Hill’s thistle (Cirsium hillii) COSEWIC assessment and status report: chapter 7

Population Sizes and Trends

For this report, the term population is used to refer to the basic unit, and is defined as, “A group of individuals that reproduce with one another and produce offspring” (Primack 1993). This term, when applied to Cirsium hillii, does not imply any genetic isolation between populations, as the species likely experiences few barriers to genetic exchange between proximal populations along the west shore of the Bruce peninsula and the south shore of Manitoulin Island, other than of course, the fragmentation of its prime habitat, i.e., it is pollinated by long-tongued bees and its seeds are wind dispersed. The term station, site, or population, is thus used interchangeably to refer to one or more populations of C. hillii that are separated from other populations by at least one km of unoccupied or unsuitable habitat. This is the definition used by the Natural Heritage Information Centre (NHIC) for an Element Occurrence (EO), thereby facilitating comparisons with, and additions to, that dataset. All populations were plotted on topographical maps in NAD 27, using the information collected by the author, e.g., pers. comm. with Morton & Johnson, and referencing these against the NHIC Element Occurrence Summaries for each population. The criteria regarding ‘lumping’ or ‘splitting’ to define a population was then applied, with those records within one km proximity treated as sub-populations. In some cases populations beyond the one km cutoff were included, because of known imprecision with the location of that particular record.

A total of 70 sites have been recorded for C. hillii; six of these are now believed to be extirpated. Of the 64 extant sites, 36 have been recorded with quantitative data (Table 1), nearly all of these population counts are from 1995 to 2003, with only three prior to this, all in the 1980s, thus rendering long-term trend analysis impossible. Of the 36 sites with good population data, only 11 of these support populations with greater than 100 plants (Sites 4, 14, 15, 16, 19, 32, 35, 44, 54, 57, and 60). The most populous documented station is Site 57, with 1175 plants, including 35 flowering, tallied by Johnson in 2003. However, it was estimated that this site could have as many as 1500 plants. Site 9 is perhaps the most abundant station for C. hillii on Manitoulin Island (Morton pers. comm. 2002), with extensive prime habitat that has not been fully quantified as yet for C. hillii. The population counts must be used with the caveat that they are often incomplete counts, as several were recorded during surveys that were not specific to Cirsium hillii, and the fact that the species can be quite difficult to quantify in its typical habitat of openings within the contiguous forested landscape. These qualifiers, where recorded by the observers, are noted in Table 1.

Table 1. Cirsium hillii Population Information Summary for OntarioStations.
Site Number of plants Observer(s)
Number 3 57 (15 flowering, 42 rosettes) Jones, 1995
Number 4 107 (mostly rosettes) (“If entire habitat were more thoroughly searched, the #s would be much higher.”) Jones, 1996
Number 6 12 (no breakdown)
39 (1 flowering, 38 rosettes)
several rosettes
Jones, 1995
Jones, 2000
Morton & Venn, 2002
Number 9 13 (all rosettes)
about 2 dozen (no breakdown)
Morton (2002) considers this to be perhaps the most abundant station for C. hillii on Manitoulin Island
Jones, 2000
Allen et al. 2002

Number 11 ca 20 plants (2 in flower) Oldham, 1994
Number 12 “rare” Oldham, 1994
Number 13 “rare” Oldham, 1994
Number 14 104 (only a few seed heads, mostly rosettes)
46 (all rosettes)
Jones, 1995
Jones, 2000
Number 15 180 (all rosettes)
347 (all rosettes)
(“The 347 is representative of a few places where rosettes were counted along trail. Actual total number is much higher for whole area”).
Jones, 2000
Jones, 2000
Number 16 121 (1 seed head, 120 rosettes)
14 (all rosettes)
Jones, 2000
Jones, 2000
Number 17 70 (1 seed head, 69 rosettes) Jones, 1995
Number 18 ca. 20 plants (no breakdown) Oldham, 1989
Number 19 187 (4 flowering, 183 rosettes) Jones, 1996
Number 22 ca. 30 (no breakdown)
31 (8 fruiting heads, 23 rosettes)
Oldham, 1994
Jones, 2000
Number 23 17 (3 flowering, 14 rosettes + many small seedlings) Jones, 1996
Number 24 13 (1 in bud, 12 rosettes) Jones, 2000
Number 25 several rosettes (“but plant is probably more widespread”) Morton & Venn, 1981
Number 32 152 (2 flowering, 150 rosettes) Jones, 1996
Number 33 10 (5 flowering, 5 rosettes) Allen & Allen, 2002
Number 35 157 (113 flowering, 44 rosettes) Allen & Allen, 2002
Number 37 “rare” Morton & Venn, 1973
Number 39 “scarce & rarely flowers” Morton & Venn, 1987
Number 41 60 (5 in seed, 55 rosettes) Schaefer, 1995
Number 42 “rare” Schaefer, 1995
Number 44 30 (all rosettes)
250 (1 flowering, 249 rosettes)
Johnson, 1984
Allen, 2002
Number 45 1 (rosette) Schaefer, 1996
Number 46 39 (14 flowering, 25 rosettes) Allen, 2002
Number 47 2 (both flowering) Oldham, 1995
Number 48 3 (2 in seed, 1 rosette) Schaefer, 1995
Number 50 4 (all rosettes) Johnson, 2002
Number 51 30 (2 flowering, 28 rosettes) Johnson, 2002
Number 52 75 (10 flowering, 65 rosettes) (a number of additional sub-populations are actually present) Johnson, 2001
Number 53Footnotea 57 (1 fruiting, 56 veg in many separate sub-populations) Johnson, 2003
Number 54Footnotea 183 (8 flowering/fruiting, 175 vegetative) in 5 sub-populations Johnson, 2003
Number 55 31 (minimum) (3 flowering, 28 rosettes)
14 (6 flowering, 8 rosettes)
Schaefer, 1995
Allen & Allen, 2002
Number 57Footnotea 19 (2 flowering, 17 rosettes)
1175 counted but 1300-1500 est. with 35 having inflorescences
Johnson, 1985Johnson, 2003
Number 59 “Rare and local” Oldham, 1997
Number 60 443 (7 flowering, 436 rosettes) Wasaga Park, 2001
Johnston & Bieniek, 2003
Number 61Footnotea
New locality
EO ID:1768
23 all vegetative Johnson, 2003
Number 62Footnotea
New locality
EO ID: 5123
43 all vegetative Johnson and Macdonald, 2003
Number 63Footnotea
New locality
EO ID:?
9 (1 flowering/fruiting, 8 vegetative) Johnson, 2003
Number 64Footnotea
New Locality
EO ID:1777
34 (3 fruiting, 31 vegetative) in two sub-populations Johnson, 2003

For Cirsium hillii it is very difficult to detect trends through time with individual sites, due to the relatively few records with quantitative data, and the uncertainty in knowing whether a population or sub-population surveyed was in fact the same as one surveyed a few years previous. This is especially the case with a species such as C. hillii, which often occurs in small openings, imbedded in rather extensive forests, and it becomes a challenge to record the precise opening or openings, and the problem is compounded by the ephemeral nature of these openings as they grow in and openings are created elsewhere by disturbance. The recent advent of GPS technology will greatly assist this situation, with the majority of recent surveys on the species possessing very precise locations.

A total of 36 of the 64 extant stations have been surveyed quantitatively, with nearly all of these counts made in the last eight years. The largest population at Site 57 consists of 1175 individuals, including only 35 mature flowering/fruiting plants in 2003. The total number of plants counted for the 36 sites surveyed is about 3915, with 253 of these being mature flowering plants, 3,662 rosettes, and 86 with no distinction made between reproducing plants and basal rosettes (There may be some overlap in the ‘counting’ of sub-populations, but for the most part the quantifications of populations represented separate sub-populations). It is quite striking that of the 253 flowering plants recorded, 45% of these (113 plants) are at one station, Site 35, which is licenced for aggregate extraction, with the site being prepared at the time of field surveys in 2002. The largest populations are likely known, however some significant populations may still be undetected at the other 28 stations. The Canadian population likely consists of at least 5,000, or perhaps somewhat more, plants (flowering plus vegetative) when taking into consideration the total estimated number at Site 57 and additional undetected sites. The total number of plants in Canada capable of reproducing in any year is estimated at 10% of the total; the total mature, reproducing individuals may be in the order of 500+ plants.

