Introduction

Human activity has led to the introduction of many alien organisms in all ecosystems, including freshwaters (Kilroy & Unwin, 2011). There is ample evidence of unintentional transport of aquatic organisms in shipping (ballast water), trade of species for aquaria, aquaculture, angling, fisheries, etc. (Incagnone et al., 2015). Other activities such as canal construction also facilitate invasions of aquatic alien species without physically moving them (for more details see Incagnone et al., 2015). All those interventions provide opportunities for species to travel long distances and colonise new areas. After reaching new habitats, these travellers may be recognised as alien. It is well known that their impact on ecosystems can pose a threat to biodiversity in many ways, including predation or herbivory, competition and hybridisation, altering ecosystems as a result (Streftaris & Zenetos, 2006).

Different aspects of the problem of biological invasions have been discussed widely in work ranging from simple lists of alien species on local (Najberek & Solarz, 2011), regional (Kaštovský et al., 2010; Wilk-Woźniak & Najberek, 2013) or global scales (Lowe et al., 2000) to complex analyses of their ecological and economic impacts (Vilà et al., 2011). Methods of assessing the impact of alien species are relatively well developed for plants, vertebrates and invertebrates in terrestrial, freshwater and marine environments (Vilà et al., 2009). In contrast, changes in freshwater ecosystems due to new introductions of microorganisms (cyanobacteria, algae), and the consequent socioeconomic impacts, are poorly studied. Part of the reason for this is that finding and evaluating alien microorganisms in new habitats is difficult due to their small size, sometimes insufficient taxonomic expertise, disagreements about a given species concept and the confusing terminology of biological invasions.

Due to problems defining what an alien microorganism is, their records are very often reported as exotic species (Cellamare et al., 2010), interesting species (Zapomělová et al., 2012), or first reports of a species in an area (e.g. Manti et al., 2005). Assembling the alien microorganisms jigsaw puzzle requires various kinds of information, among which are proper taxonomic identification and a good description of the native range of the species, so that expansion beyond the native range can be ascertained.

Cyanobacteria form one of the oldest groups of microorganisms, with fossil remains dating back 3.5 billion years (Schopf, 2000). Due to their ability to create toxic blooms (harmful algal blooms—HAB) they are considered dangerous organisms which may harm water habitats and local biodiversity and reduce the value of ecosystem services such as recreational uses or provision of potable water.

Some cyanobacteria are reported to be ubiquitous organisms with a nearly global distribution (Jungblut et al., 2009). Others are said to have extended their native range (Padisák, 1997; Wood et al., 2014). It is very difficult to assess the native/alien status of a taxon, but analyses aimed at determining the true status of these ‘alien’, ‘exotic’ and ‘interesting’ microorganisms can help elucidate their biogeography and ecology. We tracked the long-distance ‘walk’ of three species of cyanobacteria through different climate zones, to see whether we could reliably determine if they have indeed expanded their ranges. The larger aim was to establish the context in which the migration and evolution of these microorganisms should be understood, so that a proper assessment of their geographic expansion can be made.

Materials and methods

We selected 3 species of cyanobacteria, Cylindrospermopsis raciborskii (Woloszynska) Seenayya & Subba Raju in Desikachary 1972, Raphidiopsis mediterranea Skuja 1937 and Cuspidothrix issatschenkoi (Usachev) P. Rajaniemi, Komárek, R. Willame, P. Hrouzek, K. Kastovská, L. Hoffmann & K. Sivonen 2005. Those three species are widely accepted as alien for temperate zone in Europe (Kaštovský et al., 2010), and according to the current knowledge, their taxonomy is accepted (Komárek, 2015).

To gather information on the records of these species from different countries/continents, we made a thorough literature search using the ISI Web of Knowledge, Scopus, Google Scholar, local journals and local libraries. The identification keys of the Algae-Base (Guiry & Guiry, 2014) and Komárek (2013) were used to determine the taxonomic position of each species. Maps were created in ArcGIS 10.1 using the World Climate Zones—Simplified ArcGIS Online basemap. To track the route of each species’ ‘walk’ we determined the site of its first description (locus typicus) and subsequent findings. We used the locus typicus as the starting point because it is not possible to make a definite determination of the native range of any of the discussed species. On the maps we marked the site of the first record of the species on the local or regional scale (e.g. country, continent). The resolution of the maps ranges from country to global, depending on the availability of data. In the tables, we also present some basic environmental features such as water temperature, depth, trophic status and type of climate.

Results

All three cyanobacteria assessed in this study are recognised as tropical to subtropical species, and were first described from warm zones.

Cylindrospermopsis raciborskii (Woloszynska) Seenayya & Subba Raju in Desikachary 1972 is also known as Anabaena (Anabaenopsis) raciborskii (Woloszynska), Anabaenopsis raciborskii (Woloszynska) Elenkin, A. seriata Prescott, A. maksimilianii Obuchova, A. koganii Obuchova, A. wustericum Obuchova, Cylindrospermum kaufmanii (Schmidle) Huber-Pestalozzi and C. doryphorum (Schmidle) Bruhl and Biswas.

Locus typicus: Rava Demangan pond in Java (Wołoszyńska, 1912). The species was described from material collected in 1899–1900. However, it was found earlier in 1887 in the Nile, reported as C. kaufmannii (Schmidle) Huber-Pestalozzi, (Wołoszyńska, 1912).

Cylindrospermopsis raciborskii expanded rapidly to Europe in the last century and is considered an invasive alien species of the temperate zone (Padisák, 1997). The temporal sequence of the earliest local or regional records (Fig. 1) does not yield any solid conclusions about its potential routes of expansion, but the distribution of those records, ignoring their dates, suggests the following possible migration scenarios: (1) from the centre of Africa towards the north of the continent; (2) from North Africa to Europe; (3) from Java to eastern Australia; (4) from Java through South Asia to Europe and (5) from Australia to Southern Africa and on to the eastern parts of North, Central and South America.

Fig. 1
figure 1

Locus typicus and selected locations of the first records of C. raciborskii. Numbers mean subsequent years of observations: 1 Nile, tropical zone; 2 East, central, Java and Bali; 3 India, Varanasi; 4 Japan, Ganoike pond; 5 Northern Caspian Sea; 6 Greece, Lake Kastoria; 7 Turkmenia, Hindukush Res.; 8 China, Xi-hu Lake; 9 USA, Wooseter Lake; 10 Moldavia; 11 Uzbekistan; 12 Georgia; 13 Cuba; 14 Russia Zuevskoe cooling reservoir; 15 Venezuela, Lagartijo reservoir; 16 Belarus; 17 Caspian Sea near the mouth of Volga; 18 Kazahstan, Farkhadskoe Res.; 19 Brasil, Paranoa Res.; 20 USA, Minnesota; 21 Poland, Lake Pątnowskie; 22 Hungary; 23 Slovakia, Borycki Res.; 24 Australia, Queensland, Solomon Dam Palm Island; 25 Spain, Albufera la Valencia; 26 Nicaragua, Lake Masaya i Lake Xolotlan; 27 RPA, Zeekoevlei; 28 Australia, Murray Darling River Basin; 29 Germany, See Lieps; 30 Austria, the Old Danube; 31 France, small private fishpond; 32 Germany, lakes near Cottbus; 33 Lithuania, G1ūko Lake; 34 Mexico, Lake Asmolapan Vera Cruz; 35 USA, Florida; 36 Brazil, Garças Pond; 37 Canada, Constance Lake; 38 Portugal, Odivelas Res.; 39 Portugal, Caia Res.; 40 Portugal, Maranhão Res.; 41 Portugal, Ardila River; 42 Russia, St. Petersburg; 43 Italy, Albano Lake; 44 Algeria, Lake Oubeira; 45 Egypt, El-Dowyrat fish farm; 46 Croatia, Kopački Rit and Lake Sakadaš; 47 Thailand; 48 Coastal region of Uruguay; 49 Italy, Trasimeno Lake; 50 Sardinia Island, Cedrino Lake; 51 Ukraine, Dnieper River and Danube Delta; 52 Serbia, Aleksandrovac Lake

Full size image

Currently, the species is already found on almost all continents and in various climate zones: humid continental, humid subtropical, tropical wet, tropical dry, semi-arid, Mediterranean, marine and humid continental/subarctic (Fig. 1; Table 1). We have found no reports of it having colonised these other zones: subarctic, tundra and ice cap (Fig. 1). The majority of waters where C. raciborskii was first recorded at local and regional levels were shallow, warm (>20°C), of low transparency and high-nutrient concentrations (e.g. Padisák, 1997). However, there are some records from the Old Danube (Austria), Albano Lake (Italy) and other sites where local conditions were different (Table 1).

