Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards, as per the Policy on Communications and Federal Identity.
COSEWIC Assessment and Status Report on the Yellow Rail in Canada
- Assessment Summary
- Executive Summary: from the 1999 Status Report
- COSEWIC Mandate, Membership and Definitions
- Lists of Figures and Tables
- Species Information
- Population Size and Trend
- Summary of Population Size, Number of Breeding Localities and Trends
- General Biology
- Limiting Factors
- Special Significance of the Species
- Evaluation and Proposed Status
- Technical Summary
- Acknowledgements, Biographical Summary of the Authors, and Literature Cited
Population Size and Trend
The heart of the breeding range in the U.S. is in Minnesota, where recent work has boosted the number of “element occurrences” (The Nature Conservancy 1998) to 178, which can be aggregated into 70 “sites” (R. Baker, pers. comm.; Table 2). McGregor Marsh (Minnesota), the best known site, has about 30 pairs (Eckert 1983, Savaloja 1984). The Seney National Wildlife Refuge in northern Michigan, the best known of that state’s estimated 5-10 sites (as inferred from the S1S2 rank), had about 52 singing males in the early 1980s (Bart et al., 1984, Bookhout 1995), and one to 85 singing males from 1995 to 1998 (R. Urbanek, pers. comm.). After the south-central Oregon population was re-discovered in 1982 (Stern et al., 1993), research in the ensuing years showed that there were three or four general sites (Wood River Valley, Klamath, Sycan, and Big Marsh). In 1998, only three of the sites had rails and a total of 128 were heard. K. Popper (pers. comm.) estimates 200 pairs for the whole population. In Montana, the Yellow Rail breeds only in the far northeastern corner of the state in northeastern Sheridan County. The total number of records for the state is less than 10, but it appears that the species breeds in this area regularly. There are probably fewer than five localities in the state (P. Hendricks, pers. comm.). Given their SRANKS, and without any further information, North Dakota and Wisconsin should have about 6-20 sites and about 1-5 sites, respectively (Table 2).
|Province, Territory or State||Number of pairs1||Number 6 of summer locations||Source and/or reasons||Population trend|
|Hudson/James Bays||A few thousand?||Strip of coastal habitat ca. 1700 km long.||Our estimate.||Stable? Snow Goose problem a concern.|
|NWTs||20-100||4–20||Our estimate based on small range.||Unknown.|
|Alberta||500+?||Hundreds?||40 known sites; much unsurveyed potential habitat.||Has likely declined. Still declining, but more slowly.|
|Saskatchewan||500+?||Hundreds?||60 known sites; much unsurveyed potential habitat (A. Smith).||Has likely declined. Still declining, but more slowly.|
|Manitoba (not including Hudson Bay)||500+?||Hundreds?||26 known sites; much unsurveyed potential habitat (P. Taylor, R. Koes).||Has likely declined. Still declining, but more slowly.|
|Ontario (not including Hudson/James Bays||Central Ont. (incl. Rainy River region): 115-125||4-6||Austen et al., 1994, D. Elder.||Has declined. Now stable?.|
|1-2||Austen et al., 1994|
Page and Cadman 1994
|Probably declined greatly|
|Québec (not including James Bay)||Interior Québec: 20-80||2-16||Our estimate.||Has probably declined.|
|Southern Québec: 20-80||6-15||Robert et al., 1995||Has declined, probably greatly.|
|New Brunswick||0-50||0-10||Erskine 1992||Has probably declined; now stable?|
|Nova Scotia||0-50||0-10||Erskine 1992||Unknown.|
|Canada Total||1687-2015 + a few thousand?||250-800 + a strip of coastal habitat ca. 1700 km long.||Numerous.||Has likely declined and continues to decline, albeit more slowly, throughout Canadian range, except in Hudson/James Bay region where there is a new concern.|
|Oregon||200||4||K. Popper||Habitat has declined greatly, but now seems stable.|
|Wyoming||10||1||D. Wile||No information|
|North Dakota||30-100||6-20||Inferred from SRANK of S1S2||No information.|
|Minnnesota||350||70||Habitat under constant threat.|
|Wisconsin||5-25||1-5||Inferred from SRANK of S1|
|Michigan||25-50||5-10||Best known site threatened.|
|U.S. Total||640-755||91-114||Numerous sources.||Has declined and continues to decline.|
|World Total||H/J Bays : a few thousand.|
The rest : 2327-2770 pairs.
|H/J Bays : 1700 km coast.|
The rest : 341-914 sites.
|Numerous sources.||H/J Bays : Stable until recent concern due to Snow Geese.|
The rest : Has declined and continues to decline, but more slowly.
