Assessment of the DNA barcode libraries for the study of the poorly-known rove beetle (Staphylinidae) fauna of West Siberia

Introduction

Rove beetles (Coleoptera, Staphylinidae) are the second largest family of living organisms (after weevils) (Fig. 1A). They are abundant in the majority of terrestrial habitats globally. Mostly, they are generalised predators or scavengers independent from the single impact of particular factors like, for example, distribution of the host plants or a particular prey species (Herman 2001, Thayer 2016, Betz et al. 2018). Therefore, rove beetles are a good proxy group to explore macroecological and biogeographic patterns. Staphylinids currently comprise nearly 67,000 described species globally (Newton (2022) in the Catalogue of Life at https://www.catalogueoflife.org) and they are always present in good numbers in various biodiversity assessment samples of the ground-based substrates (Betz et al. 2018).

To efficiently use rove beetles or other comparably diverse organismal groups in biodiversity studies, one needs, as a minimum, to quickly and precisely identify multiple species in the large samples. This is a daunting task requiring expensive, time-consuming and, nowadays, rare taxonomic expertise. DNA-based techniques, especially barcoding, now can serve for overcoming such an impediment. In the course of the last two decades, barcoding grew into an important and popular tool of biodiversity exploration and monitoring with sound prospects to increase its impact in the future (Schlick-Steiner et al. 2010, Taberlet et al. 2012, Yu et al. 2012, Liu et al. 2019, Grant et al. 2021). Growth of the DNA barcoding into a multi-purpose research tool led to the formation of the barcode reference libraries. Barcode of Life Data Systems (BOLD; Ratnasingham and Hebert (2007)) is the largest database of this kind that contains more than 12.4 million DNA barcodes (Grant et al. 2021) and is constantly growing. Another is the GenBank database that contains more than 3.5 million COI barcodes (Sayers et al. 2021). Performance and degree of completeness of these databases for various taxa, regions and applications vary and have been subject of investigations (for example, Gaytán et al. (2020), Piemontese et al. (2020), Cheng et al. (2023)).

Here, we want to explore if BOLD and GenBank already gained critical mass of data to study Staphylinidae in West Siberia, i.e. a region where we know that the fauna is poorly known and DNA barcodes were never sampled. In general, we know that, for well investigated regions like, for example, Central Europe or Canada, many beetle specimens can be quickly and reliably identified by its DNA barcode alone already (Hendrich et al. 2014, Gwiazdowski et al. 2015, Schmidt et al. 2015, Hawlitschek et al. 2016, Hebert et al. 2016, Pentinsaari et al. 2019). For the more poorly explored taxonomic groups and regions, i.e. the frontline of biodiversity exploration, we remain far from such an ideal situation, with much to be done. Obviously, the use of the DNA barcoding is less straightforward under conditions of limited data and knowledge. With respect to north temperate Staphylinidae, regions vary from very well-studied like Central and Northern Europe, through reasonably well-explored with knowledge gaps like USA and Canada to poorly known like Siberia. Since many rove beetle species have the latitudinally and longitudinally very extensive distribution ranges across the Holarctic or its sub-regions (Herman 2001), we should expect some overlap amongst local faunas as, for example, is shown by the recently revised genus Quedius (Salnitska and Solodovnikov 2019, Hansen et al. 2022). Thus, under conditions of such an overlap, we want to explore how much the currently available DNA barcode libraries accumulated for regions with better known faunas of Staphylinidae can be applied for the study of their poorly-known faunas.

Study design

We select West Siberia as a target region with poorly-known rove beetle fauna; as well as Fennoscandia in Europe and Canada and Alaska in North America (Fig. 1B), i.e. the regions with, expectedly, very or somewhat similar faunas with West Siberia, respectively. These comparative regions were also chosen because it was relatively easy to generate their comprehensive faunistic lists of Staphylinidae.

West Siberia is a large part of Eurasia with rather clear boundaries defined by topography. From the west, it is outlined by the eastern foothills of the Ural Range, in the south by South Siberian Mountains and Kazakh Uplands, in the east by the Yenisei Ridge and the Central Siberian Plateau and in the north by the Kara Sea. The total area of West Siberia is almost 3 million km², the length from north to south is almost 2500 km, from west to east about 1900 km. West Siberia is a flat lowland spread from tundra in the north through extensive taiga forest to steppe in the south (Gvozdeckiy and Mihalov 1978). Due to its large and hardly accessible terrain, rove beetles and other arthropods of West Siberia are poorly studied. However, we were able to pull together all scattered information on rove beetles of that region for this study, as described below.

Fennoscandia (Ramsay 1898) is outlined by the Baltic Sea in the south, the North Sea in the west, the Norway and the Barents Seas in the North and the White Sea and the East European Plain in the east. The total area is 1.88 million km². Its landscapes vary from plains adjoining the Baltic Sea to mountains adjoining the Norway Sea (Kulikov and Kulikova 2013). Its natural zones span from tundra in the north through boreal to broad-leaved forests in the south (Golubyatnikov and Mammarella 2018). The Fennoscandian biodiversity is intensively studied in many ways (Bjørnstad et al. 1995, Angerbjörn et al. 2001, Kozlov et al. 2022). Due to historical reasons, the Fennoscandian rove beetle fauna (Silfverberg 2011) is amongst the best investigated in Europe (Benick 1934, Hansen et al. 1939, Herman 2001, Hansen et al. 2018).

Contrary to West Siberia and Fennoscandia, Canada and Alaska comprise a much larger area, outlined by the shores of the Atlantic Ocean (Baffin Bay and Labrador Sea) in the east, the Arctic Ocean (Beaufort Sea) in the north and the Pacific Ocean (Bering Strait) in the west (Baulig 1936). Its southern boundary is not defined with the relief, but separated by biogeographic patterns as Canada and Alaska almost match with the Arctic subregion of the Nearctic Region (Escalante et al. 2021). The total area of Canada and Alaska is almost 12 million km² (Hall et al. 2023, Lynch and Miller 2023). Its natural zones stretch from tundra in the north to mixed forests in the south (Larocque et al. 2006, Johansson et al. 2013). It is possible to extract a reliable list of species of rove beetles for this area due to existing literature compilations (Bousquet et al. 2013).

We mine barcodes in BOLD and GenBank for each of these three study regions in order to know: (1) how complete these databases are for their rove beetle faunas; (2) which of both databases is more complete for Staphylinidae as a whole or for the chosen regional faunas, how do they overlap and what is the difference between them; (3) which database is better to use and for which purposes. For seeking answers to these questions we: (1) compile a species list of rove beetles for each of three target regions; (2) compare their faunas based on these species lists; (3) explore both BOLD and GenBank for the presence and diversity of the COI barcodes for the listed species, for all three target geographic regions; (4) explore an overlap and peculiarities of each database in order to eliminate duplication of information and focus on the unique barcodes of each species; and (5) reflect on the observed patterns.

Material and Methods

Results

Discussion and Conclusion

Our data exploration stresses the poor state of knowledge of the Staphylinidae beetles of West Siberia, which is an impediment for using this common and ubiquitous group of the soil macroinvertebrates as a proxy for exploring ground patterns of biodiversity in Eurasia and globally. Contrary to the comprehensive faunal lists of Fennoscandia or Canada and Alaska, both made available by single comprehensive summary resources, the West Siberian faunal list had to be compiled from 27 scattered publications. These publications with reasonably reliable, not outdated species identifications, were filtered from a larger pool of older publications where many identifications were ambiguous. The Palaearctic Catalogue (Schülke and Smetana 2015) is the most significant source that lists 591 species of Staphylinidae for West Siberia, albeit it does not provide the data its records are based on. A high proportion of Staphylinidae species (206 of 726) are known for that region from no more than these catalogue records without any precise baseline data. Most of the georeferenced West Siberian records originate from the southern, better investigated, areas of this region.

Based on the published data, the rove beetle fauna of Canada and Alaska is the largest by the number of species, followed by the fauna of Fennoscandia and then West Siberian fauna. This seems natural as Canada and Alaska cover a territory the largest of all three compared regions. Interestingly, Fennoscandia, an area which is several times smaller than Canada and Alaska, has the rove beetle fauna that is smaller than the fauna of the latter much larger area, only by a few hundred species. This can be explained by the combined effect of two factors. Firstly, the Fennoscandian fauna is much better explored compared to the fauna of Canada and Alaska and, secondly, the European rove beetle fauna is rather species-rich. The number of Staphylinidae species in West Siberia, an area which is geographically larger than Fennoscandia, is about two times smaller than the number of species in the former smaller area. Again, this can be explained by the much poorer degree of our knowledge of the West Siberian fauna, as well as perhaps by the naturally poorer fauna of the latter with more continental and harsh climate and with more homogeneous flat relief. With the future faunal explorations of all study areas, we expect significant increase of rove beetle species for Canada and Alaska and for West Siberia.

