Murmansk Region is a top-level federal subject of the Russian Federation, situated in the north-western part of European Russia; this territory is also known as Russian Lapland in historical literature. It is largely situated on the Kola Peninsula, surrounded by the Barents Sea in the north and by the White Sea in the south. Its total area constitutes 144,902 km². This territory is part of the Subarctic Zone; its vegetation is represented by tundra in the north, forest tundra in the major part of the mainland, and northern taiga in the south-west, next to the borders with Finland and Russian Karelia (Chernov 1971). The relief is largely flat and nearly monotonous, except for rocky mountain groups in the western part of the territory, among which the Khibiny and Lovozero Mts. are the highest to reach the maximum of 1191 m above sea level (Fig. 1A). Smaller but important hills and isolated outcrops are situated in the Kovdor and Kandalaksha Districts, whereas the north-western coast is incised by fjords and the other coastal areas are traversed by deeper river valleys. Forested areas are extensive in the western part, and forests are significantly present in the basins of the Ponoi River and the rivers flowing into the southern coastal waters of the White Sea (Fig. 1B).

Figure 1. 

Study area: Murmansk Region, Russia A hypsometric map, major rivers and lakes B distribution of forested landscapes (denoted by green colour). Maps were created using ArcGIS software by Esri. ArcGIS is the intellectual property of Esri and is used herein under license. Copyright Esri. All rights reserved.

This study was based on a comprehensive sampling of all herbarium specimens available from the study area and kept at H, INEP, KAND, KPABG, LE, LECB, MW, OULU, PTZ, TROM (herbarium acronyms according to Index Herbariorum (Thiers 2023)) and the Herbarium of the Apatity Branch of the Murmansk Arctic University (unregistered, provisional acronym ARCT). Documented observations (iNaturalist 2023) were also used. The specimens (258) and observations (3) were georeferenced and databased, and their data were made available as a taxonomic dataset (Sennikov and Kozhin 2023), which includes 261 herbarium specimens and documented observations altogether. Point distribution maps were generated in ArcGis 10.3.1 (https://www.esri.com) from the database, taking into account the residence status and period of introduction of the species in every locality; we distinguished three categories of species residence: native, pre-industrial alien (introduced prior to the industrialisation in the USSR, i.e. before the 1930s) and industrial alien (introduced with or after the industrialisation in the USSR, sometimes after 1930 but usually after 1960).

The herbarium specimens were examined taxonomically by A. Sennikov either de visu (H, OULU), or as high-quality scanned images (ARCT, INEP, KAND, KPABG, LE, MW), or as low-quality scanned images (TROM), or as photographs (LECB, PTZ). Scanned images were also used as illustrations.

Special effort was used to trace validly published names at the level of species and subspecies, in order to produce a stable synonymy whenever the species or subspecies rank is preferred. More precise publication dates were traced from a variety of bibliographic sources, which helped to establish the sequence of publication and to ensure priority in difficult cases.

As in the previous revisions (Tzvelev 1994; Kurtto and Väre 1998; Šída 1998, 2000), we examined the following diagnostic characters: stem and phyllary colouration; stem, leaf and phyllary pubescence; length and density of pubescence; number, shape and density of cauline leaves; shape of synflorescence; ligulate flower colour. The value of individual characters may be limited due to high genetic variability or phenotypic plasticity, whereas a complex of characters may unambiguously characterise the taxa.

In this work, we used a classical method of morphological comparisons, observing discontinuities in plant variability. We did not use statistical methods (like employed by Olander and Tyler (2017)) because, in order to distinguish between very closely related taxa, infra- and interpopulational variability must be considered separately, whereas our material was represented by previously collected herbarium specimens, i.e. a random selection from multiple populations that do not allow to assess their interpopulational variability.

We follow the concept of narrowly defined taxa in the Erigeron acris group, which is universally adopted nowadays (Halliday 1976; Tzvelev 1994; Kurtto and Väre 1998; Šída 1998, 2000, 2004; Greuter 2006; Olander and Tyler 2017; PoWO 2023). This concept is justified because of the limited variability and morphological distinction of the accepted taxa. Ranking in this group is disputable; species level is currently accepted in the Czech Republic (Šída 1998, 2000, 2004) and Eastern Europe (Tzvelev 1994), whereas subspecies level is preferred in Finland (Kurtto and Väre 1998) and Sweden (Olander and Tyler 2017), and in European (Halliday 1976; Greuter 2006) and global (PoWO 2023) compilations. We accept species rank because the ranked taxa are clearly defined by morphology and commonly co-occur in the same territory with limited hybridisation, thus complying with the biological species concept due to apparent reproductive isolation (Gao and Rieseberg 2020).