Table 1 Selected locations of the first records of C. raciborskii at local and regional levels (for more information about the spread of Cylindrospermopsis raciborskii before 1997 see Padisák, 1997)
Full size table

Raphidiopsis mediterranea Skuja 1937; heterotypic synonym: R. subrecta Frémy ex Skuja 1949.

Locus typicus: Kastoria Lake, Greece (Fig. 2; Table 2). It is considered an invasive alien species of the temperate zone (Kaštovský et al., 2010).

Fig. 2
figure 2

Locus typicus and selected locations of the first records of R. mediterranea. Numbers mean subsequent years of observations: 1 Greece, Lake Kastoria; 2 USA, Minnesota Lakes; 3 Sweden, Lake Trummen; 4 Canada, Bay of Quinte; 5 India, Osman Sagar, Husain Sagar, South Indian Lakes; 6 New Zealand, Lake Hakanoa; 7 Argentina, Salado River; 8 Brazil, Garças Pond; 9 China, Fishpond in Wuhan; 10 Czech Republic, South Moravia; 11 Argentina, The Yacyretá Res.; 12 Israel; 13 Australia, tropical reservoirs; 14 Spain, reservoirs: El Atazar, El Vello, Cazalegas, Lozoya, Picadas, Rio Sequillo, Rosarito, San Juan, Santillana, Valmayor; 15 Spain, Arcos Reservoir; 16 Spain, Zujar Reservoir; 17 Spain, Puentes Viejas Reservoir; 18 Egypt, El-Dowyrat fish farm; 19 Japan, Lake Biwa; 20 Lithuania, Lake Gineitiškes; 21 Vietnam, Huong River, Hoamy Reservoir

Full size image
Table 2 Selected locations of the first records of Raphidiopsis mediterranea at local and regional levels
Full size table

As in the case of C. raciborskii, the temporal sequence of the earliest local or regional records (Fig. 2) does not easily lead to hypotheses on possible expansion routes but the spatial clustering of records suggests the following routes: (1) from Southern Europe to northern and western parts of the continent; (2) from Southern Europe to North Africa and (3) from Europe to the east coast of North America. The Australian and South Asian records also seem to form a cluster separated from the European ones, a potential base for expansion to the east coast of South America.

The species has been recorded on all continents except Antarctica. Currently, it occurs in different climate zones: Mediterranean, humid continental, humid subtropical and arid. There is no information indicating a shift towards these cold zones: subarctic, tundra, ice cap or montane (Fig. 2). The locations of the first local or regional records of R. mediterranea included a wide variety of waterbodies, from shallow to very deep (Lake Biwa), and from eutrophic to hypertrophic (Table 2).

Cuspidothrix issatschenkoi (Usachev) P. Rajaniemi, Komárek, R. Willame, P. Hrouzek, K. Kastovská, L. Hoffmann & K. Sivonen 2005; homotypic synonyms: Anabaena issatschenkoi Usachev 1938, Aphanizomenon issatschenkoi (Usacev) Proshkina-Lavrenko 1968.

Locus typicus: Caspian Sea and Sea of Azov; shallow (max. depth 14 m) and of low salinity. It is known as a halophytic species. This cyanobacterium is considered to be an alien (but not invasive) species of the temperate zone (Zapomělová et al., 2012). The distribution of its records (Fig. 3) may suggest the following expansion routes: (1) from Western Asia to Europe; (2) from Europe to the east coast of North America; (3) from New Zealand and Australia to Far East Asia and (4) from New Zealand and Australia to Southern Africa and further to the east coast of South America. It has been found on almost all continents and is known from different climate zones: humid continental, tropical wet and arid. There are no reports of its progress towards these colder zones: subarctic, tundra or ice cap (Fig. 3). Many waters where C. issatschenkoi was first noted at local or regional levels were shallow and slightly saline (e.g. Baltic Sea). Salinity may be the key factor driving its expansion because it corresponds to its habitat preference. However, there are also some records from non-typical environmental conditions, including Lake Volvi (deep freshwater) in Greece, Lake Okaro in New Zealand (Table 3) and waters of low conductivity (Ferrari et al., 2011).

Fig. 3
figure 3

Locus typicus and selected locations of the first records of C. issatschenkoi. Numbers mean subsequent years of observations: 1 North Island of New Zealand, Lake Okaro; 2 Caspian Sea; 3 Sea of Azov; 4 Poland, Lake Gardno; 5 Hungary, Lake Balaton; 6 Canada, Bay of Quinte; 7 Japan, Lake Kasumigaura; 8 Greece, Lake Volvi; 9 Czech Republic, fishponds in South Bohemia, 10 Australia; 11 Slovakia; 12 Australia, New South Wales; 13 Ukraine, Zaporozh’ye Nuclear Power Station; 14 Portugal, Montargil Res.; 15 Argentina, Otamendi Natural Reserve, Laguna Grande; 16 North-Eastern France, Saint-Rémy; 17 Finland, Lake Tuusulanjärvi; 18 Spain, Guadiana River between Merida and Badajoz; 19 New Zeland, Lake Hakanoa; 20 Portugal, Lake Vela; 21 Baltic Sea, Gulf of Finland; 22 Lower Uruguay River; 23 Germany, Lake Stolpsee; 24 China, Nanhu Lake, Chidonghu Lake, Honghhu Lake, Donghu Lake, Xinhu Lake, Dongping Lake; 25 China, Yuehu Lake; 26 RPA, Hartbeespoort Dam

Full size image
Table 3 Selected locations of the first records of Cuspidothrix issatschenkoi at local and regional levels
Full size table

Discussion

Cyanobacteria are very abundant in all types of waters but our knowledge of their biogeography, migration and evolution is still insufficient. The ubiquity hypothesis (Finlay, 2002), which attributes almost unlimited dispersal abilities to microorganisms, leads to the common belief that geographic isolation does not operate in the micro-world. The resulting conclusion is that allopatric speciation is rare or absent and that therefore, the global morphospecies diversity of microbial organisms is low (Finlay, 2002). The moderate endemicity model (Foissner, 2006), on the other hand, suggests that even microorganisms can have restricted geographic ranges (Papke & Ward, 2004; Boenigk et al., 2006). In the changing world, with increasing pollution, global warming, carbon dioxide effects and rising salinity, species expansions are often recorded (Paerl & Huisman, 2009). However, in the micro-world, the uncertain provenance of species obscures their expansions even more. The scattered bits of information we have on the geographic range, locus typicus or ecological requirements of a species are puzzle pieces that can help complete the picture and answer the question of its status.

Cylindrospermopsis raciborskii, Raphidiopsis mediterranea and Cuspidothrix issatschenkoi—products of migration and speciation?

It is believed that these three species originated from a region of warm climate (Kaštovský et al., 2010). According to the classification proposed by Schabetsberger et al. (2009), C. raciborskii, originally reported from habitats between the Tropic of Cancer (23°27′N) and Tropic of Capricorn (23°27′S), is a circumtropical species. R. mediterranea and C. issatschenkoi were classified as tropicopolitan species, meaning that they are found frequently throughout the tropical and subtropical zones (Schabetsberger et al., 2009).

According to Padisák (1997), the primary evolutionary centre of C. raciborskii is in tropical Africa, from where it expanded to Europe, and the second evolutionary centre, from where the species expanded to South America and Asia, is Australia. Although it is difficult to support this by tracking the temporal sequence of the first local or regional records, we drew similar conclusions from our assessment of the spatial distribution of records irrespective of their dates. The mechanisms of those expansions are not clear, however. So far the presence of C. raciborskii has been confirmed in all climate zones except for the subarctic or any other cold zones. Its northernmost locality lies within humid continental climate.

Several hypotheses might explain the success of C. raciborskii outside its evolutionary centres. One of them considers its wide physiological tolerance of various environmental conditions (Kokociński et al., 2010; Piccini et al., 2011). Others invoke the trend towards warming of waters as an effect of, for example, the North Atlantic Oscillation (Briand et al., 2004), or global warming (Paerl & Huisman, 2009).