1 When the number of pairs was estimated using the number of sites, we assumed a mean of 5 pairs per site.
We have made some rough estimates of the number of summer locations and the number of breeding pairs in each state using available information. Assuming an average of five pairs in each site (unless an estimate already exists), we estimate 600-750 pairs at about 100 sites in the U.S. (Table 2).
Wetland drainage is probably responsible for the loss of the southernmost breeding areas during this century, as well as the loss of several breeding sites located in the northern U.S. (Bookhout 1995). For example, in Oregon, ditching and draining of wetlands for agriculture have been responsible for the loss of several known breeding sites since 1985 (Stern et al., 1993). In Michigan, Yellow Rail habitat in the Seney Wildlife Refuge is threatened in the long term by the virtual elimination of the prescribed fire program, which itself is a result of heavier restrictions for prescribed burning (R. Urbanek, pers. comm.). In Minnesota, Yellow Rail habitat is under constant threat of either being drained or flooded for agricultural, industrial, and/or waterfowl management projects (Coffin and Pfannmuller 1988).
Overall, although the Yellow Rail has probably always been a rare species of limited distribution in the U.S., there is ample evidence that there has been some reduction in its numbers and range. It has been extirpated as a breeding species from at least three states, and there is also good evidence that many breeding sites have disappeared from states in which it still breeds. There is no evidence of increases in numbers or range anywhere in the U.S.
Estimating true numbers of Yellow Rails is particularly difficult on the wintering grounds, where the birds are not only secretive, but they do not vocalize or respond to recordings (K. Mizell, pers. comm.). Yellow Rails are especially common in the salt marsh zone along the Texas coast in March-April, but are difficult to locate during winter. In winter the species may rely more on fresh marsh areas (C. Cordes, pers. comm.). It is clear that at least some wintering habitat types used commonly are declining under continual pressure (see Habitat Trends).
Given the species’ small range in the Northwest Terrritories, we suggest that no more than 100 pairs may breed in this area, which represents the northern extremity of its range; but this is no more than an educated guess.
Several years ago there were only four records of the Yellow Rail in British Columbia (Campbell et al., 1990), but the species is now recorded almost every year somewhere in the province. Nowhere has it been recorded year after year (S. Cannings, pers. comm.), but because of the small number of surveys conducted, this does not necessarily mean that it has not occupied any sites in successive years. Nevertheless, the recent increase in observations is likely due to an increased effort to survey wetlands in the province’s interior for birds, and to include night-time observations (W. Campbell, pers. comm.).
The Yellow Rail is thought to be locally common (McGillivray and Semenchuk 1998), although efforts to census them have been few and without significant success (Semenchuk 1992). There are 40 known locations of presence during the breeding season (Table 2). While there were no reports of Yellow Rails in Alberta from 1961 to 1970 (T. Sadler, pers. comm.), there were 18 from 1973 to 1980. This change corresponded to an explosion of ornithological data in the 1970s (Pinel et al., 1991).
The species was reported in only 16 of 2206 squares surveyed during the Atlas. Breeding was not confirmed anywhere, but was considered probable in six squares and possible in 10 (Semenchuk 1992). No information is available on population trends per se, but the species’ habitat has been eroded considerably, largely for agriculture, and it continues to be eroded for agriculture, and more recently for a wide variety of industrial endeavours (see Habitat Trends).
Of the 724 maps at the 1:50 000 scale covering the province (each map sheet covers on average about 900 km2), only 37 (5%) have evidence of possible or probable breeding – nowhere in the province has breeding been confirmed (Smith 1996). There are 60 known locations of Yellow Rail presence during the breeding season in Saskatchewan (Table 2). A. Smith (pers. comm.) suggests that there could be many more breeding sites in Saskatchewan than the number of known sites and the atlas data would suggest. (There are very few birders in Saskatchewan, and they tend to visit the same sites rather than explore far afield.) According to Smith, considerable amounts of potential habitat for this species exist in the Parklands and the southern Boreal Forest, but it is not known to what extent Yellow Rails use this habitat. Indeed, some new sites harbouring this species have been discovered since the Saskatchewan Atlas (Smith 1996) was published.