In agreement with Pyšek et al. (2002), origin status and invasion status were determined for each accepted species, and residence status was determined for each recorded locality. We determined the origin status as native or alien; a taxon was treated as alien if it arrived to the territory with human assistance at any time period, including the remote past from which no historical plant records or other direct evidence are available. Each non-native species was classified according to its invasion status either as casual or naturalised, with further estimation of invasiveness (Richardson et al. 2000). Records of alien plants were classified according to their period of introduction (residence status), using the major subdivision between archaeophytes and neophytes as in Pyšek et al. (2002) and major periods of the recent political history as in Sennikov and Lazkov (2021). For temporal classification of neophyte alien records of the Erigeron acris group in Murmansk Region, we used the 1930s as a temporal limit; this limit reflects a major change in economic activities, transportation and human migrations, which was linked with the beginning of industrialisation in the USSR (Lewis 1979).

The history of introduction was determined based on the history of human activities in a certain locality and in the territory as a whole. The local history was obtained from historical accounts in cases of the distant past, or from technical reports and local knowledge in cases of the recent past. Local plant introductions were linked to the local human activities and their time periods. We cross-checked our information against the knowledge available from the neighbouring territories, i.e. Finland and Russian Karelia.

We inferred pathways of introduction for alien taxa, based on direct evidence as recorded by field collectors or on indirect evidence as derived from the local history. The pathways were categorised according to Hulme et al. (2008) and interpreted as recommended by Harrower et al. (2018).

The examined collections are comprehensive historical materials and include all periods of the botanical history in the present-day territory of Murmansk Region. Many of these collections were taken into account in various botanical publications. So far, no proper overview of botanical collections and their corresponding publications exist for Murmansk Region; for this reason, we provide a more detailed description of the Erigeron collections in order to uncover their link with historical publications and major events of the botanical exploration.

The first record of Erigeron acris s.l. from the Kola Peninsula was published by Jacob Fellman (1831), who reported this group from the south-western part of the territory. Nowadays, Jacob Fellman’s herbarium collection is fragmentary; its reliquiae are preserved at the University of Helsinki (Väre 2011). However, no specimens of E. acris s.l. survived in this collection.

The earliest historical specimens are available from Russian academic expeditions and the Finnish botanical exploration of the Kola Peninsula. The first extant specimens were collected by A.F. Middendorf in 1840 during his academic expedition along the Barents Sea coast (Sukhova and Tammiksaar 2015). Further collections were exclusively Finnish, linked with botanical explorations of Russian Lapland (Uotila 2013; Väre 2017). During 1861 and 1863, N.I. Fellman and his team made very important collections along the southern and eastern sea coasts and in Kola Town (Sennikov and Kozhin 2018), thus bringing evidence for an early introduction of alien plants by the Pomors. A.J. Malmberg in 1870 (Malmberg 1926; Lappalainen 1959), R.B. Enwald and C.A. Knabe in 1880 (Uotila 2013) acted as commercial botanical collectors and brought many well-documented specimens from the coastal areas. At the same time, geologist A. Göbel, who was dispatched by the Russian Academy of Sciences to Russian Lapland in 1868–1870, made some collections along the northern and eastern coasts but his botanical collections were very poorly prepared and extremely inaccurately documented. V.F. Brotherus (1886) collected in the western parts of the territory in 1885, and made the first good collections from the Rybachii Peninsula. These early explorations, which aimed at the primary floristic knowledge and focused largely on coastal areas with an emphasis on its western (Kandalaksha) and eastern (Ponoi) extremities, culminated with the Great Kola Expedition in 1887, 1889 and 1892, which was organised by the University of Helsinki and the Societas pro Fauna et Flora Fennica with the major aim to cover the interior parts of the Kola Peninsula (Rikkinen 1980; Uotila 2013). With these expeditions, the basic knowledge about native and archaeophytic populations of E. acris s.l. was obtained.

Further botanical expeditions focused on the “white spot” areas from which no botanical knowledge had been available. V. Borg and A. Rantaniemi extensively collected in Kuusamo, covering the territory of the Kutsa River and neighbouring villages (Uotila 2013). This important territory became a nature reserve and had been subsequently visited several times by various Finnish researchers before it was ceded to the USSR after the Second World War. More recently, it was revisited by T. Ulvinen who published a synopsis of the flora of the Kutsa Nature Reserve and its vicinities (Ulvinen 1996).

K. Regel made extensive explorations of plant communities in the Kola Peninsula but collected rather few specimens. We traced only two specimens of E. acris s.l. which he collected along the Ponoi River (Regel 1914, 1927). Another great early expedition focused on plant communities was made in 1927 by G. Zinserling, who collected many important specimens along the southern coast of the Kola Peninsula (Zinserling 1935).

The Polar-Alpine Botanical Garden-Institute was established in 1931 in the Khibiny Mts. This botanical institution triggered a new period of regular botanical studies of Murmansk Region. Eventually, these academic activities led to the five-volume book “Flora of Murmansk Region” (Gorodkov 1953; Poyarkova 1954, 1956, 1959, 1966), which was considered among the best regional synopses in the USSR. Distribution maps based on point occurrence data, which were provided for each species treated in this book, were digitised and made available online, including two accepted species of E. acris s.l. (Kozhin et al. 2020a).

Higher mountains of the western part of the Kola Peninsula were in focus of botanical studies in the 1930s, when their exploration for mining of natural resources had been initiated. This study was summarised by B. Mishkin in his monograph on the flora of the Khibiny Mts. (Mishkin 1953). The neighbouring Lovozero Mts. were studied in detail much later, during the 1970s–1980s (Belkina et al. 1991).

The Kandalaksha Bay, with its many islands, has been thoroughly explored for 75 years due to the existence of the Kandalaksha Nature Reserve (Kozhin and Sennikov 2020). The herbarium collections from its territory are kept also in a dedicated herbarium repository in the reserve.

Another nature reserve with a long-standing record of botanical explorations is Pasvik, situated at the border with Norway. Despite its tiny territory, its vascular plants were completely inventoried three times, but only the latest revision included records of E. acris s.l. (Kravchenko 2020). Alien plants of this territory were studied in the course of a transborder project that involved botanists from Norway, Finland and Russia (Alm et al. 1997).

As native vascular plants of Murmansk Region were considered rather sufficiently studied, their alien counterparts remained largely neglected (Kozhin et al. in prep.). During the latest 20 years, the effect of the revegetation of slag dumps in electric power stations and stockyards in mining factories was examined (Evdokimova et al. 2005; Kapelkina 2014; Timofeeva et al. 2016), however, without paying a proper attention to the introduction of alien plants. Their inventory in Murmansk Region has been started recently, with a few minor contributions published to date (Kozhin et al. 2020b; Kozhin and Sennikov 2022).

During the whole history of botanical studies, some populations of the Erigeron acris group occurring in Murmansk Region were reported under wrong names.

Quite commonly taxa were treated in very broad circumscriptions; such examples are E. acris s.l. that included either the whole complex or its hairy representatives, or E. politus that included its hybrids with E. rigidus. Such misidentifications are too impractical to mention because of their exceedingly high number.

Sometimes, more precise identifications were published, which were mostly wrong due to vague taxonomic concepts of the past. Such identifications were rather few, and such species names are in current use for narrowly defined taxa. We traced the background for these wrong records in order to provide their correct identity (Table 1).

Although we relied upon diagnostic characters on which the previous works (Tzvelev 1994; Kurtto and Väre 1998; Šída 1998, 2000; Olander and Tyler 2017) have been based, every character was reassessed in the course of our revision.

Easy to catch is the feature of purple colouration, which may affect all vegetative parts of the plant: stems, leaves and phyllaries. We found this character to be of good subsidiary value: although it may quite widely vary in plants of the same taxon being nearly green to completely purplish (e.g. E. politus), it may reliably serve for primary diagnostics between E. rigidus (purplish) and E. acris s.str. (green).

Synflorescence shape (corymbose vs. racemose) is found to be a strong and highly reliable character, in agreement with the work of Šída (1998, 2000, 2004). In weak plants, synflorescences may be reduced and their shape may appear uncertain; in such cases, this character can be inferred from the length of pedicels (synflorescence branches) and the number of capitula, as already used by some researchers (Tzvelev 1994; Kurtto and Väre 1998).

Size of capitula, measured as length of phyllaries, is an important character apparently correlating with the synflorescence shape. It can be used as a proxy for the latter, too.

Number, shape and density of cauline leaves have been commonly used to distinguish between some taxa in Central and southern Eastern Europe (Tzvelev 1994; Šída 1998, 2000, 2004), which may differ in the density of foliage and ultimately the absolute number of stem leaves below the synflorescence. Although these characters are undoubtedly useful in the E. acris group, their use in the European North is rather limited because of the lack of the densely leaved taxa. In general, our plants with corymbose and racemose synflorescences may differ in the number of cauline leaves and their density, but this difference is not so prominent and this character can be used as auxiliary here.

Flower colour (ligulate flowers) varies between pale and dark lilac, rarely (in E. decoloratus) white ligules were observed in plants outside the study area. We noticed that this character correlates with the purple colouration of stems and phyllaries and is therefore similarly variable, and sometimes may vary within a single plant when one branch is purplish and the other is greenish. For this reason we do not give a separate diagnostic value to this character.

Length of ligulate flowers was sometimes used (Fries 1846; Olander and Tyler 2017) but we found this character variable within the same taxon. Its value is rather uncertain.

Pubescence (presence or absence of simple hairs) is considered another primary taxonomic character (Tzvelev 1994; Kurtto and Väre 1998; Šída 1998, 2000; Olander and Tyler 2017). Plants in the E. acris group differ remarkably in the density and length of pubescence on stems, leaves and phyllaries. Despite a certain level of variability, we find this character reliable in distinguishing taxa within the groups with different types of synflorescence, with the hairiness ranging from extremely sparse (E. politus, E. droebachiensis) through moderate (E. rigidus, E. uralensis) to abundant (E. acris s.str., E. brachycephalus). The length of pubescence is important to help distinguishing between some taxa with shorter (E. rigidus, E. brachycephalus) and longer (E. acris s.str.) hairs, and can be used when the other characters are expressed ambiguously. The differences in pubescence are most easily observable on the involucres (Fig. 12).

Figure 12. 

Flowering heads of the Erigeron acris group in Murmansk Region A E. politus Fr. (KPABG 040062) B E. × pilosiusculus Sennikov (KPABG 039993) C E. rigidus Fr. (KPABG 040017) D E. × intercalaris Sennikov (KPABG 039998) E E. acris L. (KPABG 043965) F E. droebachiensis O.F.Müll. (KPABG 047662) G E. uralensis Less. (KPABG 043994) H E. brachycephalus H.Lindb. (KPABG 040025).

Achene characters are difficult to use in herbarium specimens, which are collected mostly in flower (Olander and Tyler 2017). However, we noticed that achene hairiness may vary within the same narrowly defined taxon, and the diagnostic value of this character is therefore doubtful.

The first attempt to classify the diversity of the Erigeron acris group was made by Müller (1782) who described E. droebachiensis from a single locality in Viken County, Norway. This species was separated not because of the advanced knowledge in this taxonomic group; instead, it was compared with E. canadensis L., a distantly related species which appears to be only superficially similar because of its racemose synflorescence.

Early botanical works recognised only a single species in the E. acris group, but its apparent morphological variability was reflected in varieties. Hartman (1820) noted some differences in the size of flowering capitula and the colour of ligulate flowers as taxonomically significant characters; at the same time, he merged the previously described E. droebachiensis even without a note on its name.

Elias Fries (1843a, 1843b, 1846) recognised several taxa in this group, at the level of species and below. He used several characters to ground his taxonomy, including synflorescence shape, size of capitula, features of foliage, pubescence of all parts of the plants, ligulate flower colour and length. Most notably, he distinguished between the southern E. acris s.str. with an abundant soft pubescence and the northern E. rigidus with sparser stiff hairs, the distinction being of a wide phytogeographic importance but commonly neglected in later taxonomic works.

Fellman (1864) reported E. droebachiensis from Russian Lapland; however, this was not the taxon in our current understanding.

Mela (1877, 1884, 1895, 1899) developed a complicated taxonomic classification for the E. acris group in Finland, in which he recognised 10 infraspecific taxa at the level of subspecies, variety, subvariety or forma. His primary distinction was laid between E. acris s.l. and E. droebachiensis s.l., which he distinguished on the basis of hairiness and subdivided further for the size and colour of capitula and minor details of pubescence. The two-taxon system in the E. acris group developed by Mela was used in contemporary Finnish publications (e.g., Saelan et al. 1889). His elevation of taxonomic ranks in this group (Mela 1895, 1899), giving the species status to E. droebachiensis, was not supported by other botanists.

Botschantzev (1959) attempted to revise the E. acris group in the USSR. He recognised only two taxa in the European North, the glabrous E. politus and the hairy E. acris, between which abundant hybrids were allegedly found. His treatment was based on the collections at LE. The first detailed taxonomic revision of this group in Murmansk Region was provided by Orlova (1966), who followed Botschantzev (1959) in the taxonomy. Orlova revised the collections at KPABG and provided point distribution maps of both taxa accepted in the territory.

Tzvelev (1994) provided a new taxonomic treatment of the E. acris group in the European part of the USSR. In the North, he maintained the two taxa recognised by his predecessors in the same circumscription but added the ill-defined collective species E. uralensis, which he distinguished by its numerous flowering heads (paniculate synflorescence) and believed to have originated from interspecific hybridisation between E. politus and E. acris. He collected three synonyms under this species name: E. brachycephalus, E. decoloratus and E. elongatiformis. By doing so, Tzvelev disregarded the character of paniculate vs. corymbose synflorescences (E. brachycephalus and E. uralensis have paniculate synflorescences, whereas E. decoloratus and E. elongatiformis have corymbose synflorescences) and apparent differences in pubescence. He used many diagnostic characters to distinguish between his species, and the number of taxa that he accepted was the highest among all taxonomic revisions to date. Nevertheless, the inconsistent use of morphological characters in his treatment did not allow him to achieve a clear picture of plant taxonomy and distributions.

The latest Finnish synopsis of the E. acris group (Kurtto and Väre 1998) was most detailed in Eastern Europe. It accepted five subspecies of E. acris s.l. on the basis of pubescence, presence or absence of a purple tint, and number of flowering heads. The subspecies were assessed according to their residence status, as native or alien (archaeophyte), and mapped according to traditional biogeographic provinces of East Fennoscandia, thus revealing their distribution patterns. This treatment was quite fairly and accurately set, except for the broad treatment of E. acris subsp. acris (which included E. rigidus) and the lack of recognition of hybrids.

