A nearly complete database on the records and ecology of the rarest boreal tiger moth from 1840s to 2020

Global environmental changes may cause dramatic insect declines but over century-long time series of certain species’ records are rarely available for scientific research. The Menetries’ Tiger Moth (Arctia menetriesii) appears to be the most enigmatic example among boreal insects. Although it occurs throughout the entire Eurasian taiga biome, it is so rare that less than 100 specimens were recorded since its original description in 1846. Here, we present the database, which contains nearly all available information on the species’ records collected from 1840s to 2020. The data on A. menetriesii records (N = 78) through geographic regions, environments, and different timeframes are compiled and unified. The database may serve as the basis for a wide array of future research such as the distribution modeling and predictions of range shifts under climate changes. It represents a unique example of a more than century-long dataset of distributional, ecological, and phenological data designed for an exceptionally rare but widespread boreal insect, which primarily occurs in hard-to-reach, uninhabited areas of Eurasia.

The Menetries' Tiger Moth Arctia menetriesii (Eversmann, 1846) may be considered the rarest and most enigmatic species among boreal representatives of the tribe Arctiini 32,33 . This species has a continuous range expanding throughout the entire Eurasian taiga biome from Finland to the Sakhalin Island and northeastern China 26,32,34 (Fig. 1a). Despite its enormous range, Menetries' Tiger Moth occurs very rarely, and less than 100 specimens were recorded since its initial description in 1846 32,35,36 (Fig. 2a-e). It could primarily be found in hard-to-reach, uninhabited areas of the continent, harboring primeval woodlands 32 (Fig. 1b-d). Earlier, this species was placed within its own monotypic genus, Borearctia Dubatolov, 1984 33 but it was recently transferred to the genus Arctia Schranck, 1802 based on a comprehensive multi-locus phylogenetic research 14 . Host plants and life history of A. menetriesii were discussed in a few works [36][37][38][39] and are rather poorly known.
The holotype specimen of this species (Fig. 2c,d) is imprecisely labelled as being collected from a broad historical region, the Songoria (currently eastern Kazakhstan; most likely somewhere within the Tarbagatai or Saur Mountains) 33,35,40,41 . Next records come from Finland 37,42,43 , Yakutia 44 , North Manchuria 34 , and Sakhalin 34 in the 1910s-1920s. Kurentzov 44 presented the first range map of A. menetriesii with eight records known to date (1965). Later, several authors listed additional records from the Urals and Asiatic Russia 28,32,33,39,[45][46][47][48] . The species was considered extinct in Finland 49 but was rediscovered in 2003 50,51 . Dubatolov 26 published a more detailed  52 . The size of circles indicates the uncertainty of the geographic co-ordinates (see Legend). The color areas represent three regions discussed in this study as follows: yellow Europe (records from Finland, Northern European Russia, and the Urals); red Siberia (records from Western and Eastern Siberia in Russia, and East Kazakhstan); and green the Far East (records from the Russian Far East and northeastern China). The map was created using ESRI ArcGIS 10 software (https://www.esri.com/arcgis); the topographic base of the map was created with Natural Earth Free Vector and Raster Map Data (https://www.naturalearthdata.com) and Global Self-consistent Hierarchical High-resolution Geography, GSHHG v. range map with 40 records (37 precise points and three vague localities) but, unfortunately, any data on certain records has not been listed. Our team started to compile information of A. menetriesii records in the early 1990s. The first georeferenced dataset was published in 2013 32 . This compilation contained the data on 43 records, including three samples from vague localities such as Lake Baikal 32 .
The present study describes our final compilation, the Menetries' Tiger Moth Range and Ecology Database (1840s-2020) 52 , which contains nearly all available information on this extremely rare species collected since its original description (Fig. 3). In total, the database comprises geographic, environmental, and temporal information on 78 records of A. menetriesii (Online-only Table 1, Figs. 4-6). Here, we consider individual specimens as separate records with unique IDs, even if these specimens were collected simultaneously from the same locality. All the records were delineated into three larger geographic regions: Europe (Finland, Northern European Russia, and the Urals), Siberia (Western and Eastern Siberia in Russia, and East Kazakhstan), and the Far East (Russian Far East and northeastern China) (Fig. 1a). Available images of habitats and samples are linked to the database via file names (examples: Figs. 1b-d and 2b-e). The database is a unique example of a more than century-long dataset of distributional, ecological, and phenological data designed for a rare boreal moth, having a broad trans-Eurasian range.
The present work was also motivated by a broader aim linked to historical and recent rapid defaunation processes on Earth 53,54 and, more precisely, to the so-called insect apocalypse in the Anthropocene 2,3,5 . A plethora of recent surveys revealed that there is a drastic decline in insect (and moths in particular) abundance and biomass over the past few decades 11,55,56 . The causal mechanisms and drivers of this phenomenon are poorly understood [57][58][59] but, at first glance, could be linked to agricultural intensification, pesticide use, urbanization, habitat loss, and climate warming 3,60,61 . Conversely, it was shown that many insect species show no declines and that the most dramatic examples of declines come from highly populated agricultural areas of the Northern Hemisphere 10,62 .
Based on these considerations, A. menetriesii may be used as a model organism to address the long-term distribution and abundance shifts in larger boreal insects, having a broad trans-Eurasian range. Hence, the Menetries' Tiger Moth Range and Ecology Database (1840s-2020) 52 could serve as the reliable basis for a variety of future research such as distributional and phenological modeling and predictions of range shifts under various climate change scenarios.