Apart from one site (Peter Marsh) where up to 22 birds have been heard, at most sites only one to five birds are generally heard. Numbers at a given site likely vary from year to year according to water conditions. A. Smith (pers. comm.) suggests that there are likely more than 100 breeding sites in the province, and these may harbour several hundred to 1000 birds. As is the case with Alberta, no information is available on population trends per se in Saskatchewan, but the species’ habitat has been eroded considerably for agriculture and continues to be (see Trends in Quality and Quantity of Critical Habitat).
There are 26 known locations of Yellow Rail presence during the breeding season in Manitoba (Table 2). Even though it is considered uncommon (Manitoba Avian Research Committee 1986), there are thought to be more than 100 recent occurrences with at least 3000 individuals (Duncan 1996). R. Koes (pers. comm.), with about 30 years birding experience in the southern third of the province and in the Churchill area, feels that “there are undoubtedly more than 100 breeding sites in Manitoba, and likely hundreds”. He argues that there are vast areas of potential habitat in the Parklands, the southern Boreal Forest and the Hudson Plains that are not often visited by birders, particularly at night. P. Taylor (pers. comm.) states that it is extremely difficult to estimate the Manitoba population or trends, but suggests that the Yellow Rail population is in the thousands rather than hundreds. Occupancy of particular sites often varies from year to year according to water levels, but some sites are occupied every year.
R. Koes (pers. comm.) suggests that the Yellow Rail in Manitoba is probably much more common and widespread than many people suspect. In his view, it is probably not a rare bird in the province. Although there was likely considerably more habitat 100 years ago than now, R. Koes suggests that the conversion of this habitat, mostly for agriculture, has decreased considerably. Yet it must be noted that the species’ habitat has been and still is being eroded considerably (see Trends in Quality and Quantity of Critical Habitat).
P. Taylor (pers. comm.) writes, and some of his comments may apply to the Parklands and Southern boreal forest regions of the three prairie provinces: “It is a little ironic that we probably know the best localities in the extreme south and north, and very little about the main breeding range in central Manitoba. However, that is true of most Manitoba birds…I have visited the [Lee Lake Wildlife Management Area] in daylight, and there are huge expanses of sedge fen that look perfect for this species. Such habitat is widespread in that part of the Interlake region, and would merit more intensive surveys for this species. I suspect that this region, in addition to large marshes bordering Lakes Winnipeg and Manitoba, represents the most important Yellow Rail habitat in Manitoba. Unfortunately it is somewhat remote and extremely under-birded!”
The Yellow Rail likely had a widespread breeding distribution historically in Ontario. Today, the Hudson Bay Lowland is clearly Ontario’s stronghold (Cadman et al., 1987). Abundance estimates from the Atlas (Cadman et al., 1987) indicate that the Yellow Rail is locally common at least at some sites on the Hudson/James Bay coasts: two 100 km x 100 km blocks were estimated to contain 101-1000 pairs, two 2-10 pairs, and one 11-100 pairs. Cliff Hope and two other observers estimated 100 birds on 17 July 1940 at Fort Severn (Speirs 1985). Of the 12 Atlas blocks of 100 km x 100 km covering Ontario’s north coast, all have had observations of Yellow Rails. During the atlas itself, only 11 did, but in the remaining one (in which Cape Henrietta Maria lies) breeding was confirmed before field work was conducted for the Atlas (Speirs 1985) (Appendix 2).
In the remainder of the Ontario range, breeding is apparently scattered and numbers of Yellow Rails are thought to be small. In the Rainy River region, their last stronghold, Big Marsh, at the mouth of the Rainy River in Lake of the Woods, is protected by the provincial government. This 6 km x 1 km area has much sedge meadow. The observations made every year of 1-6 individuals are made in a small accessible portion representing only about 5% of the Big Marsh (D. Elder, pers. comm.). The Ontario Rare Breeding Bird Program estimated that more than 100 pairs summered in the Rainy River region annually between 1981 and 1990 (Austen et al., 1994), and that number still seems a reasonable reflection of the current situation (D. Elder, pers. comm.). There have also been occasional reports from Quetico Provincial Park (D. Elder, pers. comm.).