From his rangewide status assessment, Penskar (2001) concluded that “Although Cirsium hillii is an endemic Midwestern taxon far from being demonstrably secure, it is not sufficiently rare nor endangered throughout its range to merit consideration for federal listing at this time. Despite its extreme rarity in Iowa, Indiana, and to some extent Illinois, there are numerous records of viable populations within the remainder of its range, including Ontario, Michigan, Wisconsin, and Minnesota. Within states such as Michigan, the stronghold for Hill’s thistle rangewide, this species persists well even though as a special concern or “watch-list” plant it has no legal protection, and can be expected to persist well into the future. Moreover, there are likely additional large to moderately large populations that will be discovered through ongoing inventory in Michigan and elsewhere. Overall, the current G3 rank assigned by The Nature Conservancy is well justified, based on the current data.” However, Penskar adds a cautionary note, stating that, “Even though there is insufficient merit for recommending federal listing, Hill’s thistle remains vulnerable to continued decline in many portions of its range without determining the most appropriate management strategies. It is clear that the biology and ecology of this species is rather poorly known. Without intervention through active management and a better understanding of the natural history of this species and its habitat, it is very possible that Hill’s thistle could significantly decline over the next several decades.”

A downward trend in Ontario populations appears inevitable, for the reasons cited in the section on Limiting Factors and Threats. Locally, however, in some areas, as along sections of the south shore of Manitoulin Island (Jones pers. comm. 2003), and on the Bruce Peninsula, from Dorcas Bay to Little Pine Tree Point, it could be considered common (Owen Sound Field Naturalists, 2001). As with the US populations, new populations will continue to be discovered and ‘new’ sub-populations will be added to the existing stations. However, a cautionary note is also warranted for the 60 extant stations. To control, let alone reverse this trend, with respect to development pressures in the shoreline alvars, lack of fire disturbance in the species’ habitat, and successional closure of open habitats supporting Cirsium hillii, will pose a considerable challenge.

Extirpated populations

Hopkins Bay, St. Edmunds Township, Bruce County. First recorded in 1949 by H.A. Senn et al. (coll at DAO) from "Hopkins Bay, 3 ½ kms south of Tobermory, upper beach, wooded". No observations since and believed extirpated. The site is private propoerty.

Pine Tree Point, St. Edmunds Township, Bruce County. Probably first observed in 1934 by Krotkov (coll at TRT) from "open woods" and not recorded since a 1948 collection (DAO) by C. Frankton et al. from "burned over limestone barrens". Site investigated 8 August 2002 and very end of Pine Tree Harbour Road is posted “No Trespassing” and fenced (owner Bauman). Shoreline of White Cedar on alvar is all carved up into new lots for sale. There is complete exclusion of public access down Pine Tree Harbour Road. Also investigated the alvar on the north side of Pine Tree Harbour Road. This appeared to be good habitat, and was quite extensive, with quite open white cedar, tamarack, white birch, with Schizachrium scoparium dominant in the herbaceous layer, and Danthonia spicata, Juniperus horizontalis and Potentilla fruticosa common, but no C. hillii could be found. There is good alvar on the south side of the road, from Concession 6-7 west, but this was posted “No Trespassing” so was not accessed. “For sale” signs were erected on individual lots.

Fishing Islands, Amabel Township, Bruce County. John Macoun collected C. hillii on these islands in 1874 "In dry thickets", with no mention as to the specific island (coll at CAN). There have been no other confirmed records for the Fishing Islands since Macoun’s.

Sauble Beach South, Saugeen First Nation Reserve, Amabel Township, Bruce County. First observed in 1950 from “South Sauble Beach” by J.H. Soper & Shields (coll at CAN, TRT). Last observed in the 1980s by J.W. Johnson, with no abundance or habitat information. No observations since and believed to be “probably extirpated” (Johnson pers. comm. 2002).

Walpole Island First Nation, Lambton County. Last noted in 1914 by Dodge (1914) as, "occasional in prairie-like ground on Squirrel Island. Apparently rare." Dodge does not seem to have made a voucher, as no collection exists at MICH. Not observed by Gaiser (1966) or during the Natural Areas Inventory (Woodliffe & Allen 1988). If C. hillii ever existed on Squirrel Island, it certainly does not currently, as little remnant savannah exists on Squirrel Island, and this was covered intensively during the inventory work in the 1980s.

Wasaga Beach, Nottawasaga Bay Shore, Simcoe County. This station was known in the 1970s by A.A. Reznicek (pers. comm. in Brunton 1989). It is believed to have been destroyed by recreational development which is quite prevalent inthe area.

Historical populations of unknown status

Cove Island, St. Edmunds Township (location is approximated from Morton’s description). Morton & Venn (1987) cite a record for Cove from Cuddy and Norman (1972) as “needing confirmation”. Morton (pers. comm 2003) notes that, “[The record] could be correct because there is a suitable bit of alvar behind one of the exposed bits of shore on the west side of Cove about half way up the coast.” Not recorded since 1972 (Promaine pers. comm. 2003).