Methods
Original field research. The authors of this paper were looking for A. menetriesii specimens during their fieldworks starting from the early 1990s (ca. 30  www.nature.com/scientificdata www.nature.com/scientificdata/ by hands or by an entomological net. Here, we present eight unpublished records collected by ourselves. Additionally, a few own records were published in our earlier works 28 Second, we searched for published records of A. menetriesii through online tools such as Web of Science, Scopus, and Google Scholar using the scientific name of this species (in combination with both generic names: Borearctia and Arctia) and its English, Finnish, and Russian common names ("Menetries' Tiger Moth", "idänsiilikäs", and "медведица Менетрие", respectively) as keywords. Additional data was collected from references cited in the works we have found. None of the public records was found in the iNaturalist online portal (https:// www.inaturalist.org).
We consider that the Menetries' Tiger Moth Range and Ecology Database (1840s-2020) 52 represents a nearly complete resource on available records of A. menetriesii. Perhaps, a few additional specimens may be found in museum and private collections in China (not checked by us), and some singletons in private collections of www.nature.com/scientificdata www.nature.com/scientificdata/ Russian and European amateurs may have been overlooked. However, we estimate that the amount of records overlooked by us (if any) may not be higher than 8-12 (10-15% of the total number of records in our database) due to an exceptional rarity of this species. Data processing. The collecting localities were georeferenced and verified using the Google Earth tool (https://www.google.com/intl/ru/earth). The same tool was used to assess the altitude of each locality, which was rounded to the nearest tenth. The altitude of mountain localities was additionally checked using available topographic maps and the ACE2 global digital elevation model 63 . More than half of the localities were precisely georeferenced (±10…500 m; N = 50; 64.1% of the total number of records). Other part of the localities share less precise co-ordinates (±1…10 km; N = 22; 28.2% of the total number of records). Four records (AM-075, AM-076, AM-077, and AM-078) were ascribed for vague localities 34,35 , and their co-ordinates listed in the database are approximate (±100…200 km). Finally, two records (AM-036 and AM-053) were left without co-ordinates because their localities were too vague for referencing purposes (the entire Lake Baikal and an unknown mountain in Yakutia, respectively).
Unfortunately, the holotype specimen of A. menetriesii (AM-075) is also among the group of rather uncertain records. The label of the holotype reads as "Songoria" (Fig. 2b). The original description states that "Haec Eupreria, ab illustrissimo D. Ménétriés ad describendum mihi communicata, campos Songariae inhabitat" 35 . In this case, Songoria most likely means the so-called Songoria rossica [=Russian Dzhungaria; currently East Kazakhstan]. This region contains desert areas and the Tarbagatai and Saur mountain ranges. Based on available data on the habitat spectrum of this species, it became clear that the holotype was collected somewhere in the mountains. www.nature.com/scientificdata www.nature.com/scientificdata/ The Tarbagatai Mountains as the most probable type locality of the species first appeared in the Palearctic Lepidoptera catalogue of Staudinger and Rebel in 1901 but with a question mark 40 . Subsequent researchers in Finland and Russia followed the catalogue, placing the type locality within that mountain range 37,41,42,64 . Later, Dubatolov assumed that the type locality might be somewhere within the Saur Range 33 , which represents the eastern extremity of the Tarbagatai. In a subsequent work, Dubatolov listed the type locality as follows: "Mountains around Lake Zaysan, East Kazakhstan Region, Kazakhstan" 65 . This large lake is situated between the Southern Altai and Tarbagatai mountains. As a conclusion, the type locality of A. menetriesii cannot be identified with certainty but it is most likely somewhere within the Tarbagatai -Saur mountain system in East Kazakhstan. We therefore placed it in the middle of this range but with a high uncertainty level (±200 km).
The WWF's Global 200 Project established a system of Earth's ecoregions and biomes 67,68 , which is widely used in biogeographic, ecological, and conservation surveys 69,70 . We therefore placed each georeferenced locality within these categories using ESRI ArcGIS 10 software (https://www.esri.com/arcgis).
All types of habitat were converged into five larger categories: mountain forest; plain forest; riparian forest (forest patches in river valley, stream valley or on lake shore); alpine meadows and tundra (open alpine habitats); and town (urban environments). For example, two specimens (AM-006 and AM-007) were collected in a half-open bog site surrounded by coniferous forest in Kuhmo, Finland. This habitat was www.nature.com/scientificdata www.nature.com/scientificdata/ assigned to the plain forest category. In each case, the original information on habitat and available data on plant cover at the collecting site could be seen in the "SAMPLE_DATA" field 52 .
The Julian calendar date of a historical sample (AM-043) collected in Russia before 1918 was changed to the Gregorian calendar (actual) date but both the historical and the actual date could be seen in the "SAMPLE_ DATA" field 52 .