In southern Ontario, which has lost a large proportion of its wetlands, the Yellow Rail population has been almost extirpated (Austen et al., 1994). In the early-1900s, it was found regularly, albeit in small numbers, in large marshes of southern Ontario from the St. Clair River east to Toronto (Austen et al., 1994). In recent years, however, it has been recorded only sporadically in summer at a few widely scattered marshes in southern Ontario (it was found in only six of 1824 Atlas squares (Cadman et al., 1987)), only two of which have recent breeding evidence – Richmond Fen, in the Ottawa region, and Holland Marsh, along the south shore of Lake Simcoe (Sankey 1987, Goodwin 1992, Page and Cadman 1994). Up to 20 males have been heard at Richmond Fen, and a nest with eggs was found in 1982 (Austen et al., 1994). The species was found in the Holland Marsh area during the Atlas in the 1980s, but not since 1989, despite the fact that there do not appear to have been any significant changes to the habitat during that period (Austen et al., 1994, R. Ridout, T. Hofmann, pers. comm.). It may still breed in the less accessible east portion of the Holland Marsh, which is about 6 km long by 1.5 km wide (T. Hoffman, G. Bennett, pers. comm.). From 12 to 30 pairs might breed annually in southern Ontario (Page and Cadman 1994).
The Yellow Rail is considered a rare summer resident in Québec (David 1996). Even though no surveys have been carried out specifically for this species in the coastal marshes of James Bay, this region probably harbours as many, if not more, Yellow Rails than the rest of Québec. Todd (1963) noted that they were “plentiful” in July 1941 at Neck of Land, just north of Boatswain Bay. Numerous large bays, notably Rupert and Boatswain Bays, contain vast suitable high marshes (M. Robert, pers. obs.). The presence of Yellow Rails in a number of those coastal marshes suggests that dozens (or even hundreds) of pairs breed in that vast region (Robert et al., 1995), despite the fact that nesting has never been confirmed there. The number breeding there may have changed little, because the region has been subjected to little disturbance (Robert et al., 1995). Nevertheless, two potential threats are hydro-electric development and overgrazing by Snow Geese, which has been responsible for major habitat changes on the west coast of Hudson Bay (see Trend in Quality and Quantity of Critical Habitat for details).
The situation is much different elsewhere in Québec. There are very few Yellow Rails inhabiting the St. Lawrence and Saguenay corridor. Up to 9 have been heard in the Lake Saint-François National Wildlife Area and its surrounding areas (Robert and Laporte 1996), 3 at Île du Moine (ÉPOQ database), 5 at Cap Tourmente (ÉPOQ database), 20 at Île aux Grues (Robert and Laporte 1996), 2 at Sainte-Anne-de-la-Pocatière (Campagna 1931, Meredith 1935), 9 at Cacouna (Robert and Laporte 1996), 3 at Pointe-aux-Outardes (Robert and Laporte 1996), 4 at the rivermouths of the Gaspé Peninsula (Robert and Laporte 1996), 5 at Coin-du-Banc (Terrill, unpublished data, Canadian Museum of Nature), 8 at Saint-Fulgence (Cormier and Savard 1991, Savard and Cormier 1995), and 2 at Saint-Gédéon (Robert and Laporte 1996). The remaining reports of Yellow Rails along the St. Lawrence corridor involve only one individual. In the Abitibi region, up to 3 were heard at Marais Antoine in June and July 1995 (Robertet al., 1995), and 7 were heard at Marais Maine in June 1996 (van de Walle 1997). It should be noted that the numbers given above are historical maxima and that the number of individuals heard at these sites is usually smaller. At most of these sites the species’ occurrence is irregular or uncommon (see below).
The irregular nature of this species’ occupation of sites seems to be generalized in southern Québec. At Saint-Fulgence (Saguenay River), it was first found in 1964 and was reported irregularly until 1996 (Browne 1967, Cormier and Savard 1991, Robert and Laporte 1996); it does not seem to have been present since then (G. Savard, pers. comm.). Similarly at Cap Tourmente it was first reported in the early 1970s and has been observed only irregularly during the past decades (Otis et al., 1993, Robert and Laporte 1996); contrary to Saint-Fulgence, however, it has been heard at Cap Tourmente during the past few years (S. Labonté, pers. comm.). Its presence seems to be irregular also at Marais Antoine and Marais Maine in the Abitibi region, where several individuals were heard in 1996 (see above) but where a thorough survey the following year yielded none (M. Robert, pers. obs.). At Marais Antoine, it is possible that the desertion was associated with water level changes for a waterfowl management project (water level management for waterfowl at Marais Antoine was initiated in autumn 1996).
The number of Yellow Rails breeding in southern Québec is likely much lower than it used to be. Numerous marshes along the shores of the St. Lawrence River disappeared this century, and since the Yellow Rail uses the higher portions of marshes, which are often the first and easiest to be drained, the population must have been affected by the loss of habitat to humans (see Habitat). For example, the number of birds breeding at sites along the river at Sorel or Yamachiche or at sites along the estuary (e.g. Château-Richer, Sainte-Anne-de-Beaupré, La Pocatière, Kamouraska) must have been considerably larger than today (see Campagna 1931, Meredith 1935).
Despite the Yellow Rail’s seeming irregular occupation of sites in southern Québec, there are sites that are known to have been traditionally used over the past few decades. At Coin-du-Banc, for example, Terrill (unpublished data) observed Yellow Rails regularly from 1939 to 1949, and the species was still present in 1997 (P. Poulin, pers. comm.). In the area of the Lake Saint-François National Wildlife Area, the species started to be reported at the end of the 1960s and it still occurs there (Bannon 1992, 1993, Robert and Laporte 1996, M. Robert, pers. obs.).
Most of the marshes along the St. Lawrence River and the Saguenay River with habitat suitable for Yellow Rails were surveyed in the early 1990s (Robert and Laporte 1996), and the results showed the extent to which the species is uncommon and locally distributed there. Yellow Rail habitat is rare along the St. Lawrence River and its tributaries. At present, the largest areas are at Lake Saint-François (in Québec’s extreme south-west) and on the south shore of the St. Lawrence River (east of Québec City), especially at Île aux Grues. About 130, 530 and 30 ha of suitable habitat were found at Lac St-Saint-François, Île aux Grues and Cacouna, respectively, over the last few years. No more than 40 males were found at the three sites combined each year, although sometimes the number was much lower (Robert and Laporte 1996). Given these results, it is probable that the main localities in southern Québec are all known, but Yellow Rails could potentially also use other sites.
In summary, the Yellow Rail inhabits a specific type of marsh that is rare in southern Québec. Most of the suitable sites are known and they harbour few rails. Furthermore, the few large fens that have been visited in Québec’s interior do not seem to harbour the species (M. Robert, pers. obs.), suggesting that most of the individuals inhabiting southern Québec frequent the St. Lawrence corridor and a few of the large rivers feeding it. There is little room for Yellow Rails to expand in the St. Lawrence and Saguenay corridors (Robert et al., 1995). At most a few new sites will be found in the future, and these will likely harbour few individuals. We suggest that 20-80 pairs currently breed in southern Québec, and that about the same number breed in Québec’s interior (Robert et al., 1995).
Erskine (1992) suggested that fewer than 50 pairs breed in New Brunswick. In fact, the only site where the Yellow Rail is known to occur regularly in New Brunswick is Grand Lake Meadows, in the upper estuary of the St. John River, where three to 24 calling males have been heard annually from 1991 to 1996. During that period, the area with Yellow Rails varied from 35 to 131 ha (P. Kehoe and G. Forbes, pers. comm.). Even though the habitat seems very good for Yellow Rails, breeding has never been confirmed there. Water levels in the St. John River may be too high for the species during the breeding season and it is possible that the birds observed at Grand Lake Meadows have been molting birds that bred elsewhere, such as in Québec or in the interior of New Brunswick. Efforts are currently underway to determine whether the species breeds at Grand Lake Meadows (P. Kehoe, pers. comm.).
After 11 years of aerial and ground wetland surveys throughout New Brunswick, P. Kehoe (pers. comm.) found that of about 20 000 ha of wetland in the St. John River Valley downstream from Fredericton (including tributaries of the river), there are about 4 000 ha of potential habitat for Yellow Rails, i.e. sedge meadows. According to the latest figures from the provincial government, 11% of the 20 000 ha (2 300 ha) of the Lower Saint John River (Fredericton south to Saint John) have been impounded. Only a portion of this would have been sedge meadow, and it is not known whether these areas had Yellow Rails, as the species has been heard only at Grand Lake Meadows (G. Forbes, pers. comm.). In the province’s interior there are about 30 sites with sedge meadows that could have breeding Yellow Rails (P. Kehoe, pers. comm.). Night surveys for Yellow Rails at these interior sites are needed.
Several Yellow Rails were heard in recent years at the Tantramar marshes near Sackville. This was the area of a “colony” in the 1940s-1970s (G. Forbes, pers. comm.).
Erskine (1992) suggested that fewer than 50 pairs breed in Nova Scotia.
- Date Modified: