Araneae of Canada

Abstract In 1979 nearly 1400 spider species in 32 families either had been recorded (1249) or were believed to occur (~140) in Canada. Twenty years later, although significant progress had been made in survey efforts in some regions, Canada’s spider inventory had only increased by approximately 7% to roughly 1500 species known or expected to occur. The family count had increased to 38 but only two additions were truly novel (five family additions and one family deletion were the result of advances in family-level systematics). The first comprehensive taxonomic checklist of Canadian spider species was published in 2010 documenting the regional distributions of 1376 species representing 42 families (three novel since 1999). From 2010 through 2017 new national records steadily accumulated resulting in the current (2018) Canadian inventory of 1477 species classified in 45 families (one novel since 2010). Although there has been close to a 20% increase in the number of spider species recorded in Canada since 1979, much greater increases have occurred in some of the regional species checklists, indicating increasing knowledge of the regional distribution of species previously recorded elsewhere in Canada. For example the regional checklists for Newfoundland, British Columbia, and Prince Edward Island have increased by 69%, 339%, and 520%, respectively. The national and regional increases reflect significant advances in the first two decades of the 21st Century in spider faunistics research in previously under-sampled habitats and regions and the development of molecular techniques and consequent barcoding of spiders. Of the 1477 species recorded in Canada, 92% have been successfully DNA barcoded resulting in 1623 unique Barcode Index Numbers (BINs). At least 25 of the BINs are associated with relatively easily distinguished but undescribed morpho-species. The majority, however, appear to indicate the existence of many cryptic species within Canada’s known spider fauna. These data, coupled with the fact that novel Canadian or even Nearctic spider species records (including of undescribed species) continue to accumulate annually (especially in habitat-diverse regions such as British Columbia), suggest that Canada’s tally of spider species may approach or even exceed 1800.


Introduction
Canadian national spider faunistics work commenced with Emerton's (1920) initial listing of the names of and basic locality data for the spider species then known to occur in Canada. Emerton's list was far from complete, reflecting the scant knowledge of Canada's spider fauna at that time. Nearly 60 years later the accumulation of sufficient data allowed production of the first reasonably comprehensive faunistic overview of the spiders of Canada (Dondale 1979). Subsequently, a series of increasingly detailed studies of the fauna has significantly improved the situation. Dondale (1979) provided the initial faunistics framework: a family-level table outlining general distributions and estimated species numbers for each, accompanied by a brief introduction to spiders, Canada's spider fauna, and basic biology of the major families.
Twenty years later, Bennett (1999) supplied a similar table with revised family-level nomenclature and species statistics, as well as more detailed family-level biology and comments on the then-current state of araneology in Canada. Subsequently, Paquin et al. (2010) published a checklist of all spider species found in Canada up to that date including species-level provincial and territorial distributional data, tables with summary family-level statistics, and a list of introduced species.
Most recently, the Canadian Endangered Species Conservation Council (2011,2016) provided updated species-level checklists, including novel records and nomenclatural changes, as well as national and provincial/territorial conservation ranks for all species using NatureServe distribution and abundance analysis methodology (see Faber-Langendoen et al. (2012) for a description of NatureServe's methodology). Since then, two of us have maintained and updated the Canadian national and regional checklists (R Bennett and G Blagoev unpubl. data, summarised in Table 1). Unsurprisingly, the known Canadian spider fauna has been (and continues to be) dominated by species of Linyphiidae (Table 2).
Prior to 1979, regional spider species checklists existed for only two Canadian regions (Table 3): British Columbia (Thorne 1967, Bragg andLeech 1972) and the island of Newfoundland (Hackman 1954). Species-level data now exist for all Canadian provinces and territories (Table 3; Paquin et al. 2010) and a majority of these have series of three or more checklists showing increases in the accumulation of novel jurisdictional species records of greater than 20% (Alberta, Manitoba, Northwest Territories, Québec, and Yukon), nearly 70% (island of Newfoundland), to well over 300% (British Columbia) or 500% (Prince Edward Island).
Novel species records, including of undescribed species, continue to accumulate annually in Canada especially through the work of the University of Guelph's Centre for Biodiversity Genomics (CBG) at the national level and, regionally, where intensive spider inventory work is on-going, e.g., in British Columbia through the efforts of the Royal British Columbia Museum (RBCM). Ideally, faunistics research generates sufficient distribution and abundance data to allow for conservation ranking analyses. For more than one third of Canada's recorded spider species, however, only simple presence data exists and conservation ranks remain undetermined for those species. It is clear that much work remains to be done on Canadian spider faunistics. Canada's national and provincial/territorial spider species checklists will continue to grow, our knowledge of Canadian spider morphological and molecular taxonomy will continue to improve, and accumulating inventory data will allow for conservation ranking of a significant proportion of Canada's currently unranked spider fauna.
Here we present an overview of Canadian spider faunistics work from 1979 to 2018, focusing on the accumulation of national and regional species-level data. Classification follows Wheeler et al. (2017); nomenclature follows the World Spider Catalog (2018). Standard postal abbreviations are used for Canada's provinces and territories except for LB and NF, which refer to mainland Labrador and the island of Newfoundland, respectively: AB -Alberta, BC -British Columbia, MB -Manitoba, NB -New Brunswick, NS -Nova Scotia, NT -Northwest Territories, NU -Nunavut, ON -Ontario, PE -Prince Edward Island, QC -Québec, SK -Saskatchewan, and YT -Yukon.

In the beginning (Dondale 1979)
Although Emerton (1920) presented a cursory list, including basic locality data, of about 340 spider species then known to occur in Canada, the first serious summary of faunistics of Canada's spider species was published by Dondale (1979). The single table in that paper provided the faunistics framework for future studies of Canada's spider fauna: a list of the 32 families recorded to that date accompanied, for each family, with general Canadian ecological distribution data and the number of species known in Canada. Loxoscelidae (now Sicariidae), a 33 rd family listed in the table, was then unrecorded in Canada and has not become established subsequently. Dondale predicted a total of about 1400 species for Canada's spider fauna: 1249 recorded (Table 3; number here revised down from 1256 because of enumeration errors in Dipluridae, Antrodiaetidae, Dysderidae, and Segestriidae (CD Dondale pers. comm.)) with a further 144 either known but undescribed or expected but not yet recorded.
In 1979, only two Canadian jurisdictions, BC and NF (Table 3), had spider checklists, and these were not comprehensive for either province. Thorne (1967) listed 212 species in 20 families for BC, subsequently upgraded to 259 species by Bragg and Leech (1972). Thorne noted BC's spider fauna was poorly known and implied that significant increases to the provincial checklist would result from better knowledge, in general, of the province's northern interior and northern and eastern border areas and, in particular, of the province's erigonine linyphiid fauna (see Bennett (2001) for a brief history of araneological work in BC from the late 1800s to 2001). Hackman (1954) documented 216 species (originally 220, revised down by Pickavance and Dondale (2005)) in 19 families for NF. Because his study was based almost entirely upon only two sets of collections made over two summers (1949,1951) by visiting scientists from Fennoscandia, it is clear that Hackman's (1954) checklist was not comprehensive.

Making progress (Bennett 1999)
Twenty years after Dondale's (1979) treatment of the national fauna, Bennett (1999) updated Canadian family-level spider species statistics. Estimates, rather than exact counts, were made for most of the larger entelegyne families (e.g., Dictynidae -"75-80", Linyphiidae -"> 500") and the total number of species then recorded in Canada was estimated to be approximately 1400 (Table 3). Bennett (1999) speculated that the  Thorne 1967. 7 Bragg and Leech 1972. 8-9 West et al. 1984, 1988. 10 Bennett 2001. 11-16 Bennett et al. 2006, 2010, unpubl. data. 17 Perry et al. 2014  species count could reach 1500 with concerted inventory work in particular habitats, especially forest floors and canopies. Increases in species counts (Table 2) were, as expected, primarily in Linyphiidae (more than 50 species added) but also in groups that had been subject to significant taxonomic revision since 1979, such as Gnaphosidae (about 40 species added) and Lycosidae (20 species). Species counts for other families were relatively unchanged. Although Bennett's (1999) checklist recorded 38 spider families in the Canadian fauna, six more than reported by Dondale (1979), only two families, Zoridae and Mysmenidae, were truly novel additions through the discovery of populations in BC of species known to be relatively widespread in the western United States. Five other family additions were not novel but the result, rather, of advances in spider systematics and classification: in Dondale (1979) Cybaeidae was classified in Agelenidae, Liocranidae and Corinnidae in Clubionidae, Titanoecidae in Amaurobiidae, and Pimoidae in Linyphiidae. Dondale's (1979) Erigonidae was not included in Bennett (1999) and subsequent lists because of its reclassification as a subfamily of Linyphiidae.
In the 1980s and 1990s, considerable progress was made on Canada's regional spider checklists (Table 3), a reflection of significant spider inventory work in some regions. BC's spider species count more than doubled to 570 (West et al. 1984(West et al. , 1988 and initial checklists were produced for MB (483 species; Aitchison-Benell and Dondale 1990), QC (549; Bélanger and Hutchinson 1992), and YT (297; Dondale et al. 1997). As with the earlier checklists for BC and NF, the regional checklists of the 1980s and 1990s were still far from complete and major changes would be made through the next couple of decades.

The recent past (Paquin et al. 2010)
Early in the first decade of the 21 st Century, DJ Buckle (pers. comm.) began compiling a checklist including the names and basic regional (Canadian provinces and territories and/or Alaska) presence data for all spider species then recorded in Canada and Alaska. The database was subsequently published (Paquin et al. 2010), making widely available for the first time a reasonably comprehensive Canadian national and regional species-level checklist. Thirty-six of the species were known from Alaska but not from anywhere in Canada (one species in each of Clubionidae and Dictynidae, three in Lycosidae, and 31 in Linyphiidae) and one South American species (Sicariidae, see above) was not established in Canada then and has not become so subsequently. Removing those 37 species from consideration, the 2010 checklist recorded 1376 species (Table 3) representing 42 families in one or more of Canada's provinces and territories. Although this species count was not appreciably different from Bennett (1999), four families were newly recorded by Paquin et al. (2010). Three of these families were first Canadian records of non-native species in Amphinectidae (one species, now classified in Desidae), Oonopidae (two species), and Zodariidae (one species). The addition of a fourth family, Miturgidae, was the result of advances in systematics (transfer of two well-known species from Clubionidae). Among the new introductions, the status of the desid and the oonopids is uncertain. No new records are known since the first (and only) record of the desid in AB in 1991. One of the oonopids has been collected only once, albeit from a natural (but disturbed) outdoor habitat in Victoria BC; the other species is known from a substantial indoor population in Toronto ON (Platnick et al. 2012) and is probably established there. The zodariid is almost certainly an established member of the Canadian fauna; it is now known from relatively natural habitats in southern localities in three provinces (R Bennett and G Blagoev unpubl. data) and appears to be well established in the northern United States (Ramseyer 2016). As in Bennett (1999), family-level species counts in Paquin et al. (2010) showed substantial increases only in Linyphiidae, up about 30 since 1999 (Table 2).
In addition to revising the species names and count data for regions with existing species-level checklists (AB, BC, MB, NF, QC, and YT), a major benefit of Paquin et al. (2010) was the provision of species-level data for those Canadian regions where spider species checklists had not previously existed (Table 3). The low species counts for PE and NU reflected the fact that no serious spider inventory work had yet been done in PE (Bowden et al. in press) and very little in NU (e.g., Leech 1966). The species numbers recorded in 2010 for the remaining "first-time" regions are an indication of substantial spider inventory work in those regions, largely led by CD Dondale and JH Redner of the Canadian National Collection (CNC) but also with significant input of specimens and data from CNC entomologists, Canadian Forest Service researchers, and non-government arachnologists such as DJ Buckle, R Holmberg, and RE Leech.
In the first decade of the 21 st Century, Canadian regional spider faunistics work continued to be especially active in BC and QC where, by 2010, both provinces had recorded increases of about 130 species to 700 and 677, respectively (Table 3). In addition, a short burst of intensive spider faunistics work in NF resulted in an increase of more than 150 species for a total of 363 species on the island's spider checklist (Pickavance and Dondale 2005). Based primarily on large numbers of spider specimens collected during research in the 1990s in localised areas of southern and central BC led by DCA Blades (Blades and Maier 1996), GGE Scudder (unpublished), N Winchester (Copley 2010, Copley andWinchester 2010), and BS Lindgren and J Lemieux (Lindgren et al. 1999), Bennett (2001) and Bennett et al. (2006) updated, respectively, the BC spider species count and the checklist of names and locality data. Family and species-level nomenclature in Bennett et al. (2006) were in agreement with current World Spider Catalog (Platnick 2006) standards, thus simplifying tracking of taxonomic synonyms and other nomenclatural changes in preceding and subsequent BC spider checklists. Two of us (R Bennett and C Copley) and D Copley used locality data in Bennett et al. (2006) to map unsampled or poorly sampled regions and habitats in BC and plan a series of annual field surveys, primarily targeting alpine and subalpine habitats across the province, to address the knowledge gap. With support from the RBCM and a wide range of collaborators, these surveys commenced in 2008 and the first two field seasons resulted in the addition of more than 40 species to the known spider fauna of BC (Bennett et al. 2010), some of which were new records for Canada or even the Nearctic.
In QC, the significant efforts of the team of Dupérré (2003, 2006) in the field (as well as in examining specimens in the CNC for overlooked new records) were largely responsible for the increases in the list of spider species known in QC.

The current era and into the future (Canadian Endangered Species Conservation Council 2016 and the rise of BINs)
In 2010, using a draft version of the species list in Paquin et al. (2010) as their baseline, attendees at an ad hoc workshop organised by NatureServe Canada produced an initial consensus-based assessment of the national and regional conservation status of all spider species recorded in Canada (Canadian Endangered Species Conservation Council 2011). In 2013, the Canadian Wildlife Service (Environment Canada) began a comprehensive reanalysis of the conservation status data using NatureServe methodology (Faber-Langendoen et al. 2012) and incorporating all available spider species distribution and abundance data as well as the significant number of novel species records that had accumulated since 2010. The results of the analysis included updated national and regional checklists recording 1399 spider species (Table 3) representing 42 families (data current to 2013; Canadian Endangered Species Conservation Council 2016) and, in 2018, this is the contemporary published resource for species-level faunistic data for Canadian spiders. Family number remained unchanged from 2010 although Zoridae was incorporated into Miturgidae and Anapidae was newly recorded for Canada. Notably, the anapid record, Comaroma mendocino (Levi), collected in BC, is one of the rarest of all native spider species in North America (see Lopardo and Coddington 2017). New records of Linyphiidae accounted for two thirds of the relatively minor increases in Canada's family-level spider species counts reported in the 2016 publication (Table 2).
Since 2013, i.e., data incorporated in Canadian Endangered Species Conservation Council (2016), new national records have accumulated rapidly. Seventy-eight new species records have been documented from across Canada (with the majority from BC) resulting in the current Canadian inventory of 1477 spider species in 45 families (Table 1). Linyphiid species account for more than one third (27) of the new records (Table 2) but some families which had not seen much change for several decades also show substantial increases, for example araneid and salticid species counts have increased by 9 and 23 (Table 2), respectively. All three of the new family names in Table  1 are the result of recent classification changes: Eutichuridae was previously classified in Miturgidae while Phrurolithidae and Trachelidae were extracted from Corinnidae.
As with previous checklists, the increased species counts largely reflect the continuation of serious regional spider inventory work. However, molecular data have become increasingly important in Canadian spider faunistics and systematics, and the current Canadian spider species checklist (R Bennett and G Blagoev unpubl. data) is the first to benefit from the use of molecular data. The molecular work, led by the CBG, has combined laboratory work with comprehensive sampling of arthropod specimens in a wide variety of often difficult-to-access and/or previously poorly known habitats in Canada (deWaard et al. 2017). The RBCM has closely collaborated with the CBG since 2012, resulting in large amounts of novel regional as well as national specieslevel data (Bennett et al. 2017). In particular, molecular data have been critical in the identification of specimens of described species that previously were unidentifiable (or easily misidentified) because of the lack of good morphology-based diagnostic images or reliably identified voucher specimens (e.g., various linyphiid taxa).
More than 50,000 Canadian spider specimens have been analysed in the course of the barcoding work, representing 92% of the 1477 spider species currently recorded in Canada (Table 1; R Bennett and G Blagoev unpubl. data) and generating 1623 individual Barcode Index Numbers (BINs) (Ratnasingham and Hebert 2013). Unsurprisingly, because nearly 40% of all Canadian spider species are linyphiids, nearly one third of the sequenced specimens are Linyphiidae and those specimens account for nearly 40% of the resultant BINs. BINs in excess of known species-level taxon numbers usually indicate the potential for undescribed species and complexes of cryptic species. This is particularly evident among some smaller families (e.g., Telemidae, Theridiosomatidae, Tetragnathidae, and Trachelidae) which have generated 2-3 times as many distinct BINs as there are named species in the Canadian fauna of those families. Such BIN disparity can be particularly striking for individual species. For example, 18 BINs are associated with a single species, Tetragnatha versicolor Walckenaer. For this species, the disparity is not a big surprise: T. versicolor is well known for its morphological variability and probable masking of a cryptic species complex (Levi 1981). Similarly, 19 and 25 unique BINs have been generated from specimens of Evarcha hoyi (Peckham and Peckham) (Salticidae) and Grammonota angusta Dondale (Linyphiidae), respectively.
The accumulation of new provincial/territorial records across Canada has been especially impressive for some jurisdictions (Table 3). Most dramatically, the species count for PE increased by 521% between 2010 and 2018. This was almost entirely because of the combination of data accumulated through CBG's national barcoding project (Blagoev et al. 2016, deWaard et al. 2017) and a "BioBlitz" in 2015 targeting PE's previously poorly known spider fauna (Bowden et al. in press). (BioBlitzes, a relatively new faunistics tool, can be important sources of novel data. For example, in 2014 and 2015 the participation of a team of arachnologists in two BioBlitzes in the small area of Ojibway Park in Windsor ON resulted in the addition of twenty new species records (including six new genus records) to the checklist of Canadian spiders.) The annual intensive surveys of alpine and subalpine habitats in BC begun by the RBCM in 2008 have continued and, along with significant input of specimens and data from a wide variety of collaborators, have resulted in a series of provincial checklists documenting a 25% increase in BC's recorded spider fauna between 2010 and the end of the 2017 field season (Bennett et al. 2012(Bennett et al. , 2014(Bennett et al. , 2017. Spider faunistics in northern Canada benefited from intensive surveys at arctic and subarctic sites by S Loboda and C Buddle (McGill University), R Wisseman (Aquatic Biology Associates, Inc., Corvallis), Perry et al. (2014), the CBG, and others; spider species number increases since 2010 of 71% (124 to 212) and 22% (267 to 325) for LB and NT, respectively, are largely the result of that work. The CBG has also continued to support regional work led by others, such as in BC and PE. This has been mutually beneficial, helping the CBG to complete its library of BINs for Canadian spider taxa and the provinces and territories to achieve a more complete understanding of their regional spider faunas.

Summary (and some predictions for the future)
The rate of increase in Canada's spider species count shows little sign of slowing down. In 2018 the count is well beyond the approximately 1400 predicted by Dondale (1979) and is rapidly closing in on Bennett's (1999) estimate of about 1500. For example, consider the case of BC, which has a longer and more consistent history of spider faunistics research than do other regions of Canada (Table 3) and is the nation's most ecologically diverse province (Morrison and Turner 1994). On average, more than 20 species (primarily linyphiids) have been added annually to the BC regional checklist since 2006 (many of the new records are the result of the RBCM's on-going collaboration with the CBG) and more than 70 of the additions to the BC checklist are of species not previously recorded in Canada and, in some instances, the Nearctic (Bennett et al. 2016(Bennett et al. , 2017. Although the inventory work conducted annually since 2008 by RBCM researchers and their colleagues has resulted in a major increase in the area of BC surveyed for spiders (see fig. 1 in Bennett et al. 2017), habitats in much of the province remain unsurveyed, especially in the province's Boreal Cordillera, Taiga and Boreal Plains, northern Montane Cordillera, and central Pacific Maritime ecozones. Nearly two thirds of all spider species recorded in Canada occur in BC (Table 3). The Nearctic region is home to close to 3800 spider species (Cushing 2018), nearly one quarter of which have been recorded in BC. Clearly, BC is an important region for Nearctic spider diversity and, assuming the RBCM-led work continues, the BC spider species count eventually is likely to exceed 1000, as first predicted by Bennett et al. (2012). Some proportion of those species will be novel Canadian or even Nearctic records. Although other regions of Canada lack BC's complex ecological diversity, with the exception of PE, all have vast amounts of habitat in which the spider fauna has never been surveyed (or only poorly so). Canada's boreal forest, southern hardwood forest, alpine/subalpine, and subarctic tundra habitats are excellent candidates for the discovery of novel regional and national species records given sufficient search effort. As well, provinces sharing borders and particular terrestrial habitats with the United States (e.g., BC's dry southern interior valleys; the southern grasslands of AB, SK, and MB; and southern ON's Carolinian forests and tallgrass prairies) have been the source of novel national species records historically (usually northern range extensions of southern species) and likely will continue to be so in a warming climate.
Finally, DNA barcoding has proven to be an effective complement to classic morphological techniques for identifying spiders and contributing to the advancement of Canadian spider faunistics (Barrett and Hebert 2005, Robinson et al. 2009, Blagoev et al. 2013, 2016, Blagoev and Dondale 2014. At this time the CBG's reference library, BOLD (Ratnasingham and Hebert 2007), contains DNA barcodes for 92% of Canada's 1477 known spider species. Many BINs, appear to reflect cryptic species complexes within well-known taxa (e.g., the example provided by Tetragnatha versicolor) but nearly 60 BINs are associated with Canadian taxa that remain unidentified for a variety of reasons. These have interim, temporary names in BOLD and eventually may be matched to named taxa pending sufficient laboratory detective work or future sequence matching. At least 25 of the BINs, however, are associated with species that are clearly new to science -unnamed and undescribed but morphologically easily discernible. For example, during the period 2014 to 2017 many hundreds of specimens of an unidentifiable but morphologically distinctive species of Perro Tanasevitch (Linyphiidae) were collected at alpine localities in northern BC and southern YT. DNA barcoding and associated taxonomic detective work at the CBG confirmed that this taxon, although morphologically distinctive and common and abundant in relevant habitat, is a Beringian species new to science. Another example is provided by the first Nearctic records of the Palaearctic genus Mughiphantes Saaristo and Tanasevitch (Linyphiidae), an undescribed species initially revealed in 2009 by barcoding of originally unidentifiable specimens from Churchill MB (Blagoev et al. 2013). Subsequent barcoding of further unidentified specimens from other localities has demonstrated that the species is relatively common and widespread in subarctic habitats in AB, BC, LB, MB, NF, QC, and YT.
Future additions to the checklist of Canadian spiders will likely follow the historical precedent shown in Table 2: dominance by linyphiids augmented by new data from difficult groups (e.g., salticids) or emerging from formal taxonomic revisions (e.g., such as occurred with gnaphosids and lycosids between 1979 and 1999). The most recent published estimate for the "actual total" number of spider species in the Canadian fauna (Paquin et al. 2010) is approximately 1720. This is probably a reasonable estimate, given the continuing accumulation of substantial numbers of named species to the checklist of the Canadian spider fauna and of species new to science and awaiting description prior to their addition to the national checklist. However, if most of the excess BINs reflect real but cryptic species, it will not be surprising to see the species count approach or exceed 1800. Tune in again in 20 or 30 years… erine Scott, Janean Sharkey, John Swann, Melissa Todd, Susan Wise-Eagle, Ken White, Charlene Wood, the folks in the Flathead Wild coalition, and our RBCM field work colleagues Richard Hebda, Ken Marr, and Erica Wheeler.
The Royal BC Museum has provided significant financial support for the necessary field work we have been undertaking in the province. Financial support for the International Barcode of Life project was provided by the Ontario Ministry of Research and Innovation, Genome Canada, and the Ontario Genomics Institute. We also thank all our colleagues at the CBG for making the Barcode of Life project work successfully. We are grateful to Parks Canada and Provincial and Territorial Parks staff for supporting collection programs in Canada's national and regional parks.
Finally, we wish to extend special thanks and appreciation to Charlie Dondale and Don Buckle. Charlie led the way for years, putting Canada on the arachnological map of the world, mentoring many past and present arachnologists, and continuing to be active and interested in Canadian spider taxonomy and faunistics decades after his retirement. Don has been the "go-to" person who provided much of the ID work on Canadian spiders for years after Charlie's retirement. Don remains very active as well, and both (perhaps unwittingly) are icons of arachnological socialism (the free, unfettered, and unconditional sharing of arachnological data), a philosophy without which much of what we do would be far more difficult than it is.