Black Creek Provincial Park, Estnor Township, Bruce County. First observation 1 July 1982 by I.D. Macdonald as “scattered in various locations and observed in several communities, from Stokes Bay to Myles Bay.” (Macdonald 1982). Also observed in the 1980s by J.W. Johnson, with no abundance information provided. Searched unsuccessfully for C. hillii 8 August 2002 by G.M. Allen, including Juniperus communis-Pteridium aquilinum openings; back dunes of Schizachyrium scoparium and Danthonia spicata in white spruce; and the frontal dunes backing the parking lots with very narrow dunes backing onto closed mixed forest. However the areas to the west toward Shute Point, and north toward Stokes Bay were not surveyed, and C. hillii could certainly be present here. Publicly owned as a Provincial Park by the Ministry of Natural Resources.

Erroneous reports

Flowerpot Island, Bruce County, is erroneously reported as an Element Occurrence in the NHIC database, with the source cited as Morton & Venn (1987). There is no such reference in The Flora of the Tobermory Islands, and in personal communication with Dr. Morton (2003), when queried on the record, he stated that “The only location for Hill’s Thistle on the Tobermory Islands that I know of is on the alvar on Bear’s Rump Island. There is no suitable habitat for it on Flowerpot.” 

Cape Crocker (Nawash) First Nation, Bruce County. NHIC notes that, “Varga’s (1995) report references the NHIC database as the source of the pre-1949 observation, however no record of this observation can be located.” Not found by Varga in 1992 and 1995 field surveys (Varga et al. 1995).

Johnson’s Harbour, Grey County. Although a separate EO in the NHIC database, the “Comments” in the NHIC EO summary sheet for “Johnston’s Harbour, Bruce County” note that, “A 1933 collection record by Krotkov labelled “Johnson’s Harbour” has been included in this EO – the ARVPO database listed it as ‘Johnson Harbour’ northeast of Owen Sound. Johnson’s (1982) report may contain a recent record of C. hillii at this location, but this report is presently unavailable at the NHIC.” Joe Johnson has no knowledge of this record of Krotkov’s from “Johnson’s Harbour” in Grey County, nor is he aware of any other records from Grey County (pers. comm. Johnson 2003). This record is thus considered to be Krotkov’s record from “Johnston’s Harbour” in Bruce County.

Fishing Islands, Wildman's Island, Amabel Township, Bruce County. The species has been reported from the Fishing Islands, in Bruce County (Parker et al. 1985), specifically from Wildman’s Island. The record of C. hillii from this study is believed to be erroneous, as there was apparently no collection made, and no specimen was ever verified by Joe Johnson, who was acting in that capacity for the project (Johnson, pers. comm. 2003).

Cirsium hillii was cited for Manitoba by Fernald in Gray’s Manual of Botany (1950). Scoggan, in his Flora of Manitoba (1957) pointed out that the related species C. drummondii was the basis of reports of C. hillii in Manitoba. Subsequently, Moore and Frankton (1966) concurred, stating that, “We question the reported occurrence [of C. hillii] in Manitoba, and believe that these refer to the related species C. drummondii.” In the intervening years no records of C. hillii have been confirmed from Manitoba.

Potential sites for investigation

The following sites are offered as having good potential for supporting Cirsium hillii (all UTMs are NAD 27). I have provided the originator of the suggested site in brackets.

  • Great La Cloche Island – There is extensive alvar habitat as yet unexplored for C. hillii through the southern portions of the island (Dr. J.K. Morton and Judith Jones).
  • Little La Cloche Island – I was able to investigate the alvars at either end of the island during the 2002 fieldwork, but was not able to visit the south end at Mary Point, also suggested by Dr. Morton.
  • Providence Bay, Manitoulin Island - Habitat to the west and east of town at UTMs 399400 5057300; 399800 5058000; 301700 5057700; and 302700 5057600 affords good potential for C. hillii, and would help bring some precision to this record Dr. J.K. Morton).
  • Shrigley Bay, Manitoulin Island – No records of C. hillii, despite good habitat supported (Judith Jones).
  • Barrie Island – Good alvar at extreme west end of the island between Salmon Bay and Sturgeon Bay was not checked during fieldwork (Dr. J.K. Morton).
  • Wikwemikong First Nation Reserve - South end supports good habitat that should be checked for C. hillii (Judith Jones).
  • Wikwemikong First Nation Reserve– Possible good alvar surrounding the bay on the east side of James Bay (Dr. J.K. Morton).

Page details

Date modified: