The earliest botanical collections on Victoria Island were gathered during a four-year (1850–1854), two-ship (Enterprise and Investigator), British expedition in search of Sir John Franklin’s missing ships Erebus and Terror, and by John Rae in 1851. These collections are housed at the Royal Botanic Gardens Kew (K; herbarium acronyms follow Thiers (continuously updated)) and are summarized in Simmons’ (1913) account of the early history of botanical exploration of the Canadian Arctic islands. Travelling aboard Investigator, commanded by Robert McClure, Moravian missionary and interpreter Johan A. Miertsching made five collections along Prince Albert Sound (Draba hirta L. [=D. glabella], Saxifraga oppositifolia, Potentilla rubricaulis, Astragalus aboriginorum Richardson [=A. richardsonii], Arctostaphylos alpina [=Arctous alpina]), and a collection of Salix arctica, taken at “Mount Adventure” [Adventure Mountain] on Prince Albert Peninsula, is attributed to McClure (Simmons 1913). Enterprise, commanded by Richard Collinson, was separated from Investigator in the Pacific, but eventually reached the western Arctic. The surgeon, Charles Anderson, who was among the crew on Enterprise, made botanical collections at Minto Inlet and Cambridge Bay, representing 94 species according to Simmons (1913) (Suppl. material 1). A collection of Oxyria digyna from Walker Bay, taken by Collinson, is also recorded by Simmons (1913).

John Rae made collections along the “south shore” of Victoria Island in 1851 while exploring and mapping the area (Rae and McGoogan 2012). He travelled to the island from the mainland twice that year. On 5 May, Rae crossed, by sledge, the frozen Dolphin and Union Straight and reached Douglas Island (68°28'N, 113°29'W), just west of Wollaston Land, the next day. Rae then travelled eastwards along the southern shore of Victoria Island as far as Wilbank Bay (68°38'N, 110°10'W). Upon his return to Douglas Island, on 15 May, he travelled northwards along the west coast of Wollaston Land to ca. 14 miles beyond Cape Baring (70°02'17"N, 117°21'33"W), before turning around. He returned to the mainland on 2 June. On the second leg of the 1851 expedition, Rae travelled by boat. He left the mouth of the Coppermine River in early July and travelled eastwards along the south shore of Coronation Gulf to the east end of Kent Peninsula. On 27 July he crossed the strait to Victoria Island, reaching Cambridge Bay. Leaving the bay, Rae followed the coast eastwards to Albert Edward Bay, and then travelled north overland on foot to a point ca. seven miles south of Pelly Point (70°12'N, 101°01'W) on the Collinson Peninsula, before turning around. On the return trip Rae apparently followed the southern coast of Victoria Island, as he describes stopping north-east of Cape Peel (69°02'30"N, 107°16'15"W) along Dease Strait and Point Ross [Ross Point; 68°31'N, 111°10'0"W) at Johansen Bay, which was reached on 28 August 1851. We are not aware of information as to precise localities of Rae’s botanical collections; there is no mention of them in his account of his travels on Victoria Island (Rae and McGoogan 2012). Given that his travels on the southwestern portion of the island occurred in May, when there would have been snow on the ground and conditions would not have been suitable for collecting plants, we assume that his collections were gathered on the “south shore” of the southeastern part of the island during July and August. Hooker (1856) listed Rae’s 1851 collections from Victoria Island and from the adjacent mainland, but he did not indicate in the list which specimens were from the island. Simmons (1913), having reviewed the collections at K, reported that Rae made 34 collections on Victoria Island (Suppl. material 1). These collections are also summarized in Porsild (1955).

A substantial amount of information on the flora of Victoria Island accumulated in the 20^th century. The herbaria in which the collections described below are housed are listed in the Annotated Checklist, unless otherwise indicated. Diamond Jenness, anthropologist, made collections on “Wollaston Land” (Wollaston Peninsula) in 1915, during the Canadian Arctic Expedition, 1913–1918. These were reported in Macoun and Holm (1921). A list of 41 plant collections made by Jenness on the island was also included in Jenness (1991).

Considerable plant exploration of Victoria Island occurred in the 1940s, when many plant collections were made at fur trading posts on the eastern shore of Walker Bay (Fort Collinson), at Holman and Cambridge Bay, and on Read Island just off the south coast of the Wollaston Peninsula. Father Arthème Dutilly, associated with the Catholic University of America, collected in 1940 at Holman, at a site referred to on his collection labels as “Willows Patch” [Boot Inlet; see comments under Salix alaxensis about the location of this site], at Kookyoak River [=Kuujjua R.] and at Cambridge Bay (Louis-Marie 1961). Dutilly’s collections are distributed widely in herbaria in Canada, the United States and Europe (Boivin 1983). Although collected over 70 years ago, most of Dutilly’s Victoria Island collections were not included in the distribution maps in Arctic floristic treatments that include the island (Porsild 1955, 1957, 1964, Porsild and Cody 1980, Aiken et al. 2007). The geologist Albert L. Washburn made collections at Cambridge Bay in 1940. In 1941 and 1942, Lillian Ross, spouse of trading post manager Ray Ross, collected at Holman Island post, Read Island post and sites on southern Victoria Island adjacent to Read Island. In 1947, the geologist Y.O. Fortier made collections at Greely Haven (1 August) and Cambridge Bay (3 August). Also in 1947, Nicholas Polunin, Oxford University, made collections at Cambridge Bay (24 August); these collections have not previously been accessible for research, being part of the unprocessed backlog of the National Herbarium of Canada (CAN). In 1948, John L. (Pete) Jenness, son of Diamond Jenness, made a few collections in the vicinity of Richard Collinson Inlet and on the Storkerson Peninsula.

Botanist A. Erling Porsild, with the National Museum of Canada, collected on Victoria Island in 1949 at several sites, spending no more than two days at each: Read Island (27 July), the interior of the Wollaston Peninsula (27 July), the head of Prince Albert Sound (4 August), Holman Island (Ulukhaktok; 8 August), “Jackpot Lake” (an unnamed lake east of the head of Minto Inlet; 16 August), Walker Bay (at Fort Collinson; 25 August), the head of Minto Inlet (2–3 August), Cambridge Bay (5 August), an unnamed lake ca. 60 miles north of Cambridge Bay (6 August), Tahoe Lake and Washburn Lake (4 August), and a few other places inland. Those collections reportedly doubled the number of species known from the island, bringing the then-known flora to 201 species (Porsild 1950a). The collections were reported in Porsild (1955), a monograph on vascular plants of the Western Canadian Arctic Archipelago, including Axel Heiberg, Amund Ringnes, Ellef Ringnes, Borden, Mackenzie King, Cornwallis, Bathurst, Prince Patrick, Melville, Banks, Victoria, Prince of Wales, and King William islands, and numerous smaller islands. Porsild returned to Cambridge Bay in 1959 and again collected there (12 August).

Margaret Oldenburg, an amateur botanist from Grand Marais, Minnesota, travelled and botanized extensively across Victoria Island in the 1940s and 1950s. She travelled primarily by chartered aircraft, typically collecting during brief stops. In 1943, she made collections at Holman Island (=Ulukhaktok; 1 August) and Read Island (23–24, 31 August, 2–3 September). In 1944, she collected at Cambridge Bay (14–15 August) and again at Read Island (20 August). In 1945, she collected at Ulukhaktok (12 August), Walker Bay (12 August) and “Oldenburg Lake” (11 August), an unnamed lake on northwestern Victoria Island west of Peel Point. We determined the location of “Oldenburg Lake” (73.11750, -114.58) based on information in Whyard (1984). In 1946, she collected on the north side of Prince Albert Sound (19 August). In 1954, she collected at “Char Lake” (15 August), which is a small lake southwest of the head of Richard Collinson Inlet on the Prince Albert Peninsula, “Yellow Lake” (8–9 August), which is an inland site east of Deans Dundas Bay, and on a pingo on northwestern Wollaston Peninsula (16 August). We determined the locations of “Char Lake”, “Yellow Lake” and the pingo based on maps on Oldenburg’s specimen labels, on which the sites are mapped as “Sta. 20” [station 20], “Sta. 10” and “Sta. 21”, respectively. Presence of pingos in the area indicated by Oldenburg is confirmed by Geological Survey of Canada (2017). Despite the importance of Oldenburg’s collections, which are extensive and, in many cases, represent the first (and in some cases only) collections from sites on the island, they were not considered in earlier Arctic floristic works. They were, however, known to Porsild (1955) as being unnamed and inaccessible in the Herbarium of the University of Minnesota (MIN). A subset of Oldenburg’s collections from Victoria Island are housed at GH, MIN and UBC, but the majority were recently “discovered”, along with her collections from other Arctic and non-Arctic sites, in the unprocessed backlog cabinets of the National Herbarium of Canada.

Botanical documentation of the Victoria Island flora continued through the 1950s. Entomologists E. Smith and D.K. Sweatman, Department of Agriculture, Ottawa, collected in 1950 at Holman Island (Ulukhaktok). In 1952, entomologists D.P. Gray and B. Gibbard, also with the Dept. of Agriculture, made collections at Holman Island (Ulukhaktok). In 1959 and 1960, W.D. Stretton, Army Survey Establishment, made collections at Cambridge Bay and at numerous sites in interior Victoria Island, during defence mapping operations. Areas where Stretton collected include the north side and head of Prince Albert Sound, Berkeley Point, Burns Lake, Freshwater Bay, Gordon Point, Kuujjua River, Murray Point, Richard Collinson Inlet, Storkerson Peninsula, Ulukhaktok and the west end of the Diamond Jenness Peninsula. Botanists J.A. Calder, D.B.O. Savile and I. Kukkonen collected at Cambridge Bay on 12 August 1959.

Collections in the 1960s were focused on southeastern Victoria Island and in the 1970s in the Ulukhaktok area. Homer A. “Steve” Stephens (and his associate George M. Sutton) made extensive collections in 1962 in the vicinities of Cambridge Bay and Mount Pelly, in association with research on the White-rumped Sandpiper at Cambridge Bay. Stephens also made collections on Jenny Lind Island immediately southeast of Victoria Island in Queen Maud Gulf (Sutton 1967, Parmelee et al. 1968), which are not considered here. In the same year, Robert Hainault made collections at Mt. Pelly, Mt. Lady Pelly and Ferguson Lake north of Cambridge Bay. Ecologist Josef Svoboda, University of Toronto, made collections at Holman (Ulukhaktok) in 1974. Ecologist Lawrence C. Bliss, University of Alberta, collected there that same year.