Synflorescence shape was considered a primary character by Šída (1998, 2000, 2004), who based the uppermost-level division in his classification (ranked as series: E. ser. Trimorpha (Cass.) Šída, E. ser. Macrophylli Šída) on this character but also added a separate series (E. ser. Politi Šída) to accommodate completely glabrous plants. The main classification rank in this system was species. Šída provided a brief overview of the E. acris group in the whole of Eurasia (Šída 1998) and a detailed taxonomic treatment for the Czech Republic (Šída 2000, 2004). He precisely defined the species and mentioned the presence of hybrids. His work provided highly useful insights for our treatment but could not be used at the species level.

The latest taxonomic revision of the E. acris group in Sweden, with taxonomic implications for Fennoscandia (Olander and Tyler 2017), was based on a complex statistical analysis of numerous measures rather than traditional observations. Olander and Tyler (2017) believed that their accepted taxa (ranked as subspecies) may be distinguished by a combination of the characters of pubescence and dimensions (of stems, leaves and phyllaries, also including leaf and capitula number). However, they failed to observe minute differences in pubescence between E. rigidus and E. acris, and also reduced E. brachycephalus to a synonym of E. droebachiensis despite their apparently different types of pubescence, probably because they had insufficient material for examination.

The available herbarium collections indicate that flowering period may be highly variable due to the differences in particular season, vegetation zone or even local conditions. In the same locality but in different years, plants of the same species may start flowering with a difference as high as a month. At the same time, plants collected within one day in the same place may show differences corresponding to one week of observations within a single population. Plants collected in different vegetation zones (tundra vs. taiga) may start flowering with a delay of two weeks or even greater, whereas the flowering period may be quite short and limited to two or three weeks. This makes summarily observations within the whole territory of Murmansk Region practically meaningless.

At the same time, plants of different taxa may develop in clearly different periods when observed as co-occurring within the same locality. In this case, plants of E. rigidus and its hybrids may develop significantly earlier than those of E. politus, whereas plants of E. brachycephalus start flowering apparently later than those of E. rigidus. Since in such cases the difference in flowering periods is approximately a week or less, the summary observations may practically coincide.

Distribution patterns of each species in the Erigeron acris group are individual. Although some species may co-occur in the same locality and may share some part of their history of dispersal, their main sources and drivers seem to be different.

Due to a considerable confusion between the segregate taxa in this group even in the most detailed treatments (Mäkelä 1980; Tzvelev 1994), their major distribution areas remain partly obscure. The lack of separation between the native and secondary parts of distribution areas in these works makes the identification of their origin even more difficult.

Erigeron politus is a native taxon, which occurs in mountains and uplands, and in some fjords and river ravines along sea coasts in Murmansk Region. This species is a largely subarctic (oroarctic) plant in Fennoscandia (Mossberg and Stenberg 2018). Its secondary dispersal is very minor and limited to short-distance transfer of diaspores from native populations to neighbouring anthropogenous habitats (roadsides, waste lands, populated places). The species is favoured by disturbance and can be found growing rather abundantly in industrial waste lands.

The present-day distribution of E. rigidus in Murmansk Region seems to be a fair reflection of its historical dispersal. The species distribution is limited to two major areas: the entire White Sea coast with adjacent islands, which was a traditional area of the Pomors economy, and the Alakurtti–Salla road, which was a historical traffic route between northern Russia and Finland (Fig. 13A). Two isolated localities can be explained by further dispersal: Kola (by historical trade) and Apatity (more recent dispersal in the early industrial times, but most likely with some local and traditional-like traffic). In undoubtedly native habitats (tundra and remote river ravines, in which E. politus typically occurs) this species is always lacking.

Figure 13. 

Main transport networks and their major populated places in Murmansk Region A pre-industrial period B industrial period. Red dashed lines (A) show the Pomors transport route along the White Sea, the Kola road to the north, and two trade communication roads between the White Sea and northern Finland. Red solid lines (B) indicate main roads, black dashed lines (B) are railways. Special signs (B) denote electric thermal (black) and hydro (blue) power stations, and nepheline mining area (brown) mentioned in the text. Maps were created using ArcGIS software by Esri. ArcGIS is the intellectual property of Esri and is used herein under license. Copyright Esri. All rights reserved.

The distribution of E. brachycephalus follows the same pattern but is much sparser, also suggesting its connection with the pomors. However, two of its known localities have a recent origin. In the Pasvik area, the species was collected from a recently abandoned Russian military camp (erroneously reported as E. acris: Kravchenko 2020), which had a post-war origin, whereas its presence along the channel of a hydroelectric power station may be linked with revegetation activities during the late Soviet times.

Despite the broad occurrence erroneously reported in the past, E. acris s.str. was found only in a few scattered localities, mostly very recently. The only old locality of this species is known at Tetrino, along the southern coast of the Kola Peninsula. This population is part of a local hybrid swarm, found together with E. rigidus and their hybrids in 1937, and therefore can also be linked with traditional activities of the pomors.