Usually, the first local records of C. raciborskii outside its locus typicus were from shallow, warm, turbid waters with different concentrations of nutrients. Exceptions include deep and cold waters in different parts of the world such as Austria (Dokulil & Mayer, 1996), Canada (Kling, 2009) or Uruguay (Bonilla et al., 2012). Such diversity of conditions under which this cyanobacterium occurs may be explained by the existence of different ecotypes and genotypes (Piccini et al., 2011; Sinha et al., 2012, 2014). Physiological and ecological studies showed that genetically and ecophysiologically differing ecotypes of C. raciborskii occur in temperate lakes (Kokociński & Soininen, 2012). They are able to persist at temperatures lower than the laboratory optimum (Briand et al., 2004). The occurrence of this cyanobacterium in such habitats suggests its wider ecological tolerance (Kling, 2009) and high phenotypic plasticity (Bonilla et al., 2012), and a capacity for fast adaptation to various conditions, which may lead to changes in genetic structure and consequently to speciation. This falls in line with the conclusion that the success of C. raciborskii is “due to a combination of several factors” and that it is “based on the existence of different ecotypes” (Piccini et al., 2011). It also points to the need for genetic studies as a key component of any framework developed for distinguishing between native and alien species.

As different ecotypes reflect different genotypes, genetic analyses should lead to a better understanding of the expansion of small planktonic organisms. A phylogenetic tree constructed using the 16S rRNA gene sequences of strains of C. raciborskii isolated from freshwaters in Australia, Brazil, Germany, Hungary, Portugal and the USA showed phylotypes grouped into Australian, European and North/South American ones (Neilan et al., 2003). Other work showed that the taxon found in Constance Lake (Canada) corresponded to one of the two basic morphologies first described in Australian reservoirs, and to some strains from Florida lakes (Hamilton et al., 2005). Those findings support the suggestion that it dispersed from Africa. Using different techniques, Haande et al. (2008) separated strains of C. raciborskii into American, European and Australian-African groups (ITS1, nifH), or American and European-Australian-African groups (PC-IGS, rpoC1). Wood et al. (2014) showed a close relation between strains from New Zealand and South America. But the high genetic similarity between the Brazilian and Uruguayan strains suggests recent population exchange (Piccini et al., 2011). Using sequences from the phycocyanin intergenic spacer region, Haande et al. (2008) and Dyble et al. (2002) found that the American C. raciborskii populations split very early and posited early evolution on the American continent. The possible expansion pathways inferred from the map of the sequence of first records of C. raciborskii are in partial accord with genetic findings. The claim of Australia–New Zealand as a potential starting point of its expansion is supported by its long presence in New Zealand (Wood et al., 2014).

When the ecological (Padisák, 1997; McGregor & Fabbro, 2000), physiological (Burford & Davis, 2011; Sinha et al., 2012) and genetic data (e.g. Gugger et al., 2005) are put together, the hypothesis that C. raciborskii expanded from Africa and from Australia to Eurasia and the Americas seems highly probable. The contrasting hypothesis that the recent ‘invasion’ into Europe was due to its spread from warm refuges within the continent (Gugger et al., 2005; Haande et al., 2008) would support the claim that C. raciborskii is native to Europe. According to Gugger & Hoffmann (2004), Kling (2009) and Sinha et al. (2012), C. raciborskii survived multiple glaciations and climatic changes in the Pleistocene by forming akinetes, and recent climate change and increasing eutrophication have promoted its spread through Europe and the Americas.

Vectors of introduction are another significant piece of the alien/endemic puzzle. For example, if the African route involved the Suez Canal, then the expansion was human caused, and hence, C. raciborskii should be classified as alien in Europe. It is also possible that it was brought from Africa with migrating birds, a natural phenomenon which would confer native European status to this cyanobacterium.

In natural environments, C. raciborskii displays high morphological plasticity which is not fully explained by genetic distance (McGregor & Fabbro, 2000). Due to its wide morphological variability, the species has been misidentified as Anabaenopsis, Raphidiopsis and Cylindrospermum (Hawkins et al. 1997). Thus, some authors speculated that R. mediterranea may in fact be a form of C. raciborskii (Moustaka-Gouni et al., 2009).

Raphidiopsis mediterranea is believed to have originated from the Mediterranean zone (Greece; Kaštovský et al., 2010). Due to morphological and ecological similarity and the frequent co-occurrence of Raphidiopsis and Cylindrospermopsis, the idea that the two genera are in fact one genus was proposed in the 1970s (Hill, 1970). The possible relation between the two species has been discussed on the basis of an examination of samples from lakes in New Zealand (Hill, 1970). Komárek & Mareš (2012) suggested that identification of both of species is difficult and that young filaments of C. raciborskii without heterocytes are often misidentified as R. mediterranea. In studies provided in Lake Kastoria, morphological and phylogenetic evidence suggests that R. mediterranea represents a non-heterocytous life-cycle stage of C. raciborskii (Moustaka-Gouni et al., 2009, 2010). The two taxa co-occur in tropical and subtropical reservoirs (McGregor & Fabbro, 2000; Chellappa & Costa, 2003; Mohamed, 2007; Fonesca & de Bicudo, 2008). According to Komárková et al. (1999), R. mediterranea populations mixed with C. raciborskii were most likely filaments of C. raciborskii lacking heterocysts. Other researchers considered R. mediterranea trichomes to be environmental morphotypes of C. raciborskii, which could be induced to form heterocysts under conditions of abundant phosphorus (McGregor & Fabbro, 2000), as found earlier in Lake Paranoá, Brazil (Cronberg, 1978).

Genetic studies suggest that C. raciborskii and R. mediterranea are different forms of one genus (Gugger et al. 2005; Moustaka-Gouni et al., 2009) or that they descended from a common ancestor that produced heterocytes (McGregor et al., 2011). Morphological changes led to the loss of the ability to produce heterocytes, separating them into two genera (McGregor et al., 2011). If they share a congenetic origin, Raphidiopsis is expected to lose its nitrogen fixation ability and deviate from Cylindrospermopsis (Wu et al., 2011).

The third species, Cuspidothrix (Aphanizomenon) issatschenkoi, is also difficult to separate from the others. The non-heterocytous stage of it can be misidentified as R. mediterranea (Wood et al., 2007). Molecular studies showed that there are at least three ecotypes of C. issatschenkoi, and that some trichomes identified as R. mediterranea var. grandis are one ecotype of C. issatschenkoi (Wood et al., 2007). The genotypic diversity of C. issatschenkoi, accompanied by ecotypic diversity, is the key to its successful range expansion, allowing it to adapt to a wide range of environments (Hodoki et al., 2013). This is also true for other cyanobacteria such as M. aeruginosa (Briand et al., 2009). Again, the possible pathways of migration/expansion are quite similar to the other species. Although its locus typicus is in the Ponto-Caspian region, recent studies of sediments from Lake Okaro revealed that C. issatschenkoi was already present in New Zealand 120 years ago (Wood et al., 2009).

Our consideration of all these results demonstrates how complex and uncertain is the task of understanding the migration, expansion and evolution of microorganisms. With all these pieces of the puzzle at our disposal, we still cannot unequivocally decide which species are alien. Tracking three species commonly regarded as alien through their long-distance walk through different climate zones, we were not able to reliably determine if they have indeed expanded their geographical range. By adding more parameters, such as ecological requirements, physiological responses and genetic structure, we were, however, in a position to suggest that the three species may be close relatives or may descend from a common ancestor.

Conclusion

We discussed the routes of expansion of three cyanobacteria commonly regarded as alien species of the temperate zone. Global warming has become the subject of much research focusing particularly on species that have expanded from regions of warmer climate. Many pieces of the puzzle are missing from the discussion of the warm origin of species and their expansion, but the picture is gradually being completed. To have a better view of the expansion and evolution of microorganisms we need to combine many types of information, including the history of local expansions, ecological ranges, and data from studies in morphology, ecology, genetics and paleolimnology. In the case of microorganisms, assessing the role of human intervention in range expansions will always be a problem. As direct evidence seems to be virtually non-existent, the only option is to rely on circumstantial evidence.