In the 1980s, Sylvia A. Edlund, Geological Survey of Canada, made substantial contributions to knowledge of the flora of Victoria Island, associated with research characterizing the relationships between surficial geology and plant communities. She made more than 1800 collections at numerous sites across the island, including many from sites where no other collections have been made. Sites visited and collected by Edlund in 1982 include Armstrong Point, Boot Inlet, Cape Baring, Cape Wollaston, Kugaluk River, Kuujjua River, the head of Minto Inlet, Mt. Bumpus, Murray Point, Natkusiak Peninsula, Peel Point, the north side and head of Prince Albert Sound, Richard Collinson Inlet, Sangraun Hills, Shaler Mountains near Richard Collinson Inlet, Tahiryuaq (a large lake north of the head of Prince Albert Sound), Ulukhaktok and southwestern Wollaston Peninsula (Edlund 1983). In 1983, Edlund collected at Cambridge Bay and the head of Minto Inlet, and in 1986 at Hadley Bay and Storkerson Peninsula on northeastern Victoria Island. Edlund collected in 1987 with salicologist George Argus (National Museum of Canada) at Albert Edward Bay, Anderson Bay, Byron Bay, Cambridge Bay and vicinity (including “Starvation Cove”), Cape Colborne, Collinson Peninsula, Diamond Jenness Peninsula, Ferguson Lake, Greiner Lake, Jonnessee Lake, Namaycush Lake, the head of Prince Albert Sound, Surrey Lake and Washburn Lake. Other sites visited in 1987 (without Argus listed as a collector) include the vicinities of Burns Lake, Cape Peel and Namaycush Lake. Edlund’s vouchers are deposited in the National Herbarium of Canada; most are unicates. None of Edlund’s collections have been published, but many were mapped in Aiken et al. (2007). Some 322 of Edlund’s Victoria Island collections, however, were not included in Aiken et al. (2007), as the material was, at the time, part of the unmounted backlog in the National Herbarium of Canada and inaccessible for research.

Botanical exploration of Victoria Island continued in the 1990s. William T. Dushenko made collections at Byron Bay in 1992, in association with research on environmental contaminants (Dushenko et al. 1996). In 1993, Anne H. Weerstra made collections in the Kagloryuak River valley east of the head of Prince Albert Sound, during field research on the breeding biology of the King Eider. Botanist Lynn J. Gillespie, Canadian Museum of Nature, collected in the vicinity of Cambridge Bay in August 1994, and again with Laurie L. Consaul, Canadian Museum of Nature, in July 1997. In 1999, William A. Gould, International Institute of Tropical Forestry, Puerto Rico, made collections at Cambridge Bay, Mt. Pelly, Hadley Bay and along the Tuktu River on central Victoria Island during fieldwork associated with the Circumpolar Arctic Vegetation Map project (Gould et al. 2003). In the same year, Carolyn L. Parker made collections in the vicinity of Cambridge Bay, and Bente Eriksen, Esther Levesque and C.L. Parker made collections in the vicinity of Falaise Bay on Wollaston Peninsula, as part of the Tundra Northwest Expedition 1999 (Eriksen et al. 2006). In 2003 and 2004, collections were made at Ulukhaktok by Robert Bandringa in association with ethnobotanical work (Bandringa et al. 2010).

Over three decades, ecologist Dietbert Thannheiser, University of Hamburg, Germany, conducted phytosociological research at sites on Victoria Island, including Cambridge Bay (1983, 1984, 1986, 1987, 1998), Holman (Ulukhaktok; 1986, 1973), the head of Minto Inlet (1986), Mt. Pelly (1984), Wellington Bay (1983), Hadley Bay (1986), Surrey Lake, Richardson Islands and Johansen Bay (Thannheiser 1986, Thannheiser et al. 2001). Thannheiser et al. (2001) published a list of species occurrences for each of these regions, including numerous first records for Victoria Island. No vouchers are cited, but the paper indicates that vouchers are in personal herbaria of the authors; most or all are now housed in the herbarium of The Arctic University of Norway​ (TROM) but were not examined here. It is unclear, however, if all, or only a subset of material, is vouchered. Some records reported in that paper were considered in Gillespie et al. (2015), but none were considered in Aiken et al. (2007) nor, to our knowledge, in other taxonomic works.

Despite over a century of exploration and documentation of the vascular flora of Victoria Island, only a small subset of the collection data has been published in detail. Material gathered on the island during the Canadian Arctic Expedition was published by Macoun and Holm (1921). Porsild (1955) included voucher information for Victoria Island collections known to him at the time, including details of the collections he made on the island in 1949. It is relatively straightforward to align records reported by Porsild (1955) to maps published in his later treatments (Porsild 1957, 1964, Porsild and Cody 1980), but for collections that were not cited in Porsild (1955), substantial effort is needed to match dots on distribution maps in floristic or taxonomic treatments with vouchers. None of the critical Arctic treatments that cover Victoria Island (Porsild 1957, 1964, Porsild and Cody 1980, Aiken et al. 2007) include details of the specimens on which the works are based. In many instances, Porsild (1955) noted unvouchered observations of species occurrences made by him in 1949, and he included these site records on his subsequent maps without any indication that they are based on unvouchered reports (Porsild 1957, 1964, Porsild and Cody 1980). Aiken et al. (2007) mapped many thousands of records using digital tools but did not publish the dataset underlying the distribution maps. The large dots on the maps in that treatment each cover ca. 40 km, making it difficult or impossible, in some cases, to align a dot to a particular voucher specimen or area, especially in areas where numerous collections have been made, such as in the Cambridge Bay region.

2008, 2010 and 2017

In July 2008, our team, comprising L.J. Gillespie, J.M. Saarela, L.L. Consaul and R.D. Bull (Canadian Museum of Nature), explored and collected plants along southern Victoria Island, Nunavut. Research was carried out under Nunavut Research Institute Scientific Research Licence 0401308N-A, Nunavut Wildlife Research Permit No. WL 2008-1039, Nunavut Water Board Permit No. 3BC-AFP0813 and Polar Continental Shelf Program (PCSP) Project Number 515-08. We established three base camps: (1) at a site ca. 13 km north of Oterkvik Point and 10 km north of the coast at Coronation Gulf (68°36'50"N, 112°34'21"W; 3–11 July); (2) 8 km east-northeast of Johansen Bay airstrip along the Nakoyoktok River at its outflow from a large unnamed lake (68°39'25"N, 110°42'30"W; 12–21 July); and (3) Sinclair Creek North Warning System site (abandoned DEW-line site) (68°45'5"N, 109°06'20"W; 22–24 July). In each area, we explored as many habitats as possible, by foot, at each camp, and made collections as we encountered taxa. Our aim was to document all of the vascular plant species in the vicinity of each of our three camps with at least one voucher specimen. We also explored several remote sites accessed by helicopter. Staging from Oterkvik Point, we visited five sites via helicopter on 7 July: two sites along a bay on Coronation Gulf east of Oterkvik Point (68°30'46"N, 112°33'60"W; 68°29'17"N, 112°40'13"W); an esker ca. 21 km north of the coast at Coronation Gulf and 24 km north-northeast of Oterkvik Point (68°41'59"N, 112°26'23"W); a low rocky ridge between two lakes near the previous locality (68°42'48"N, 112°30'08"W); low rocky hills at the coast in the vicinity of Oterkvik Point (68°31'32"N, 111°59'58"W; longitude coordinates recorded on specimen labels for this site as 112°59'58"W are erroneous). Collections made while exploring the Oterkvik Point area on foot and during the five helicopter stops noted above are recorded in the annotated checklist as being from Oterkvik Pt. Staging from Oterkvik Point, we also visited two more distant sites via helicopter on 8 July: the eastern-most slopes of the Colville Mountains (69°32'45"N, 112°41'27"W) and the vicinity of a river flowing into Clouston Bay, 3–4 km from the river mouth (69°02'39"N, 113°25'15"W). These are treated as separate sites in the annotated checklist. Staging from Johansen Bay we visited five sites via helicopter on 20 July: a flat-topped steep-sided hill, 11 km northeast of the Johansen Bay airstrip (68°39'12"N, 110°54'47"W); a pingo 23 km west of the Johansen Bay airstrip (68°26'23"N, 111°40'22"W); the Johansen Bay airstrip (68°35'50"N, 111°6'59"W); the west end of Johansen Bay at the mouth of Mackenzie Creek (68°36'4"N, 111°21'7"W); and Mackenzie Creek, about 1 km from its mouth at the west end of Johansen Bay, along a river canyon above a waterfall and rapids (68°36'28"N, 111°22'10"W). Collections made while exploring the Johansen Bay area on foot and during the five helicopter stops noted above are recorded in the annotated checklist as being from Johansen Bay. Staging from Sinclair Cr., we visited one site via helicopter on 21 July: Murray Point on the west side of Wilbank Bay (68°35'33.5"N, 110°18'24"W). Murray Point is treated as a separate site in the annotated checklist. In total we made 1091 collections (numbers) of vascular plants, one of lichens and 27 of bryophytes.

In July 2010, we conducted fieldwork on northwestern Victoria Island, Northwest Territories, again aiming to document all species present at each site visited. Our team comprised L.J. Gillespie, J.M. Saarela, Jennifer Doubt, R.D. Bull and P.C. Sokoloff (Canadian Museum of Nature). Research was carried out under Aurora Research Institute Licence No. 14733, Inuvialuit Land Administration Licence No. ILA10HN004, and PCSP Project Number 509-10. We established three base camps: (1) at a site ca. 8 km inland from the head of Minto Inlet adjacent to a Geo-Mapping for Energy and Minerals (GEM) program, Natural Resources Canada, camp (71°37'10"N, 115°26'22"W; 7–8, 19–26 July); (2) at a site on the northeast side of a small round unnamed lake (ca. 1 km diameter) ca. 4 km north of Boot Inlet on the north side of Minto Inlet (71°30'35"N, 117°20'35"W; 9–12 July); and (3) at a site on the southeast side of “Fish Lake” on the lower Kuujjua River between two small lakes, on the south side of Minto Inlet (71°12'28"N, 116°22'46"W; 12–18 July). Collections were also made in Ulukhaktok by J.M. Saarela and R.D. Bull (5–6 July; Saarela nos. 1410–1508). Staging from the Kuujjua River camp we visited three sites by helicopter on 17 July: the base of north-facing cliffs 68 km east-northeast of Ulukhaktok (71°2'60"N, 116°9'48"W); the sandy banks of the Kuujjua River south of “Fish Lake”, ca. 17 km southeast of the head of Minto Inlet (71°6'43"N, 116°6'21"W); the shore of the Kuujjua River delta at Minto Inlet (71°15'23"N, 116°49'35"W). Staging from the Minto Inlet camp we visited three sites via helicopter on 25 July: a deep canyon on an escarpment south of the head of Minto Inlet in the vicinity of large waterfall (71°25'30"N, 115°12'2"W), a coastal saline flat along the end of the eastern most inlet (north arm) at the head of Minto Inlet (71°31'7"N, 115°6'30"W), and an esker on a plain 3 km south of the head of Minto Inlet (71°27'10"N, 115°17'3"W). We made 1048 collections (numbers) of vascular plants, one of lichens, five of bryophytes and 14 of fungi. Jennifer Doubt made extensive collections of bryophytes during this expedition, which are not considered here.

In 2017, 20 collections were made in Cambridge Bay by J.M. Saarela (nos. 5296–5301) during the Canada C3 expedition (https://canadac3.ca/en/homepage/).

All 2008, 2010 and 2017 collections were dried in the field in standard plant presses. For each collection we preserved a small sample of leaf tissue in silica gel for future molecular analyses. In most cases, we tagged the plant from which we obtained the sample. These tissue samples are preserved in the National Biodiversity Cryobank of Canada at the Canadian Museum of Nature. These collections were variously determined by L.J. Gillespie, J.M. Saarela and P.C. Sokoloff unless otherwise indicated (Suppl. material 2). Willows were identified by George Argus (CAN). A subset of our Draba, Papaver and Potentilla collections were determined by R. Elven and confirmed by us. The first set of our collections is deposited in the National Herbarium of Canada (CAN), Canadian Museum of Nature. Duplicate specimens have been distributed to the following herbaria, as noted in the specimen citations and Suppl. material 2: University of Alaska Museum of the North (ALA); Aurora Research Institute, Inuvik (indicated in the annotated checklist as “ari”, as the herbarium does not have an official acronym); the University of Alberta Vascular Plant Herbarium (ALTA); Icelandic Institute of Natural History, Akureyri Division (AMNH); the B.A. Bennett Herbarium (BABY), Yukon; the Botanical Museum in Oslo (O); the Missouri Botanical Garden (MO); the Marie-Victorin Herbarium (MT), University of Montreal; the Ayre Herbarium (NFLD), Memorial University of Newfoundland; the herbarium in the Beaty Biodiversity Museum (UBC) Beaty Biodiversity Museum; the United States National Herbarium (US), National Museum of Natural History, Smithsonian Institution; the Intermountain Herbarium (UTC), Utah State University; the Royal British Columbia Museum (V); the University of Manitoba Herbarium (WIN); and the herbarium of the Institut für Systematische Botanik (Z), Universität Zürich.

2013, 2014

Fieldwork was conducted by B.A. Bennett in the vicinity of Cambridge Bay in 2013 and the Cambridge Bay and broader vicinity in 2014. The aim of this work was to document all species of vascular plants in the area to contribute to understanding of the ecology of the greater Canadian High Arctic Research Station (CHARS) (Polar Knowledge Canada) research area. In 2013, collections were made at numerous sites within the community, along the road to Mt. Pelly, within Ovayok Territorial Park, at the east end of Greiner Lake, and along the road west of the community, as far west as the hills above Long Point. In 2014, collections were made within the hamlet, across the bay from the hamlet, north of the DEW line site, at Long Point and at three sites more distant from the hamlet: “30-Mile Creek”, near the Ekalluk River and the west end of Ferguson Lake, and “Trunsky Lake”. Vouchers of material collected in 2014 has been distributed to ALA, BABY, CAN, MO, UBC, V as well as the CHARS herbarium (indicated in the annotated checklist as “chars”, as the herbarium does not have an official acronym) in Cambridge Bay; Stanley L. Welsh Herbarium (BRY), Brigham Young University; National Collection of Vascular Plants (DAO), Agriculture and Agri-Food Canada, Ottawa; Herbarium (MICH), University of Michigan; Michael J. Oldham herbarium, Peterborough, Ontario (indicated in the annotated checklist as “od”, as the herbarium does not have an official acronym); Herbarium (SRP), Boise State University; Herbarium (UAAH), University of Alaska Anchorage; and Herbarium (WTU), University of Washington. We here consider 319 collections made by B. Bennett and associates in 2013 and 2014.

2016, 2018 and 2019

Botanical fieldwork was conducted by CHARS ecologist S. Ponomarenko during three field seasons. From 3–18 August 2016 a vegetation survey was carried out for the CHARS Intensive Monitoring Area (IMA), an area restricted to two small watersheds within the Greiner Lake watershed that drains to the northern shore of Greiner Lake approximately 20–30 km to the northeast of Cambridge Bay, an area measuring about 50 km². This work was conducted as part of the high resolution vegetation mapping of the CHARS IMA project (Ponomarenko et al. 2019).

In the 2018 and 2019 field seasons, floristic collections were obtained during helicopters surveys of a broader area. In 2018, five days of helicopter surveys were undertaken within the Greiner Lake watershed, an area measuring approximately 1,500 km². In 2019, eight days of helicopter surveys were undertaken north and east of Greiner Lake watershed along the border of bioclimatic subzones C and D. This survey covered an area of about 2,000 km² north and east from the Greiner Lake watershed. This work constituted a part of the vegetation inventory and mapping project aiming to assemble baseline data for the CHARS ERA.

Vouchers collected during these three years of fieldwork are housed at CAN and CHARS. In total more than 700 specimens were collected, of which 394 collections were reviewed for this work. The remaining 300 collections, mainly collected in 2019, are not yet processed.

Herbarium research and data curation

In addition to accounting for our new collections, we attempted to locate and confirm or revise determinations of all unique collections of vascular plants from Victoria Island. To find specimens collected previously from the study area, we manually searched the collections at CAN, DAO, MT, the University of Calgary (UAC), UBC and Herbier Louis-Marie (QFA), Université Laval, and we queried the Global Biodiversity Information Facility (GBIF), Canadensys (http://data.canadensys.net/explorer/en/search), the Consortium of Pacific Northwest Herbaria (http://www.pnwherbaria.org/about.php), Integrated Digitized Biocollections (iDigBio) portal (https://www.idigbio.org/portal/search), and various institutional online databases. In addition to the herbaria noted above, specimens from Victoria Island cited here are deposited in the R.L. McGregor Herbarium (KANU), University of Kansas; H.A. Stephens Herbarium (KSTC), Emporia State University; TRTE Herbarium (TRTE), University of Toronto Mississauga; Herbier du Québec (QUE), Sainte-Foy, Québec; and Harvard University Herbaria (GH). We also reviewed relevant taxonomic and floristic literature (Porsild 1955, 1957, 1964, Porsild and Cody 1980, Aiken et al. 2007), conducted online searches, and consulted an unpublished database of specimen records that was developed and used in production of the "Flora of the Canadian Arctic Archipelago" (Aiken et al. 2007). We have seen and confirmed at least one duplicate for most of the specimens cited. A subset of material was confirmed based on review of specimen images available online. Determinations for some specimens we have not seen are accepted based on the authority of previous determiners, especially for taxa that are well known and not taxonomically problematic. We have also accepted some non-confirmed records from well-collected sites for which we have reviewed and confirmed numerous other records, such that a possible mis-identification of one of many records of a species from a site would not affect understanding of diversity at the site. With a few exceptions, we did not study the 19^th century collections, for which site information is vague.

As part of the current study, three batches of unprocessed material gathered on Victoria Island in the 1940s, 1940s–50s and 1980s that was stored in the backlog collection of the National Herbarium of Canada were organized, identified, mounted and inserted into the permanent collection. One batch comprised 134 sheets collected by Nicholas Polunin at Cambridge Bay in 1947. The second batch comprised 498 sheets collected by Margaret Oldenburg at various sites across the island (see Introduction). The third batch comprised 359 collections gathered by Sylvia Edlund at various sites across the island in 1982, 1986 and 1987 (see Introduction). In the course of processing this material, we discovered that 38 of these Edlund backlog collections from Victoria Island had previously been assigned CAN accession numbers and were recorded in the museum database (and thus mobilized online), but the material remained unmounted, until now. These 991 collections–of which some were gathered 70+ years ago – are now available to the scientific community and are published here. The backlog material from Victoria Island dealt with here is a subset of larger backlog batches of Arctic specimens at CAN gathered by Edlund, Oldenburg and Polunin.

During the course of this study, imaging of all vascular plant material housed at CAN from Yukon, Northwest Territories and Nunavut was completed and all images were linked to the institutional database and mobilized via GBIF. GBIF tools were used to identify and fix putative data errors in Victoria Island records. The images also facilitated completion of data entry for CAN specimens from Victoria Island for which only “skeletal” records existed (i.e., taxon name and higher-level geographic provenance – country and province/territory) or for which data entry was otherwise incomplete.

We amalgamated all collection data obtained from different resources into a spreadsheet. Substantial manual cleaning of the complete dataset was undertaken to make the dataset useable, including standardizing names of collectors, date format and locality descriptions among specimens gathered by the same collector at the same site. We combined records of duplicate specimens housed in different collections into single records, maintaining information on the disposition(s) of the duplicate specimens.

We spent considerable time improving the georeferencing of records in our dataset. Most collections from Victoria Island were made long before the existence of the Global Positioning System and many before the existence of detailed topographical maps, or of maps, period. As such, many collections either lacked coordinates or included the following: (1) inaccurate or erroneous primary coordinates, (2) accurate or inaccurate secondary coordinates but no attribution or source information for the georeferenced coordinates, or (3) accurate but imprecise coordinates determined to the nearest minute, being within ca. one nautical mile of the site, assuming the coordinates are correct. Prior to the advent of digital mapping, this level of precision was generally sufficient for dot-based distribution maps presented at a fixed scale. For example, each dot in the maps in Porsild and Cody (1980) covers about 110 km across (ca. 9,503 km²) and in Aiken et al. (2007) about 40 km across (ca. 1,256 km²). In digital environments, however, inaccurate coordinates may be problematic and misleading. For many Victoria Island collections, imprecise label coordinates often placed terrestrial collecting sites in freshwater or marine environments or far from descriptions of the places where collections were made, well beyond distances likely to have been travelled by earlier collectors by foot in the limited time that was typically available for collecting at a site.

Therefore, to improve accuracy of geographical coordinates, we secondarily georeferenced many sites following standard point-radius protocols, including determining estimates of coordinate uncertainty in metres, in cases where we were confident that existing location data could be improved upon. Georeferencing was done by J.M. Saarela and Paul Wise, Canadian Museum of Nature. Georeferencing data is included in Suppl. material 2. Where possible, we refined georeferencing data through consultation of published information about the itineraries of past collectors. We also attempted to standardize georeferencing data for collections made in the same area by the same collectors at the same time and by different collectors at different times. At times, this required superseding previous georeferencing for a site. For example, many collections over several decades have been made at “Cambridge Bay”, and that area has been georeferenced differently many times by different workers at different herbaria, including georeferencing the hamlet and the bay, which both have the same name even though it is safe to assume that collections were made on land in the vicinity of the hamlet, not in the bay (a marine environment). The name “Holman Island” on labels has also been the cause of confusion. The hamlet of Ulukhaktok, the Kangiryuarmiutun name for the area, was until 2006 known as Holman and sometimes referred to as Holman Island, as it is on many early herbarium collections. Holman Island is also the name of a small island in the Amundsen Gulf to the east-southeast of Ulukhaktok, after which the community was originally named. Holman Island was named for John R. Holman, a member of the Inglefield Arctic Expedition (1853–54). Although contemporary workers have regularly confused the two places when georeferencing, we are not aware of any botanical collections from the true Holman Island (70°39'6"N, 117°43'33"W), and we have corrected those errors. Collections made at Holman or “vicinity of Holman Post [trading post]” up to 1966 were gathered on the east side of Kings Bay, the site of the community until it moved across the bay to its current location on Queen’s Bay in that year. All collections from the community now known as Ulukhaktok are included under that name.

Distribution maps were generated in ArcMap 10.5.1. Additionally, using ArcMap 10.5.1, we generated maps showing the locations of all collecting sites on the island, maps showing the locations of collections made by S. Edlund, including for specimens previously accessioned into the CAN herbarium and for specimens that had been housed in backlog and were newly accessioned as part of this study, and collections made by A.E. Porsild, Gillespie et al., B. Bennett, S. Ponomarenko, and M. Oldenburg. A heat map showing the density of collections at sites on the island was generated using QGIS 3.4.

Annotated checklist

The vascular flora of Victoria Island is summarised in an annotated checklist. Classification of lycophytes and ferns follows The Pteridophyte Phylogeny Group (2016). Angiosperms are organized according to the linear classification of flowering plants proposed by The Angiosperm Phylogeny Group (2016). Genera are listed alphabetically within families and species are listed alphabetically within genera. Taxonomy at genus, species and infraspecific levels is based on consideration of the relevant global taxonomic literature, including Elven et al. (2011), treatments in the Flora of North America series (Flora of North America Editorial Committee 1993+), and taxonomic monographs and revisions, such as Mosyakin (2016) and Wiegleb et al. (2017). For each species we provide important synonyms, focusing on names used in critical Canadian (Porsild and Cody 1980, Aiken et al. 2007) or international (Elven et al. 2011) Arctic taxonomic treatments, more recent national or continental treatments, particularly the Flora of North America, and other taxonomic works. Common name(s) in English are mostly from the Flora of North America series and Brouillet et al. (2010+). Global distribution summaries are those described by Elven et al. (2011). For each taxon recorded from Victoria Island we include a summary of the broader distribution in the Canadian Arctic, referencing primary and secondary literature, and, in a few case, unpublished specimens. For species that are rare on Victoria Island and known from only one or few collections, we provide information on habitat, when available. Photographs are included for a subset of taxa to facilitate identification and illustrate habitats. We also provide keys to the families, genera, species and infraspecific taxa, to facilitate identification of plants on the island. Keys were adapted from diverse published sources. Measurements in the keys refer to lengths unless otherwise indicated.

All species reported for the study area are documented by one or more voucher specimens, and only vouchered records are included and mapped. Observations of species noted by Porsild (1955), which were mapped in subsequent work (Porsild 1957, 1964, Porsild and Cody 1980, Aiken et al. 2007), are mentioned in the text unless we cite a voucher obtained by a different collector from the same area, confirming the occurrence of the taxon in that area. These comments thus identify occurrences mapped in earlier treatments for which no voucher exists. For each collection, we list the collector(s), collection number or s.n. [sin nombre = without number] if no collection number exists, and the code(s) identifying the herbaria where the collection is housed. Nineteenth century collections are summarized in Suppl. material 1 and not included in the annotated checklist, except in a few cases. In a few cases, unvouchered observations of rare taxa in the Cambridge Bay are noted, but not mapped; collections should be obtained for proper documentation of these occurrences.

To simplify publication of voucher information, we assigned each collection to a general area of the island. Where possible, we used general areas as described on specimen labels. In cases where no or vague site information is given on labels, however, we assigned specimens to a nearby named place. For example, collections by Edlund reported from south of Burns Lake were gathered in 1982 and 1987 at sites ca. 40–45 km south or south-southwest of the lake, but on the specimen labels the location is described only as “Geological Survey of Canada peat study location” along with the coordinates. In most cases, the toponyms we use are recognized by the Geographical Names Board of Canada. A few sites names, however, are not officially recognized. Examples include “Oldenburg Lake”, “Trunsky Lake” and “30-Mile Creek”, also known locally as “30 Mile River” and “Halovik River”. Locations of all general areas are shown in Fig. 1. The “Cambridge Bay” area includes all collections from the immediate vicinity of the hamlet, along the road to Ovayok Territorial Park (but not including collections made within the park limit), and along and in the vicinity of the road west of the hamlet, including Long Point and the Augustus Hills. Ulukhaktok includes the immediate area of the hamlet. Collections made by S. Ponomarenko recorded from Greiner L. include specimens taken from throughout the Greiner L. watershed, except those that were gathered close to the west side of Albert Edward Bay, which are recorded as being from that general area. The following abbrevations are used in the Annotated Checklist: B., Bay; C., Cape; Cr., Creek; I., Island; Inl., Inlet; Mts., Mountains; P., Peninsula; Pt., Point; R., River; S., Sound; TP, Territorial Park.

