FIRST CONFIRMED RECORD OF AN ESTABLISHED POPULATION OF SAILFIN MOLLY , POECILIA LATIPINNA ( ACTINOPTERYGII : CYPRINODONTIFORMES : POECILIIDAE ) , IN EUROPE

The ornamental aquarium trade has been responsible for several introductions of non-indigenous species in freshwater ecosystems. This report confirms the presence of an established non-native molly population in a tectonic, brackish, geothermal lagoon located in Greece (Lake Vouliagmeni near Athens). The Vouliagmeni molly was positively identified as the true sailfin molly, Poecilia latipinna (Lesueur, 1821), using the number of scales around the caudal peduncle as the principal diagnostic character within the sailfin molly species complex. To our knowledge, this is the first documented record of an established sailfin molly population in European waters.

The ornamental aquarium trade accounts for a steadily growing number of fish introductions worldwide.Poeciliids, in particular mollies, are popular ornamental aquarium fish being deliberately or accidentally released into the wild under the erroneous assumption that these species are effective for mosquito control (Dill and Cordone 1997, Courtenay and Meffe 1989, Gutiérrez and Reaser 2005).However, some introductions of mollies in natural ecosystems have been blamed for adverse environmental impacts (McKay 1989, Williams et al. 1998).
Within the group of mollies, there are two morphologically and behaviourally distinct species complexes, which represent monophyletic evolutionary lineages: the sailfin mollies of the Poecilia latipinna (Lesueur, 1821) complex and the shortfin mollies of the Poecilia sphenops Valenciennes, 1846 complex (Schartl et al. 1995, Breden et al. 1999, Ptacek and Breden 1998).The distinction between the two molly complexes is based primarily on morphological characters and the mating behaviour of the males (Ptacek 2005, Kozak et al. 2008).The sailfin molly complex, named after the strikingly enlarged dorsal fin of males, contains four sexually dimorphic species: Poecilia latipinna, Poecilia petenensis Günther, 1866, Poecilia velifera (Regan, 1914), and Poecilia latipunctata Meek, 1904(see Ptacek 2005).The shortfin molly complex contains an assemblage of at least 12 morphologically similar but genetically quite distinct species.Shortfin mollies have a much smaller dorsal fin and do not perform courtship during mating (reviewed by Ptacek 2005).
The taxonomy of mollies is complicated and has undergone various re-classifications (Rosen and Bailey 1963, Miller 1983, Lucinda and Reis 2005).Due to the overlapping morphological characters of molly species (Ptacek and Breden 1998) taxonomic identification of feral molly populations is difficult.Indeed several records of mollies are known to derive from misidentified specimens, such as the Vouliagmeni sailfin molly which was originally identified as Poecilia sphenops (see Chintiroglou et al. 1996).Another possible misidentification concerns a Poecilia species in Kenya which has been assigned to P. latipinna and as such it is included in FishBase and other global databases of invasive alien species (Seegers et al. 2003).We suspect that there are further citations of the latter species that represent cases of misidentifications, while on the other hand, we cannot not exclude the possibility that it reflects undetected occurrences due to incorrect identifications.These misidentifications of Poecilia species highlight the importance of taxonomic studies when examining the spread and the impact of nonindigenous species.
Fish were sampled with the aid of a fry net (a 7 × 2 m seine, mesh 2 mm) towed manually over a 10 m distance towards the lake shallows (1.0-1.5 m depth).Four hauls were conducted in different parts of the lake and the collected fish were anaesthetized and preserved in a 10% formalin solution.In the laboratory, all fish were measured (standard length, SL, mm), weighed (total weight, mg), sexed, and their colour variety was recorded.Small males, lacking the elongated dorsal fin that characterises mature males, were distinguished from females based on the presence of a gonopodium.Several morphometric and meristic characters have been used by different authors in taxonomic comparisons among molly species (Sterba 1973, Miller 1983, Snelson 1985, Kittell et al. 2005, Hankison et al. 2006, Hubbs et al. 2008).For the sake of the presently reported study we concentrated on a limited number of characters, namely the number of dorsal fin rays, the number of lateral line scales, and the number of scales around caudal peduncle, proposed previously in order to discriminate between sailfin and shortfin mollies, as well as for species differentiation within the sailfin molly species complex (Table 1).