Expectedly, the best explored Fennoscandian rove beetle fauna is also the best barcoded compared to other two regions. In fact, it is barcoded to an impressively high degree, where many more than half of the species in the fauna have at least one barcode available from somewhere. One must be aware, however, that even for the well-barcoded Fennoscandian fauna, only nearly a half (45%) of the species in the fauna have barcodes generated from specimens collected within the borders of Fennoscandia. For the much less explored rove beetle fauna of West Siberia, the percentage of species with the available barcodes is even higher than for Fennoscandia, even though none of these barcodes comes from specimens collected in West Siberia itself. We explain this firstly by the high faunal similarity between Fennoscandia and West Siberia and, secondly, by the fact that the faunal list of the much poorer explored fauna of West Siberia is dominated by the common West- or Transpalaearctic widespread species which also occur elsewhere in Europe including Fennoscandia. All barcodes conspecific with species found in West Siberia were generated from specimens collected elsewhere, mainly in Europe. We foresee that further exploration of the West Siberian rove beetle fauna will add a percentage of Asian species for which barcode data are sparse. Even though the percentage of species with the available barcodes in Canada and Alaska is substantially lower than in West Siberia and Fennoscandia, it is noteworthy that barcodes are already available for more than a half of the fauna of this immensely large territory and more than a half of these barcodes are generated from specimens collected there.

BOLD and GenBank contain a number of identical barcodes, i.e. ca. one quarter of their data is doubled between both databases because the same barcode was deposited in both of them. For the majority of their barcodes, however, both databases are unique and, thus, they complement each other. For example, GenBank has approximately two thousand barcodes that appear only there. BOLD, on the contrary, has around eight thousand barcodes that belong only to this database. This could be explained by the BOLD initiatives of collecting barcodes from GenBank and, vice versa, GenBank collects data from BOLD. Most of barcodes relevant for rove beetle species of Fennoscandia and West Siberia come from the DNA barcoding initiatives launched for beetles in Central Europe, especially in Germany (Hendrich et al. 2014, Rulik et al. 2017) or in Northern Europe (Pentinsaari et al. 2014). Some researchers originally upload their data to BOLD (Hendrich et al. 2014, Rulik et al. 2017) and some to GenBank (Pentinsaari et al. 2014). However, the majority, especially in North America, is now leaning to upload barcodes to BOLD, even though some concerns were raised about BOLD not making all its tools publicly available (Meier et al. 2022). Large barcoding projects for beetles of Canada and Alaska (Hebert et al. 2016, deWaard et al. 2019), almost all stimulated by BOLD, were initiated in Canada, a home country for the BOLD headquarters. Naturally, most of the American barcodes are deposited in BOLD.

Substantial overlap in species composition between the West Siberian and Fennoscandian rove beetle faunas and some overlap between northern Eurasian and northern North American faunas, due to numerous more or less widespread species, greatly increases the available pool of barcodes for particular species due to extralimital barcodes. This is especially obvious for the West Siberian rove beetle fauna, for which all available barcodes are extralimital, i.e. mainly coming from specimens collected in Europe. The badly-needed inventory of the West Siberian rove beetle fauna can and should be facilitated by using even such extralimital DNA barcodes to quickly separate potential new species from the known ones. More efficient identification of species via barcoding will include discovery of the hitherto unrecognised molecular variants within the widespread variable morphospecies. It is, in fact, impressive that the available barcode resources can be used for that despite a complete lack of barcodes from West Siberia as such. Noteworthy is the lack of the taxonomic bias amongst the available barcodes. There is no bias towards a more comprehensive barcoding of the larger or otherwise more popular beetles, albeit, for example, the better known and larger, more commonly collected Staphylinidae are indeed relatively well barcoded, while small and taxonomically poorer known are barcoded to a lesser degree. It should be noted, however, that the slow pace of the taxonomic updates of the public barcode databases can cause some gaps or confusion in the research. For example, none of the databases includes Silphinae as a subfamily of Staphylinidae or one database may be more updated than the other with respect to taxonomy. Both GenBank and BOLD should be used as a source of barcode data and data from both sources should be critically checked as far as possible.

Supplementary Material

No conflict of interest to declare

Disclaimer: This article is (co-)authored by any of the Editors-in-Chief, Managing Editors or their deputies in this journal.