A large area of the recent introduction of E. acris is situated along the Niva cascade of hydroelectric power stations (Fig. 13B). This cascade continues from Kandalaksha Town in the south to Lake Imandra in the north. To date, plants of E. acris were recorded in connection with the second and third power stations, growing in large populations. Another isolated population was found on abandoned stockyards of an apatite-nepheline processing plant (concentrating mill) (Fig. 13B). One more locality, in Pasvik (Kravchenko 2020), is linked with another abandoned military place.

Erigeron uralensis was newly discovered in three localities along the Niva cascade of hydroelectric power stations and on abandoned stockyards of an apatite-nepheline processing plant. It was found in mixed populations together with E. acris.

Erigeron droebachiensis was collected twice in a single locality, an abandoned slag-dump at the Kirovsk thermal electric power station (Fig. 13B). This place of introduction is not connected with any other taxon of the E. acris group, thus showing its independent origin.

It may be exceedingly difficult to establish pathways and periods of introduction of particular alien plants in the territories with a long and complicated history of introductions, or even to distinguish between native and alien plants. However, the low natural floristic richness of the Arctic is particularly helpful in revealing alien plants, and its harsh climatic conditions efficiently limit the introduction and further spread of alien plants in this territory; for this reason, the diversity of non-native alien plants in the Arctic is still considerably lower than at the southern latitudes (Wasowicz et al. 2019).

The territory of Murmansk Region belongs partly to the Subarctic tundra, partly to the Northern Boreal forest (Chernov 1971). These species-poor landscapes clearly allow for detecting alien plants when present outside seminatural or human-transformed landscapes near inhabited places or places of economic activities. We considered habitats and distribution patterns of taxa of the E. acris group in Murmansk Region and determined that only one species, E. politus, is undoubtedly native in the territory because its occurrence is predominantly linked to native landscapes, whereas the other taxa are clearly alien because they are confined to the areas of historical or modern human activities. This conclusion provides a major correction to all previous treatments of this group in Murmansk Region (Botschantzev 1959; Orlova 1966; Tzvelev 1994), which were based on the belief that ‘E. acris’ (i.e. the E. acris group excluding E. politus) is also native to this territory due to its wide distribution and common occurrence in seminatural habitats. However, the occurrence of E. rigidus and E. brachycephalus is confined to the areas with strong anthropogenous influence; their localities are typically situated nearby old human settlements or places of traditional occupational activities (fishing, saltmaking, mining).The earliest records of E. rigidus, documented by herbarium specimens at H, were made in the 19^th century near old Russian villages along the southern coast of the Kola Peninsula and along the lower course of the Ponoi River and Kola River, as well as in several Finnish villages along the road from the White Sea to Salla via Alakurtti (formerly in Oulu or Lapland Region, Finland) which was situated on a major communication road between the Russian and Finnish North (Ahti and Hämet-Ahti 1971). These records evidenced that in the 19^th century the species had been firmly established and widespread in the territory, and already connected with the areas of traditional Russian economic activities. Its more precise period of introduction to the territory can be only inferred from its historical occurrences: some older Russian villages along the southern coasts were established well before the 16^th century, which is the accepted chronological limit between archaeophytes and neophytes (Pyšek et al. 2002), and the traditional Russian economy along the southern coast of the Kola Peninsula has continued since the 12^th century (Bernstam 1978). As the factors that can be linked to the present-day distribution of E. rigidus in the Kola Peninsula (economic activities of the Pomors) emerged and became strong so early in the history, we conclude that this species is most likely an archaeophyte in Murmansk Region.

The pathways of introduction of E. rigidus are linked to the economic activities of the Pomors. The taxa of the E. acris group are very minor and insignificant weeds of field crops (Krascheninnikov et al. 1935), and their effective dispersal as contaminants is unknown and therefore unlikely. Furthermore, the species occurrence in Murmansk Region strongly suggests that its local dispersal was not directly linked with transport of goods or commodities like grain; on the contrary, the species is very commonly found on fishing and sailing places in which no cargo had been discharged, with the continuous occurrence along the coasts. For this reason, we assume that the species was introduced and further dispersed adhering to clothes, footwear and other items possessed and moved by travellers, who may be trading, fishing or performing any other traditional occupation. To the Kuusamo area, it may have been introduced by Russian peddlers who were known to trade regularly in the area (Ahti and Hämet-Ahti 1971). This pathway can be classified as Transport-Stowaway: People and their luggage/equipment (Harrower et al. 2018). Local colonisation occurred not only by wind; it was apparently aided by cattle (lambs and cows), as evident from historical records on pastures and along small watercourses that are known to have been used for grazing (Zaitseva 2000).

The historical occurrence of E. brachycephalus, which is similar to the distribution of E. rigidus but much sparser, suggests that this species was introduced and dispersed using the same agents and factors but likely in later times, probably in the 16^th–17^th centuries when further large villages were established in the Kandalaksha Gulf and along the lower course of the Ponoi River (Bernstam 1978). This species can therefore be classified as an early neophyte. Its distribution was formed using the same pathway and source of introduction as those that shaped the distribution of E. rigidus.