Some toponyms on Victoria Island have changed in recent years. Ferguson Lake, a large lake north of Cambridge Bay that flows into Wellington Bay via the Ekalluk River, is officially known as Tahiryuaq, the Inuinnaqtun name. We use the English name for the lake to avoid confusion with two other lakes on Victoria Island also officially named Tahiryuaq. Both of these are in Northwest Territories, one north of Prince Albert Sound (70°56'2"N, 112°15'7"W) where collections were made by Edlund and the other southeast of the head of Minto Inlet (71°27'2"N, 114°45'8"W). The esker formerly known as Ovayuk/Mount Pelly, where many collections have been made, was renamed Uvayuq, effective 21 September 2012. The esker northwest of Mount Pelly and north of Cambridge Bay formerly known as Mount Lady Pelly (69°15'28"N, 104°48'37"W), where collections were made in 1962, was renamed Amaaqtuq, effective 21 September 2012.

For each taxon recorded from Victoria Island we summarize previous reports of the species occurrence on the island by site and indicate sites for which a taxon is newly recorded here. “Previously recorded” at a site means a species occurrence was stated or mapped for a site in one or more earlier published works, which are cited in the text here, regardless of whether or not voucher information was published, and the earlier report is supported by one or more vouchers cited in the text here, unless otherwise indicated. In most cases, the vouchers cited here are the collections on which the earlier records were based. In many cases, new material has been gathered from sites at which a species was previously reported, particularly those areas that are best collected, like Cambridge Bay and Ulukhaktok. Although we do not, in most cases, comment on the accumulation of material of a species at a particular site, we do cite all the relevant material known to us from each site.

“Newly recorded” at a site means the current study is the first to publish an occurrence of a species at a particular site. Such new records reported here include collections made during our fieldwork from 2008 on, earlier collections that previous workers did not consider in their floristic treatments, earlier collections that were not processed and available for study until recently, and new identifications of material known to previous authors, either because previous determinations were incorrect or in light of revised taxon circumscriptions. In all cases we include explicit citation of one or more voucher specimens supporting new records.

Thannheiser et al. (2001) reported species from numerous sites across Victoria Island, including many listed as new records for sites. Because we have not reviewed any voucher specimens supporting those reports, nor are we aware of the existence of vouchers for all reports, we do not include these records in our maps. Nevertheless, we recognize the importance of the work done by Thannheiser and colleagues over many years on Victoria Island and note all instances where Thannheiser et al. (2001) report a species from a site for which we have not seen a voucher. In cases where Thannheiser et al. (2001) newly reported a species from an area and we have confirmed and cite a collection made by a different collector in the same area, we state that occurrence of the species in the area as reported by Thannheiser et al. (2001) is confirmed (“conf.”); this does not mean we have confirmed a voucher taken by Thannheiser and colleagues supporting their report of the species. This is the case, for example, for many records newly reported from Johansen Bay by Thannheiser et al. (2001) confirmed to be present in that area by collections we made in that area in 2008.

Patterns of floristic diversity

To characterize patterns of floristic diversity on Victoria Island we scored, based on our dataset, the presence of species/taxa in the following regions: Nunavut and Northwest Territories, the six areas that we aimed to document comprehensively in 2008 and 2010, two additional well-collected areas (Cambridge Bay, Ulukhaktok), and Ovayok Territorial Park, the only protected area on the island (Table 2). The geographical limits of all these areas correspond to the way the sites are listed in the annotated checklist (see above), except we included Murray Point in the Johansen Bay area rather than treating it separately. We determined the area (km²) represented by each of these general study areas by drawing a bounding box around all sites in Google Earth (Google 2019) and using the polygon measuring tool: Oterkvik Point: 355 km²; Johansen Bay: 212 km²; Sinclair Creek, 1 km²; Boot Inlet: 8.77 km²; the head of Minto Inlet: 128 km²; Kuujjua River: 198 km²; Cambridge Bay: 200 km²; Ulukhaktok: 16.5 km². We constructed species discovery curves for the six 2008 and 2010 areas to characterize the cumulative number of species recorded at each as a function of the time (days) we spent searching for species diversity. We also scored the presence of taxa in the Circumpolar Vegetation Map bioclimate subzones C and D across Victoria Island (CAVM Team 2003), in the Nunavut portion of the island, and the ecological regions in the Northwest Territories portion of the island defined by Ecosystem Classification Group (2013). For the Nunavut portion of the island we additionally scored presence/absence of taxa east and west of Wellington Bay and in two subzones of bioclimate subzone C, one comprising Storkerson Peninsula, the head and west side of Hadley Bay and Natkusiak Peninsula and the other comprising the rest of the subzone across the middle of the island.