, Chorus Frogs, and Wood Frogs are often considered to be ‘temporary pond’ species because their short development times permit use of ephemeral ponds and adults have affinities for terrestrial habitats. Another habitat classification that is often made is to distinguish between ‘arboreal’ species (e.g., Grey and Cope's Treefrogs) with more ‘terrestrial’ species (e.g., American Toad, Leopard Frog, Wood Frog) or ‘fossorial’ species such as mole salamanders (Spotted, Blue-spotted, and Jefferson complex Salamanders). Species can also be distinguished along a habitat generalist-specialist gradient. American Toads are considered to be habitat generalists or opportunistic species that seem to be able to tolerate a wide range of breeding sites or foraging habitats. However, other species are more specialized such as the Pickerel Frog which tends to occur only in cold waters associated with forest habitat and the Four-toed Salamander with sphagnum bogs. Harding (1997) reviewed specific natural history information for all Great Lakes amphibian species. Reviews of habitat associations are also covered in several regional field guides (see bibliography in Harding 1997; Matson, 2000; Hulse et al., 2002; MacCulloch, 2002).
Conservation status of Great Lakes amphibians
Nearly half (17 of 35) of the amphibian species occurring in the Great Lakes basin are officially designated as being of conservation concern somewhere in the basin (Table 3). These designations include: ‘special concern,’ ‘threatened,’ ‘endangered’ and ‘extirpated.’ Because of the lengthy process from initial assessment to legally designating species, this estimate of status is likely conservative. For example, although the Northern Cricket Frog was known to be in decline since the 1970s, it was not designated as ‘endangered’ until 1990 in Canada. However, the species is likely extirpated in Canada because no records exist since 1990 despite nearly annual monitoring. Evidence of decline has also been reported for an additional six species. A more realistic estimate of status would be that about 2/3 of species are experiencing conservation concerns at least somewhere in the basin. Although numerous causes of decline are suggested, habitat loss and degradation appear to be most common (Table 3). Habitat loss has been reported for 60% (21 of 35) of species, pollution for 43% (15 of 35), over-harvesting for 14% (5 of 35), and disease for 6% (2 of 35).
Characteristics of wetlands affecting amphibians
Many biotic and abiotic characteristics of wetlands can affect the composition of amphibian communities and abundance of individuals in populations (Semlitsch, 2000). The most fundamental factors concern area, permanence of water, spatial configuration of wetlands and adjacent upland habitats, local habitat and microhabitat characteristics, and water chemistry and quality.
The species area effect (see Rosenzweig, 1995) indicates that larger areas should have higher diversity and larger population sizes of species they contain. Amphibians seem to largely follow this general principle of nature over wide spatial scales (Hecnar, in press) as do most other organisms. For example, large coastal wetlands such as the marshes at Point Pelee, Rondeau, and Long Point on Lake Erie have 11(historically), 14, and 16 species respectively (OHS database), while communities at small ponds in these areas usually have 4 or fewer species present (Hecnar and M'Closkey, 1998). Studies of amphibian faunas on islands in the Great Lakes also clearly demonstrate the importance of habitat area to species diversity (e.g., see King et al., 1997; Hagar, 1998; and Hecnar et al., 2002). Larger areas can promote amphibian species richness because they provide greater habitat diversity (Ricklefs and Lovette 1999).
Another important abiotic factor besides the areal extent of a wetland is the permanence of its water (hydroperiod). Amphibian species in the Great Lakes have development times ranging from weeks (e.g., American toad) up to two years (e.g., Bullfrog). Hydroperiod influences not only the numbers and sizes of individuals metamorphosing, but it also determines which species successfully develop and ultimately species richness of a wetland (Collins and Wilbur, 1979; Pechmann et al., 1989; Rowe and Dunson, 1995; Skelly, 1996; Wellborn et al., 1996; Snodgrass et al., 2000; Paton and Crouch, 2002). Therefore, fluctuating water levels and drought associated with climate warming (Croley et al., 1998; Schindler, 1998) may have important impacts on the Great Lakes fauna. Theory predicts that two of the most important factors affecting amphibian communities are the opposing stresses of decreasing desiccation and increasing predation as hydroperiod increases from temporary ponds to permanent waters such as lakes (Heyer et al., 1975; Wilbur, 1984).
The spatial configuration of wetlands and upland habitats is also important. Many amphibian species appear to demonstrate metapopulation spatial dynamics (Gill, 1978; Gibbs, 1993; Hecnar and M'Closkey, 1996a; Marsh and Trenham, 2001). In metapopulations, local extinctions can commonly occur but recolonization or rescue from adjacent sites allows species to persist regionally. Because most species also use terrestrial habitats, the proximity of upland habitat and the nature of the landscape between wetlands is very important (Dodd and Cade, 1998; Pope et al., 2000). Barriers to dispersal or migration between habitats such as roads can increase mortality and decrease connectivity among amphibian populations (Fahrig et al., 1995; Ashley and Robinson, 1996; Gibbs, 1998; Hels and Buchwald, 2001). Differences among species in dispersal capability and requirement of forest habitat are factors contributing to highly nested amphibian assemblages in the Great Lakes Basin (Hecnar and M'Closkey 1997a).
Amphibian species richness at wetlands is strongly associated with the extent of regional forest cover, level of urbanization, and history of land-use (Findlay and Houlahan, 1997; Hecnar and M'Closkey, 1998; Kolozvary and Swihart, 1999; Lehtinen et al., 1999; Knutson et al., 1999, 2000; Findlay et al., 2001). Forest harvesting can negatively affect amphibians by reducing habitat heterogeneity, reducing moisture levels, and restricting movements (Petranka et al., 1994; de Maynadier and Hunter, 1995, 1998; Gibbs, 1998).
Local habitat characteristics or microhabitat features can also affect habitat suitability for particular amphibian species and ultimately species richness. The nature and extent of emergent, floating or submerged vegetation, debris, shoreline features and adjacent vegetation, depth, substrates provide different microhabitats that can be exploited by individual species (Hecnar and M'Closkey, 1998; Bunnell and Zampella, 1999). The quality of shoreline features and emergent vegetation are important in territorial amphibian species for breeding sites that maximize fitness (Martof, 1953; Emlen, 1976; Wells, 1977). Removal of natural vegetation during shoreline development degrades habitat quality and reduces amphibian abundance (Woodford and Meyer, 2003). Emergent or floating vegetation provides calling sites for anurans and submerged vegetation provides habitat for newts (Stebbins and Cohen, 1997). Submerged debris or vegetation and underwater portions of stems of emergent vegetation provide structure for attachment of egg masses (Stebbins and Cohen, 1997; Hecnar and Hecnar, 1999). The structure of vegetation and shoreline features (e.g., shallows) may also provide important ‘nursery’ areas or refuges from predators (Sredl and Collins, 1992). Tadpoles often aggregate in shallows where warmer waters accelerate development (e.g., Brattstrom, 1962).
Predation is a major factor affecting amphibian populations and communities (Wilbur, 1980). Some amphibians have evolved chemical or behavioural anti-predator defences while others remain vulnerable (Kats et al., 1988). Large predatory fish are efficient predators that can restrict amphibian distribution, alter behaviour, or even eliminate some amphibian species (Collins and Wilbur, 1979; Sexton and Phillips, 1986; Wellborn et al., 1996; Hecnar and M'Closkey; 1997b, Smith et al., 1999). Predation is especially important when fish are introduced to wetlands that were previously fish-free (Bradford et al., 1993; Braña et al., 1996) or if introduced fish carry infectious diseases such as saprolegnia or chytrid fungi (Blaustein et al., 1994; Carey, 2000).
General water chemistry and water quality also play important roles in distribution of amphibians among wetlands (Dale et al., 1985; Kutka and Bachmann, 1990; Hecnar and M'Closkey, 1996b). Acidic waters increase mortality but tolerance differs among species (Clark, 1986a,b; Ling et al., 1986; Kutka and Bachmann, 1990; Karns, 1992). Low buffering capacity and the presence of metals in water can also affect amphibian distribution (Glooschenko et al., 1992). Waters polluted by fertilizers, pesticides, or industrial contaminants have toxic effects on many species of amphibians (Berrill et al., 1995; Hecnar, 1995; Russell et al., 1995, 1999, 2002; Clements et al., 1997; Bishop and Gendron, 1998; Glennemeier and Begnoche, 2002). There has also been much recent interest in determining the role of contaminants in causing deformities or endocrine disruption in amphibians (Ouelette et al., 1997; Reeder et al., 1998; Gillilland et al., 2001; Harris et al., 2001; Withgott, 2002).
Research and conservation
Understanding species distribution is of fundamental importance for both research and conservation. Large-scale distributions (ranges) of most amphibian species occurring in the basin are largely known although northern range boundaries for some are not well-established. Patterns of smaller-scale distribution (local) are not as well-known and patchy especially in northern parts of the basin. Several atlas projects have been initiated that are improving the resolution of species ranges and documenting habitat use in the basin. Most of these projects collect location data from museum records and species observations submitted by biologists and naturalists. The data are typically scrutinized and summarized to produce grid maps (resolution, 1–10 km2).
The Ontario Herpetofaunal Summary (>130,000 records) was conceived in 1984 by the Ontario Field Herpetologists (Weller and Oldham, 1988) and is now managed by the Ontario Ministry of Natural Resources-Natural History Information Centre. Atlases for several states have also been initiated and modelled largely upon the Ontario Herpetofaunal Summary. The Wisconsin Herpetological Atlas Project operated by the Vertebrate Zoology Section of the Milwaukee Public Museum and sponsored by the Wisconsin Department of Natural Resources and the Nature Conservancy started in 1986. The New York State Amphibian and Reptile Atlas Project was operated between 1990 and 1999 by the New York State Department of Environmental Conservation. The Pennsylvania Herpetological Atlas Project (>40,000 records) sponsored by Resource Conservation Fund and Indiana University of Pennsylvania began in 1996.
Historically, there was relatively little interest in documenting amphibian presence in Great Lakes wetlands aside from occasional museum collecting trips and some published notes (e.g., Logier, 1925; Ruthven et al., 1928; Adams and Clark, 1958). However, recent concerns about large-scale amphibian declines have motivated the development of numerous large-scale, long-term monitoring efforts in every jurisdiction in the Great Lakes Basin. Most of these programs use trained volunteers to conduct auditory surveys of anurans and are co-ordinated by government environment or natural resources departments. The North American Amphibian Monitoring Program (NAAMP) managed by the USGS Patuxent Wildlife Research Center co-ordinates data at the continental level and functions to standardise protocols. Ontario data are coordinated by the Canadian Wildlife Service. Programs fall within three broad categories: backyard or pond monitoring (single locations), road call surveys (multiple stations along fixed routes), and marsh monitoring (multiple plots). Field protocols are largely based on those originally developed in Wisconsin (Mossman and Hine, 1984). Typically point sample stations at fixed distances along rural roads or in marshes are visited three or four times per year. Observers conduct nocturnal surveys listening for breeding calls for several minutes to identify species present and to determine rank abundance. Data on weather conditions and habitat are usually also recorded. Road call surveys began in Wisconsin (1981) followed by Illinois (1991), Ontario (1992), Minnesota (1993), Michigan (1996), Pennsylvania (2000), New York, and Ohio. The Marsh Monitoring Program (MMP) developed in 1994 by Bird Studies Canada and Environment Canada is now supported by the U.S. EPA and other partners. The MMP uses volunteers to conduct amphibian calling surveys at over 500 locations in the Great Lakes (Weeber and Vallianatos, 2000).
These survey programs permit relatively inexpensive data collection on large spatial scales that can be used to detect temporal trends. Shortcomings include problems associated with using volunteer-based programs and the quality of presence and abundance data collected. Populations of temperate zone amphibians are widely known to fluctuate in their diurnal and seasonal activity patterns and annual abundance making assessment of status difficult to determine (Pechmann et al., 1991; Bridges and Dorcas, 2000). Because survey programs only use a limited number of visits annually, detectability of calls can differ among species, only anurans are surveyed (salamanders do not call), and there are concerns regarding low power to detect declines (Reed and Blaustein, 1995). Methods for adjusting species occupancy when detection probability is <1 are suggested (Mackenzie et al., 2002). Although many large-scale survey programs are now in place, validation studies have been conducted only in Ontario (Shirose et al., 1997) and Rhode Island (Crouch and Paton, 2002). These validation studies indicate that call indices are correlated with abundance, interobserver agreement is reasonably high, and that common species are detected within several minutes of listening.
The USGS Northern Prairie Wildlife Research Center manages the North American Reporting Center for Amphibian Malformations (NARCAM) which coordinates reports and analysis of patterns of deformities across the continent.
Shortly after the first alarms regarding global amphibian decline were sounded (Wake and Morowitz, 1991), the Species Survival Commission of the International Union for the Conservation of Nature (IUCN) formed the Declining Amphibian Populations Task Force and working groups soon formed in the Great Lakes region (Declining Amphibian Populations in Canada (DAPCAN), Central Region, Great Lakes Region). These working groups unite hundreds of biologists and resource managers from the Great Lakes region who are concerned with amphibian conservation and function to exchange ideas, promote communication, and disseminate information to the public, media and governments. Working groups hold annual meetings and have produced several key publications (Bishop and Pettit, 1992; Green, 1997; Lannoo, 1998, Seburn and Seburn, 2000). A reference text on standard field protocols for amphibians was also published (Heyer et al., 1994).
Widespread historical wetland loss has stimulated much recent interest in wetland restoration and rehabilitation (e.g., Detenbeck et al., 1999; Wilcox and Whillans, 1999). Projects have been conducted or are planned in virtually every jurisdiction in the Great Lakes Basin. For example, in Ontario over 12,000 ha of wetlands were rehabilitated under the Great Lakes Wetlands Conservation Action Plan by 2001 and a goal of 30,000 ha is set for 2020 (Shirose et al., 1995; Environment Canada, 1995). Major projects are underway at Black Creek on Lake Huron, and Cootes Paradise and Oshawa Second Marsh on Lake Ontario. In Ohio, restoration projects increased wetland area by 7,200 ha by 2000 (Ohio DNR, 2001). In Illinois, over 7,800 ha of wetland were restored since the 1940s and wetland creation is now thought to balance wetland loss statewide (Havera and Suloway, 1994). Natural history traits of amphibians such as their high reproductive output, dispersal capabilities, ectothermic metabolism, and small home range requirements seem to preadapt them for colonization. Some amphibian species quickly colonize new wetlands suggesting that wetland creation provides at least partial mitigation for losses of original wetlands (Lehtinen and Galatowitsch, 2001; Pechmann et al., 2001). However, translocation of amphibians as a conservation strategy appears to be largely ineffective (Dodd and Seigel, 1991; Seigel and Dodd, 2002). Despite recent progress through restoration, losses of wetlands still exceed gains.
Overview and summary
Wetlands in the Great Lakes of North America support a diverse amphibian fauna that possess a wide range of adaptations for life in wetland habitats. Studies of many of these populations have contributed greatly to advancing our knowledge of amphibian ecology and distribution. However, fine-scale distribution and northern boundaries remain incompletely documented for many species, and there are many gaps in our knowledge of amphibians and habitat use and in natural history and ecology of populations in the Great Lakes Basin. Although the status of amphibians prior to European settlement of the basin remains largely unknown, it seems reasonable to assume that the large extensive wetlands that formerly covered much of the coasts of the lakes and inland areas of the basin supported species rich amphibian communities characterised by large populations. All populations carry a risk of extinction (Ehrlich and Daily, 1993) but the risk decreases for larger populations (Soulé, 1987). Similarly, adding more populations to a metapopulation spreads the risk of extinction (Hanski, 1999). It seems reasonable to assume that historical losses of 50 to 90% of these wetlands has resulted in concomitant losses of 50 to 90% of amphibian populations. Fragmentation of these wetlands has also resulted in greater isolation where distances between fragments are often too great to permit dispersal. Where dispersal can occur, barriers or filters are imposed between fragments by the massive network of roads that cover eastern North America (Reh and Seitz, 1990; Forman and Alexander, 1998). These habitat fragments have also been degraded by edge effects and inputs of pollutants.
A case in evidence is the coastal marsh ecosystems of the western basin of Lake Erie. On the north shore lies the RAMSAR recognized marsh in Point Pelee National Park, one of the largest coastal wetland ecosystems (10 km2) remaining in the Great Lakes. In the late 1800s drainage for agriculture reduced the marsh to about 35% of its original extent. Over the past century the number of amphibian species has declined from 11 to 5 (Hecnar, 1999). The remaining species carry heavy body burdens of contaminants and several are completely isolated from other populations (Russell et al., 1995, 1999). Similarly to Point Pelee, the marsh on Pelee Island (2225 ha) in Lake Erie was drained for agriculture and likely played an important role in the apparent extirpation of the Northern Cricket Frog in Canada. Losses were even greater on the south shore of the lake. Marsh once stretched from near Sandusky to Detroit covering over 1,200 km2 but 90% was drained for agriculture and industry in the time span of one human generation (Ohio DNR, 2001).
Although local extirpations may be relatively common, such as the loss of ‘forest’ species from southwestern Ontario (Hecnar and M'Closkey, 1996a, 1998; Hecnar, 1997), no amphibian species has yet been extirpated across the entire basin. However, the conservation status of some species suggests that they are close to regional extinction (e.g., Lehtinen, 2002). The fact that we have not lost any species regionally should not be taken as evidence that habitat appropriation for human use is not at a critical level. This is the case simply because amphibian species occurring in the Great Lakes have large geographic ranges (i.e., Rapoport effect; Moffatt, 2000). If the equivalent level of habitat loss and degradation occurred in the tropics, numerous species level extinctions would occur (Hecnar, in press). Because wetlands are among the most productive biomes (Brown and Lomolino, 1998) and amphibians play a key role in their functioning, loss of so much habitat and so many populations has far-ranging ecological impacts. Even local extinctions of single species can alter amphibian community structure (Hecnar and M'Closkey, 1997c).
Our strategy in wetland conservation is primarily to enact legislation to protect our remaining large wetlands. This is an important effort but wetland loss continues because most of the remaining wetlands are too small or too isolated to be afforded protection by legislation or adjacent terrestrial habitats are overlooked (Dodd and Cade, 1998; U.S. DAPTF, 2001). In some areas small wetlands associated with ponds (often anthropogenic) have replaced large natural wetlands (Merendino et al., 1995). Many amphibian species also commonly use these smaller unprotected wetlands (Gibbs, 1993; Semlitsch and Bodie, 1998).
It is clear that habitat loss and degradation has had, and continues to have, a substantial negative effect on amphibians in the Great Lakes. Considering that amphibians are excellent indicators of ecosystem health (Vitt et al., 1991; Weeber, 2000) and that half to two-thirds of amphibian species are of conservation concern leads to the inevitable conclusion that many Great Lakes wetlands are unhealthy.
I thank T. Mayer for inviting this review; A. Breisch, and P. Riexinger, New York State Department of Environmental Conservation; C. Caldwell, Ohio Division of Wildlife; G. Casper, Milwaukee Public Museum; K. Genet, Michigan State University; R. Hay, Wisconsin Department of Natural Resources (DNR); A. Plocher, Illinois Natural History Survey; L. Sargent, Michigan DNR; A. Schiffli, Indiana DNR; R. Tibbott, Pennsylvania Fish & Boat Commission; W. Weller, Ontario Field Herpetologists; M. Oldham, Ontario MNR- Natural History Information Centre; B. Eliason, Minnesota DNR; and F. Schuler, Eastern Ontario Biodiversity Museum; for providing information. D. Hecnar and two anonymous referees provided valuable comments on the manuscript. Funding was provided through PREA and NSERC grants to the author.