Bernstam (1978) suggested that the Pomors originated in the Ladoga area, southern Karelia, from which they gradually colonised the coasts of the White Sea. Their origin and vector of colonisation agrees with the nearest native distribution area of E. rigidus and E. brachycephalus, which we define as central and southern Karelia. In northern Karelia, these two species were introduced by humans many hundred years ago.

Since historical records of E. acris are limited to a single village (Tetrino) without further localities along the sea coast, we assume this occurrence to have originated from old long-distance dispersal. Tetrino is an early village on the coast, dated from the 17^th century, and its history was connected with the Resurrection Monastery at Istra (now Moscow Region) that established the village (Bernstam 1978) and the Solovetsky Monastery that, before the revolution, in the 15^th–18^th centuries possessed and in the 18^th–20^th centuries supervised territories of the former Tetrino District (Dositheos 1836). The monastery traffic from the Solovetsky Archipelago (now Archangelsk Region) may have been responsible for the long-distance dispersal of E. acris and its introduction to the territory as an early neophyte.

The modern introduction of E. brachycephalus and E. acris to military camps is linked with longer-distance transportation but probably the same pathway of introduction; it occurred in the late Soviet period after the Second World War. Similarly, cargo traffic was responsible for the recent occurrence of E. acris at the Apatity railway station. So far, it is uncertain what kind of item was contaminated with the Erigeron seeds.

Erigeron acris and E. uralensis were collected on abandoned stockyards of the First apatite-nepheline processing plant (concentrating mill) of the Apatite mining and processing enterprise. These stockyards functioned during 1956–1963 (Gershenkop et al. 2010; Mazukhina 2019) to accumulate and dump tailings produced while processing apatite-nepheline ore. After this period, the stockyards were closed and their territory was revegetated (covered by ground and plant cover), in agreement with the standards (Smetanin 2000; Chibrik 2002). To achieve a higher density of the restored plant cover, seeds of rhizomatous perennial herbs and grasses have been used since 1964 (Kryuchkov 1985; Druzhinina and Mialo 1990; Timofeeva et al. 2016). Such seeds were commercially produced in the USSR during the revegetation period rather than imported from abroad (Zolotarev et al. 2017). This type of revegetation was standard for all types of industrial lands, including slag-dumps (Smetanin 2000; Belozerova 2006). When similarly revegetated slag-dumps of the Ural Region were examined botanically, the high presence of E. acris s.l. was noted as a weed, especially on bare ground patches (Glazyrina et al. 2009; Chaschikhina 2021). This presence can be explained by contamination of the seed used in revegetation, also in Murmansk Region. This pathway can be classified as Transport-Contaminant: Seed contaminant. The source territory of introduction should be situated at least at Saint-Petersburg or farther southwards, where both alien species may co-occur. A more remote location of the source area may be indicated by the fact that E. rigidus, which belongs to the most common taxa in southern Karelia and Leningrad Region, has never been found on revegetated grounds.

We also considered a possibility for the seeds of the E. acris group to arrive with contaminated soils and found this pathway practically impossible. The topsoil used for revegetation in Murmansk Region was locally excavated peat rather than any substrate imported from previously vegetated places (Kryuchkov 1985). Although many ruderal plants, including the E. acris group, were observed on revegetated grounds in Krasnoyarsk Region where fertile soil had been used in revegetation (Efimov and Shishikin 2014), local peat can be considered completely free of any unwanted botanical contamination.

An abandoned slag-dump at the Kirovsk thermal electric power station, on which the only locality of E. droebachiensis was found, has been revegetated in a similar way but after 1990 (Davydov and Redkina 2021). The nearest distribution area of this alien taxon is Southern Karelia, where it occurs together with E. uralensis and E. brachycephalus.

The same two taxa, E. acris and E. uralensis (together with E. brachycephalus), were found along the Niva cascade of hydroelectric power stations (HPS). The last power station in this cascade was completed in 1954. This construction suggests that revegetation on the channels and dams of this cascade should have occurred no later than in the 1960s. The alien species composition and their presumed period of introduction essentially coincide for the cascade of HPSs and the mining stockyards, thus indicating a likely similar origin of the seed used in revegetation in both cases.

The occurrence of E. acris and E. uralensis at the mouth of the Niva River suggests their self-dispersal downstream from the places of their original introduction. This possibility is confirmed by the experiments indicating the ability of seeds of the E. acris group to drift along watercourses and remain viable in the end (Bill et al. 1999).

All the non-native taxa of the Erigeron acris group found in Murmansk Region can be considered naturalised aliens. These plants are biennial or short-lived perennial, reproducing by seed, and self-sustaining populations are essential for their continuous presence in the territory, which has been repeatedly observed by numerous collectors in various localities.

