Distribution of Finnish mound-building Formica ants (Hymenoptera: Formicidae) based on using a citizen science approach

The distance from southern Italy to Denmark is about the same as the length of Finland from south to north. A study of the biogeography of insects, such as ants, would take a lot of effort and funding to sample the whole area. Here, a citizen science approach is used to obtain distribution records for mound-building Formica ants in Finland. This resulted in samples from 2,434 ant nests, of which 2,363 were for nests of the target species group. The data obtained helps defi ne the northern limits of the species in Finland and resulted in three new records for F. suecica Adlerz, 1902, which is a red-listed species in Finland. In addition, as a by-catch, a new imported species dispersed in a peculiar way was recorded in Finland: Lasius emarginatus (Olivier, 1792). Volunteer citizens are potential research assistants in the science of entomology.


INTRODUCTION
The mound-building Formica ants are important key species in their environment. They are effective predators and also have a role in nutrient recycling (e.g., reviews in Robinson et al., 2016 andFrouz et al., 2016). In addition, their nests are coinhabited by a multitude of other arthropods, such as beetles, spiders and oribatid mites (e.g., Päivinen et al., 2002;Härkönen & Sorvari, 2014;Elo et al., 2016;Robinson et al., 2016).
Mound-building Formica ants make nest mounds or cover their nests with organic material, mainly needles, small twigs and leaves of shrubs. The distributions of these ants in Finland were studied earlier by Baroni Urbani & Collingwood (1977) and Collingwood (1979) using geographical and administrative units. More recently, Punttila & Kilpeläinen (2009) report the distribution of these species based on data collected during the tenth Finnish National Forest Inventory (NFI10, National Resources Institute Finland). However, the species distributions in four Finish large-scale ecoregions; in addition, the NFI10 data exclude northern Finnish Lapland (approximately 1/5 of Finland), thus, leaving one of the most interesting geographic areas unstudied. The NFI10 sampling targeted forest and forested mire plots, thus leaving other environments unstudied, therefore missing the main habitats for some mound-building species, e.g., the meadow-dwelling Formica pressilabris Nylander, 1846.
In order to obtain distribution data from all over Finland and from all kinds of terrestrial habitats, a citizen science Eur. J. Entomol. 118: 57-62, 2021 doi: 10.14411/eje.2021.007

ORIGINAL ARTICLE
located in at least a 10 × 10 km square, usually well within 100 × 100 m. All but two samples could be identifi ed to species, with the two being most likely hybrids between F. aquilonia Yarrow, 1955 andF. polyctena (Czechowski, 1996;Sorvari, 2006;Kulmuni et al., 2010). Approximately one thousand people (ca. 0.018% of the Finnish population) participated in this project, mainly private citizens, but a large proportion of samples were collected by groups such as day care centre groups, school groups, scouts, etc. Biology teachers especially welcomed this project, and some even gave credit units for a biology course to those students who participated in the survey. As a bonus, many of the samples were accompanied by a story or other information and photographs. Reciprocally, the participants received a document that contained a summary of the results and information about the species.
The participants in the survey each sent approximately 2.4 colony samples, mostly one-to-two samples per participant, but it was not uncommon to receive e.g., 20 samples from a single participant. The record was an impressive 279 samples from different colonies from one enthusiastic bird ringer. Formica s. str. (Fig. 2) The distribution of this subgenus is clearly divided into those nesting all over Finland, F. aquilonia, F. lugubris and F. truncorum, and those that have a southern distribution, F. rufa and F. pratensis. While there are a few records from northern Finland, Formica polyctena generally seems to be mostly southern in its distribution. While Punttila and Kilpeläinen (2009) report a northern distribution for F. polyctena, but its exact northern range cannot be estimated based on the information provided. The northernmost record for F. polyctena in Finland, however, is included in the present citizen science data. The northern range of F. rufa and F. pratensis is a bit more northern based on the present data than that reported in previous studies (Baroni Urbani & Collingwood, 1977;Collingwood, 1979;Punttila & Kilpeläinen, 2009).

Coptoformica (Fig. 3)
Formica exsecta is clearly the most common and most widely distributed species of this subgenus. It is highly var-tifi ed to species-level using a Wild M5A stereomicroscope and the identifi cation keys of Czechowski et al. (2012) and Seifert (2018). After identifi cation, 50% ethanol was replaced with 90% ethanol. All specimens and original forms are deposited in the author's private collection.
The coordinate system used by the collectors varied, being mostly variants of WGS84 and KKJ (an outdated Finnish coordinate system). All coordinates were transformed into EUREFbased ETRS-TM35FIN coordinates using the retkikartta.fi web map platform (National Land Survey of Finland). The maps showing the locations where the samples were collected were prepared using QGIS 3.14 software.

RESULTS
The project received samples from 2,434 nests and a total 30,674 specimens of ants (for a map showing the locations of all the samples see Fig. 1). Of the samples, 2,363 were of the target species group, organic mound-building Formica ants of the subgenera Formica s. str., Coptoformica, Raptiformica and F. uralensis (Table 1). One sample of Formica polyctena Förster, 1850 did not contain any information on the location, whereas all others could be  iable in pilosity and colouration; generally, there is a hairy normal morph and a less hairy and more reddish 'Rubens' morph (Seifert, 2018). Seifert (2000) describes a new separate species, F. fennica Seifert, 2000, a meadow-dwelling species, which is almost identical to the mire-dwelling 'Rubens' morph (Ødegaard, 2013;Seifert, 2018). This 'triplet' forms a taxonomic problem, especially because a recent genetic study does not support the Finnish population of F. fennica being a separate entity from F. exsecta (Hakala et al., 2018). While in the current citizen science data the majority of the samples of F. exsecta were of the normal morph, several samples were of a reduced pilosity typical of 'Rubens' and possibly F. fennica. Here, they are treated collectively as F. exsecta.  (Baroni Urbani & Collingwood, 1977;Collingwood 1979). The other species are reported all over Finland.
Formica suecica is a red-listed species in Finland (Paukkunen et al., 2019). The survey resulted in three new records for this species, all from Lapland and one of which was well north of its known range. Formica pressilabris and F. forsslundi are habitat specialists, the former of meadows and the latter of almost open mire habitats. The distribution of F. pressilabris revealed by this study fi ts its known range, but that of F. forsslundi is well north of its known range. F. uralensis (Fig. 3) Formica sanguinea is mentioned as common throughout Finland (Collingwood, 1979). However, the survey result-  (Baroni Urbani & Collingwood, 1977;Collingwood 1979). The other species are reported all over Finland. ed in only 14 records, possibly due to its inconspicuous and irregular nests that are often located in tree stumps, under stones or in rock crevices. It covers its nests with shrub leaves rather than needles and sometimes they are not covered with organic matter.

Other ants
The survey resulted in an additional 81 records of 16 non-target species from the genera Formica (subgenus Serviformica), Lasius, Camponotus and Myrmica (Table  2). One sample was a queen of Lasius emarginatus (Olivier, 1792), which has not been previously reported in Finland. However, this is a result of a peculiar introductory event. A citizen had bought a box of peaches from a supermarket in Pori, a town in southwestern Finland, and noticed at home that one of the peaches was in a bad condition. He opened it, and the queen ant came out of the peach stone. Unfortunately, the label on the peach box was not saved and, therefore, the exact origin of this ant is unknown, but peaches sold in Finland mainly come from southern Europe and Turkey (FAO, 2020). This species' native range is Central and Southern Europe and parts of Turkey and the Caucasus (Seifert, 2018). Notably, this species has recently expanded its northern range, especially in Western Europe (Seifert, 2018).

DISCUSSION
Citizen science and/or crowdsourcing is a growing method for obtaining data. This is especially benefi cial in zoogeography since researcher-based sampling of large geographical areas would be time-consuming, very laborious and costly. There are already a few successful citizen science projects on the distribution of species, e.g., Ixodes tics, baboon spiders and bumblebees (Laaksonen et al., 2017;Campbell & Engelbrecht, 2018;Falk et al., 2019).
The citizen science approach was successful in this study of mound-building Formica ants, as a high number of samples was received and with a very low budget. There were no costs to the researcher for collecting the samples and the cost for storing samples in alcohol-containing screwcap tubes and cardboard storage boxes was below 1,500 €.
The survey records complemented earlier distribution records in a way that enable more precise estimates of distributions of species in Finland. The project received three new records of the nationally red-listed Near Threatened (NT) species F. suecica. It is noteworthy that they all came from Lapland, which could mean that the species is more common in the north than in the south of Finland. However, the species is rare, and the current data is based on only three records. Citizen science data can thus result in new records of endangered species of ants, but also of taxa other than ants.
Of course, citizen sampling is not free of errors. In this case, meadow-dwelling species such as F. pressilabris and mire-dwelling species such as F. uralensis and F. forsslundi can be underrepresented since people probably tend to seek mound nests in forested environments as only ca. 2% of samples were from meadows and 5% from mires (Sorvari, unpubl. data); thus, the citizen science data did not fully resolve the same habitat bias as that in the NFI10 data (Punttila & Kilpeläinen, 2009). Another source of error may be the size of the nests with the large mounds of the F. rufa species group, especially F. aquilonia and F. polyctena, being more attractive to citizen scientists and thus possibly increasing results for these species at the expense of other species with less conspicuous nests, e.g., F. forsslundi (making small fl at nests in mires) and F. sanguinea (making irregular shrub leaf-made nests or nests without any organic material cover).
Regardless of the potential biases, the data contained valuable information on species' distributions and provides over 30,000 individual specimens for further study. In addition, the survey can be repeated in the future to detect any northward shift in range of the southern species by comparing the results with existing data.