Data Records
The Menetries' Tiger Moth Range and Ecology Database (1840s-2020) can be downloaded from figshare 52 . The main database is presented in XLSX format (ArcMenDB_1840_2020.xlsx). The corresponding reference list and museum collection codes are available as separate PDF files (ArcMenDB_RefList.pdf and ArcMenDB_ Museum_ID.pdf, respectively). These files and the ZIP-archive with images (ArcMenDB_Images.zip) can be downloaded together with the database.
Each field name of the database is decoded and explained in Online-only Table 1. The database structure is illustrated in Fig. 3. A unique code (RECOR_ID) is assigned for each record listed in the database. Available label and field observation data on each sample of A. menetriesii is presented in the "SAMPLE_DATA" column (as a brief description). The sources of information for a given specimen are cited in the "REFS" (=References) column, while the full references for each record are listed in the "FULL_REFS" column. The complete list of references can also be downloaded as a separate PDF file (see above).
Information in the database is clustered into three large blocks: "Environment", "Timeframe", and "Biology" (Fig. 3). The "Environment" block contains geographic and environmental data on collecting localities such as the co-ordinates, their uncertainty, altitude, region, country, ecoregions (Bailey's and WWF Global 200), habitat type, and the presence/absence of a waterbody. Available habitat images (format: JPEG and TIFF) are linked to this block via file names. The "Timeframe" block presents information on the day, ten-day-period, month, and year of a given record, as well as in which (odd or even) year it was collected. Finally, the "Biology" block contains information on a given specimen as follows: developmental stage (larva, imago, etc.), sex (imaginal records only), and condition (living or died). Available specimen images (both living and collection specimens; format: JPEG and TIFF) are linked to this block via file names.
Each parameter, factor levels, and linked data used in the database are described in Online-only Table 1. Furthermore, in this table we list the number of records, for which this parameter/factor/linked data is available. In total, the database contains 78 records but some of them lack certain data points as follows: co-ordinates (2), altitude (6), habitat (7), presence/absence of a waterbody (8), Bailey's Ecoregion (2), The Global 200 Biome/ Ecoregion (2), collecting day (17), ten-day period (14), month (11), year (7), sex (20), and condition (3). The habitat and specimen images are available for 10 and 29 records, respectively.
The geographic coverage of the database can be seen in Fig. 4a,b. The color areas on this figure indicate the approximate level of probability to find A. menetriesii specimens under certain geographic conditions based on the spline interpolation of the number of collected individuals per locality. The set of available georeferenced localities expands from 42.4°N to 71.9°N by latitude, from 24.4°E to 143.3°E by longitude, and from 10 to 1740 m by altitude. From the majority of the localities, singleton specimens were collected. However, there are eight localities with two or three recorded specimens in each, indicating the presence of permanent populations in these sites (one in Europe and seven in Siberia and the Far East; Fig. 4b). The environmental coverage of the database in relation to the altitude of collecting localities and the number of collected individuals is visualized in Figs. 5-6. Most non-singleton samples were collected from lowland to upland riparian forests of Siberia and the Far East between 0 and 1,200 m altitude (Fig. 5). Furthermore, the vast majority of specimens, including almost all non-singleton records, was sampled near a river or stream (Fig. 6).

Technical Validation
Experienced entomologists verified all the records and ecological data included to the database. Furthermore, we used only records that were based on collected specimens. None of the doubtful records such as those based solely on visual observations has been included. We cited and list corresponding references for records obtained from reliable published sources, while the museum storage was listed for every collection sample, if available. An uncertainty of geographic co-ordinates for each locality was estimated. Finally, all the authors carefully checked the complete database for possible technical failures and errors.

code availability
No code was used in this study.