Coleoptera of Canada

Abstract The beetle fauna of Canada was assessed, including estimates of yet unreported diversity using information from taxonomists and COI sequence clusters in a BOLD (Barcode of Life Datasystems) COI dataset comprising over 77,000 Canadian records. To date, 8302 species of Coleoptera have been recorded in Canada, a 23% increase from the first assessment in 1979. A total of 639 non-native beetle species have become established in Canada, with most species in the Staphylinidae (153 spp.), Curculionidae (107 spp.), Chrysomelidae (56 spp.) and Carabidae (55 spp.). Based on estimates from the taxonomic community and our BOLD dataset, we estimate that slightly more than 1000 beetle species remain to be reported from Canada, either as new records or undescribed species. Renewed enthusiasm toward and financial support for surveys, especially in the central and western provinces of Canada will be critical for detecting, documenting and describing these species. The Barcode of Life database is still far from comprehensive for Canadian Coleoptera but substantial progress has been made and the number of Barcode Index Numbers (BINs) (as candidate species) has reached nearly 70% of the number of species reported from Canada. Comparison of BINs to observed species in a group of Canadian Staphylinidae suggests that BINs may provide a good estimate of species diversity within the beetles. Histeridae is a diverse family in Canada that is notably underrepresented in BOLD. Families such as Mordellidae, Scraptiidae, Latridiidae, Ptiliidae and Scirtidae are poorly known taxonomically in Canada and are represented in our BOLD dataset by many more BINs than recorded species.

. Census of Coleoptera in Canada. Information sources refer to those available since the publication of Bousquet et al. (2013).

Barcode
Index Number, as defined in Ratnasingham and Hebert (2013).

4-30
Family-level classification in Campbell et al. (1979)  Although our knowledge of Canadian beetle diversity has steadily increased between 1979, 1991, 2013 and 2018, significant contributions can still be made in each province and territory as sampling has been far from exhaustive (for overall estimates of undescribed or unrecorded beetle species, see Table 1). Most biomes in Canada are still only superficially sampled, especially those in central and western Canada. Despite much recent survey work over the past 15 years, more than 300 species were added to the provincial beetle fauna of New Brunswick only two years ago (Webster et al. 2016a). Continued survey work, using a variety of collection techniques, will be necessary for Canada to respond to important changes to its dynamic fauna, such as new invasive species and thermophilic species expanding their range northward in response to global climate change.
In total, 639 non-native beetle species are established in Canada (Table 1), although some of these may eventually be proven to be naturally Holarctic. While a few were introduced intentionally for the biological control of weeds and insects (e.g., De Clerck-Floate and Cárcamo 2011), most have been introduced into North America accidentally through various pathways including dry ballast, wood packing material, and agricultural and horticultural commodities such as stored grain, moss and plant stock (e.g., Klimaszewski and Brunke 2018). The families with the highest number of non-native species in Canada are Staphylinidae (153 spp.), Curculionidae (107 spp.), Chrysomelidae (56 spp.), and Carabidae (55 spp.).
Nineteen beetle families are currently not or only poorly represented in BOLD by Canadian specimens, i.e., the number of BINs is <20% of the number of recorded species in Canada, making it difficult to use barcode data to assess overall taxonomic knowledge (Table 1). These families typically contain few known species in Canada based on published taxonomic data ( Table 1). Sixteen of these families are not represented in BOLD by Canadian specimens: Micromalthidae (Archostemata); Georissidae and Sphaeritidae (Hydrophiloidea); Glaphyridae and Passalidae (Scarabaeoidea); Rhipiceridae (Dascilloidea); Dryopidae and Limnichidae (Byrrhoidea); Nosodendridae (Derodontoidea); Endecatomidae (Bostrichoidea); Biphyllidae (Cleroidea); Prostomidae (Tenebrionoidea); Bothrideridae, Euxestidae, Mycetaeidae, and Teredidae (Coccinelloidea). Efforts are underway to generate DNA barcodes for these families based on Canadian specimens. The family Histeridae, which has more than 130 species in Canada, is particularly underrepresented, with only 22 BINs (i.e., approximately 16% of the known diversity) currently available in BOLD. Most members of this family are small, and live in microhabitats that are not sampled frequently such as mammal and bird nests, or under bark (Bousquet and Laplante 2006). This lack of representation could also be partly due to sequencing bias against Histeridae resulting from primer mismatch, or differences in DNA preservation at the collecting and archiving stages. For example, only 3% (8/256) of a diverse sample of Histeridae specimens from the Canadian National Collection of Insects, Arachnids and Nematodes (CNC, Agriculture and Agri-Food Canada) yielded barcode-compliant (and therefore BINcompatible) sequences, versus 22% of submitted CNC Staphylinidae (522/2356).
Based on the number of BINs in BOLD for Canadian specimens, sixteen beetle families are more diverse in Canada than would appear from the recorded number of species (Table 1). The families where the number of BINs most greatly exceeds the number of species reported in Canada are: Mordellidae (+32 BINs) and Scraptiidae (+28 BINs) (Tenebrionoidea), Latridiidae (+23 BINs) (Coccinelloidea), Ptiliidae (+21 BINs) (Staphylinoidea), and Scirtidae (+21 BINs) (Scirtoidea). These families, generally with poorly known and small-sized species, require focused taxonomic studies because they may contain many undescribed species or described species yet unreported from Canada. This work should reconcile the unidentified BIN clusters with available names and describe any species new to science to adequately document the Canadian fauna. Researchers at the CNC and the Canadian Museum of Nature have made numerous contributions to the knowledge of Canadian Coleoptera. However, because most federal employees in Canada focus their research on agriculturally-significant taxa (see Bouchard et al. (2017) for plant-feeding taxa with high economic concern in Canada and in agroecosystems of our trading partners), beetle groups without either plant pests or well-known beneficial species have been given a lower taxonomic research priority. Canadian universities have until recently included taxonomic research on non-economically important beetles, although they currently support a minute fraction of research on Canadian Coleoptera.
The total estimated number of undescribed and unreported beetle species for Canada is 1080 to 1280 species (Table 1) based on expert estimates and species predicted by BINs including Canadian specimens in BOLD. The beetle families with the greatest number of taxonomist-estimated unrecognised diversity in Canada include the Staphylinidae, Carabidae, Ptiliidae, Curculionidae and Chrysomelidae, most of which include either plant pests or beneficial predators and parasitoids. These numbers represent the best available estimate of unrecorded Coleoptera diversity, although they must be interpreted with respect to limitations of expert opinion, BOLD database sampling, potential inaccuracies of the most current checklist (Bousquet et al. 2013), and BIN calculation methods. While we expect the exact numbers to change with further taxonomic research, the general trends reported herein should not.

Reconciling BINs with morphological data -a Canadian example
Barcode reference libraries for beetles and bees in taxonomically well-studied Central Europe (Hendrich et al. 2014, Schmidt et al. 2015 show that most BIN clusters are highly congruent with taxa already recognised by science. There, most discordance between BINs and recognised taxa likely reflects cases of unrecognised species diversity or species pairs with very similar COI sequences that, while considered one BIN due to shallow divergence, still clustered into species (Hendrich et al. 2014). However, it will be important to assess whether BINs closely approximate real taxa in other regions. We anticipate that barcoding will be similarly effective for characterising the Canadian fauna since both regions are climatologically and topographically similar with shared glaciation history, and with many beetle genera in common.
One DNA barcoding-related discovery is a cryptic species of burying beetle (Silphidae: Nicrophorus Fabricius) that was discovered in North America based on congruent evidence from ecological data, mating studies, morphology and DNA barcode data (Sikes et al. 2016). Nicrophorus is considered taxonomically very well studied in North America but the cryptic lineage was first highlighted by a different BIN than specimens from the Palaearctic and Alaska (Sikes et al. 2016). We anticipate that the taxonomic integration of DNA barcode data will provide many other insights about the Canadian fauna. The pressing need to carefully and authoritatively link Linnaean taxonomy with molecular reference databases such as BOLD through taxonomic research was stressed by Somervuo et al. (2017) and is re-emphasised here. This need was recently recognised in Canada, with over 5000 beetle species (summarised by Bouchard et al. 2017) added to BOLD in recent years. Although a general, species-focused analysis of the Canadian beetle fauna (similar to Hendrich et al. 2014) is premature, it is possible to examine the congruence of BINs with the taxonomy of a group of well-revised but diverse beetles.
We can partially test BIN congruence using the subtribe Quediina (sensu Brunke et al. 2016), a diverse lineage of rove beetles (Staphylinidae) and the subject of modern taxonomic revision in North America, including critical examination of male genitalia for species concepts (Smetana 1971a(Smetana , b, 1973(Smetana , 1976(Smetana , 1978(Smetana , 1981(Smetana , 1990. They are generalist predators, may be important predators of pest insects and are often abundant in decaying organic matter (Smetana 1971a). Currently, 64 species of Quedius and Quedionuchus are recorded from Canada (Bousquet et al. 2013) and of these, 42 (66%) are represented in BOLD by sequences of authoritatively identified specimens. A total of 52 BINs represent Canadian Quediina in BOLD and most BIN incongruence with existing taxonomy is due to unrecognised species diversity. Although four 'well-known' species are currently considered Holarctic in distribution, Nearctic specimens form separate BINs from their Palaearctic counterparts in three of these. One Canadian Quedius has two traditionally recognised subspecies for which BINs are 6% divergent and will likely be considered morphologically diagnosable species. Another four BINs correspond to still unidentified species and further work is needed to discern whether they belong to described or undescribed taxa. Two Palaearctic Quedius species appear present but unverified and unreported from Canada. Four cases exist where a valid species of Quedius contains two BINs that do not correspond to morphological differences. In these cases, BOLD may have oversplit species due to algorithm artifacts based on material limited in number and geographic coverage, and BINs may be later combined in BOLD when additional sequences are included. Taxonomic research involving these putative lineages, including study of type specimens for available names, is in progress. No described species of Quediina shared BINs with any other species, indicating that the species are not 'oversplit', likely due to informative variation in male genitalia. Thus, only 7.7% of BINs were incongruent with species level clusters after cases of unrecognised species diversity were removed (a further 11.5%). A similar result was found for the well-studied Quediina of Central Europe where similar diversity (71 species, Assing and Schülke 2012) was represented by about 51% coverage (36 species with 39 BINs) in BOLD and only 7.7% of BINs (involving two species) were incongruent with prevailing species concepts (Hendrich et al. 2014).
The utility of BOLD as a proxy for biodiversity should be demonstrated over a broader taxonomic and geographic scale (Bergsten et al. 2012), as the BOLD database is highly skewed toward Canadian specimens. However, it is promising that gaps in the variation of male genitalia of rove beetles, typically used by taxonomists, correspond remarkably well to gaps in sequence variation identified by BINs. This suggests that BINs may provide a proxy for beetle diversity in North America and could be useful for highlighting taxonomic groups needing research attention (as done above by family).

Future priorities
The number of known species from Canada will continue to increase as new species are described, new populations of described species are discovered and species arrive as a result of climate change or global trade. While further taxonomic work on Staphylinoidea, Cucujoidea, Chrysomeloidea and Curculionoidea will continue to add many species to the Canadian fauna, new biosystematics work on several poorly studied families (e.g., Mordellidae, Scraptiidae, Latridiidae, Ptiliidae, and Scirtidae) is greatly needed. Although recent collecting in eastern Canada has yielded many discoveries, these biomes and, especially, those of central and western Canadian provinces remain inadequately sampled. A renewed effort toward exploring the Canadian beetle fauna will be critical for the documentation of the more than 1000 unrecorded or undescribed species that are estimated to be undetected or undescribed in Canada. Since DNA barcoding is a useful tool for assessing species diversity and appears to be highly compatible and synergistic with traditional morphological taxonomy, knowledge of beetle diversity in Canada will further benefit from continued development of the DNA barcode library through focused collecting, authoritative vouchering and continued integrative taxonomic research. However, improved and continued documentation of the Canadian fauna can only be achieved with new funding for surveys, including a variety of sampling methods, and by hiring or otherwise supporting scientists that include taxonomic work on the Canadian fauna as part of their research profile.