From the sparsely scattered pattern of historical records of E. acris s.str. and E. brachycephalus and the continuous distribution of E. rigidus, all having been introduced to Murmansk Region with travelling humans but in different time periods, we infer that some kind of longer-distance dispersal may have occurred to deliver the first propagules of these species to the territory. This introduction was subsequently complemented by short-distance dispersal with the same agents and by local dispersal with cattle.

In spite of the long period of introduction, none of these tree taxa became truly invasive. All these species formed stable local populations in semi-natural landscapes or near populated places, but none of them shows a tendency to expand from their locally restricted refugia further into native landscapes.

The alien populations introduced in the post-industrialisation period (after the Second World War) were recorded in technogenic landscapes or in military areas. In most cases these alien plants were not observed outside the area of their original introduction. However, the occurrence of E. acris and E. uralensis along the Niva cascade of hydroelectric power stations demonstrates their potential for further dispersal by running water and by wind along the river corridor, when new populations successfully established downstream from the places of their original introduction.

It is commonly considered that populations of native plant species should be used in revegetation (Smetanin 2000; Chibrik 2002; Timofeeva et al. 2016). However, the list of species tried and recommended for this purpose in Murmansk Region (Kryuchkov 1985; Druzhinina and Mialo 1990; Timofeeva et al. 2016) contains plants (Bromus inermis, Lolium pratense, Phleum pratense, Alopecurus pratensis) which are not native to Murmansk Region, and therefore their introduction contributes to a further pollution of this territory by alien plants. Moreover, these seeds are produced commercially by agricultural enterprises (Zolotarev et al. 2017) far away from the area of introduction. Use of commercial seeds collected in remote territories brings seed contaminants which may appear either as easily recognisable exotic aliens or as cryptic invaders masquerading as their native relatives. The latter case is exemplified here by the E. acris group, which provides an alarming example of new invasions that can be started even when the technical standards in revegetation have been duly followed, and then completely neglected because of the lack of knowledge in this taxonomically difficult group. In order to prevent such situations, we suggest that protection from invasive non-native plants should be more efficient at the stage of seed import, and monitoring and management of invasive non-native plants should be established in places of revegetation works.

So far, there is no direct confirmation of interspecific hybridisation in the Erigeron acris group which is based on genetic studies or experimental crosses. However, the existence of putative hybrids has been noted by a number of researchers who attempted to define taxa more precisely in this group.

Botschantzev (1959), who accepted only two taxa of the E. acris group in the Russian North, the hairy E. acris and the glabrous E. politus (which he named E. elongatus), speculated that individuals with presumably intermediate morphology originated from hybridisation between these two species. He also suggested that E. brachycephalus, E. droebachiensis, E. elongatiformis and E. uralensis may have the same hybrid origin. However, the specimens at LE collected from Murmansk Region, which Botschantzev identified as hybrids, mostly belong to E. politus, except for two specimens of E. rigidus, one specimen of E. × pilosiusculus and one specimen of E. brachycephalus that do not show any regular pattern.

Tzvelev (1990, 1994) agreed with Botschantzev’s hypothesis of intermediate origin of these taxa, except for E. droebachiensis. He lumped these taxa together as a single species, which he named E. uralensis by priority, and expanded it by adding E. decoloratus as a further synonym. Like Botschantzev (1959), Tzvelev (1990) also noticed presumably hybridogenous individuals between his accepted taxa.

As E. rigidus possesses a seemingly intermediate morphology between E. acris and E. politus and its main distribution area has an altitudinal character and lies between the areas of these two species, we speculate that this species originated from ancient interspecific hybridisation. This idea may also explain the more recent hybridisation between E. acris and E. rigidus and between E. politus and E. rigidus in the places of their current co-occurrence. This hybridisation is also inferred from intermediate morphology of co-occurring individuals, which are regularly collected together and placed on the same herbarium sheets by collectors.

So far, we detected several localities with hybridisation between E. politus and E. rigidus (Kandalaksha Gulf, Turii Mys, Varzuga, the mouth and lower course of Ponoi River) and two local areas of hybridisation between E. acris and E. rigidus (Tetrino Village, Apatity Town). We expect that the hybridisation may be even more widespread and complicated but it cannot be studied in full on the basis of the morphology of historical herbarium specimens.

The authors have declared that no competing interests exist.

No ethical statement was reported.

This research received no external funding to ANS. The work of MNK was funded by the Russian Science Foundation, grant number 22-27-20009, https://rscf.ru/project/22-27-20009/.

ANS developed the taxonomic concept and revised the nomenclature. ANS and MNK collected the material. ANS identified the material with the input from MNK. ANS analysed the status of plant records and pathways of introduction. ANS wrote the manuscript. ANS and MNK revised the manuscript. Both authors agreed to the final version of the manuscript.

Alexander N. Sennikov https://orcid.org/0000-0001-6664-7657

Mikhail N. Kozhin https://orcid.org/0000-0002-0153-0287

The dataset of distributional records collected for the present work was published through the Global Biodiversity Information Facility (available online: https://doi.org/10.15468/een8vj).