In all individuals, the number of dorsal fin rays ranged from 13 to 16, which falls well outside the range of counts reported for Poecilia sphenops.Further, our results indicated that the number of lateral line scales was not a character contributing to taxonomic clarifications due to the great overlap in character values among species of the sailfin molly complex.Although the number of dorsal fin rays was a taxonomically more useful character, effectively identifying some individuals as Poecilia latipinna; it was not an "exclusive" diagnostic character, as a proportion of individuals displayed dorsal fin rays counts falling within the range of variation described for all three species.Only the number of scales in the caudal peduncle was a character definitely identifying all individuals as P. latipinna.The Lake Vouliagmeni molly population occurs under several colour varieties that can be grossly grouped into three main morphs: silver-greenish (wild type), black, and black-spotted (Fig. 2).Sexual dimorphism was apparent in all morphs, especially among larger fish.Environmental factors, predominantly ambient temperature, may have an effect on the colour structure of the population.Agnus (1983) has indicated that black spotting in sailfin molly is due to a single allele, and heterozygotes for the mutant allele exhibit variable degrees of penetrance and expressivity, depending on rearing temperature.Progeny developed under warm conditions (28°C) displayed more normal pigmentation, while progeny raised at cool (20°C) temperatures tended to develop melanistic spots.In Lake Vouliagmeni, the mean temperature range (22.5-24.9°C)falls mainly in the cool side of Agnus (1983) treatments, probably suggesting a shift of expressivity towards black phenotypes.
Large males indicated brighter colouration and a rectangular and considerably longer dorsal fin than females, which is a typical characteristic of many molly species, although females were larger than males, as expected for poeciliid species (Endler 1983).Small males (below ≈ 30 mm SL), by contrast, did not exhibit pronounced colouration and the dorsal fin elongation of large males.In all colour varieties females dominated over males, which is also typical for poeciliid species (Snelson and Wetherington 1980).The percentage contributions of the three morphs were as following: silver-greenish: 53.9%, black: 31.5%, and black spotted: 14.6%.These proportions varied in reverse order to their M ÷ F ratio: silver-greenish: 0.25, black: 0.33, and black spotted: 0.53.
In conclusion, the taxonomic identification of this molly presented acute difficulties stemming from three kinds of problems.First, the presence of melanistic morphs developed through captive breeding, reduced the utility of the natural colouration pattern as a diagnostic character.According to Agnus (1983) black and blackspotted phenotypes do occur occasionally in native populations, but they are very rare.Second, species of the sailfin molly species complex share many morphological similarities that complicate taxonomic comparisons.The character most widely used to discriminate among "sailfin mollies" is the number of dorsal fin rays, but as remarked by Ptacek and Breden (1998), three species (P.latipinna, P. velifera, and P. petenensis) have overlapping dorsal fin ray counts.And third, these three species are often hybridised by breeders to create new strains for the ornamental trade (Balon 2004, Fosså 2004).These strains may be viable and fertile (Sterba 1973, Ptacek 2002, Balon 2004) and may display confusing patterns of morphological variation.Despite these difficulties, we were able to identify the Vouliagmeni molly as the true sailfin molly, P. latipinna, using the number of scales around caudal peduncle as the principal diagnostic character.We presume that this population originated from a domestically-bred melanistic strain of P. latipinna, which was produced for the ornamental aquarium trade through selective breeding.This molly population has proliferated to high-density level, possibly because of favourable environmental conditions and absence of effective competitors and predators.The exact year this strain was introduced in Lake Vouliagmeni is not known, Fig. 2. The three colour morphs of P. latipinna found in Lake Vouliagmeni: male silver-greenish (A) and females black (B) and black-spotted (C) but local residents reported that mollies were already present in the lake in the mid-1960s.This population remains the only known sailfin molly in European waters, which is somehow unexpected on ecological grounds.Despite its reproductive requirement for warm water temperatures, Poecilia latipinna is regarded as a tolerant and adaptable species (Marchetti et al. 2004).Two other commonly traded Poecilia species with broadly similar life histories and environmental tolerances, namely Poecilia reticulata Peters, 1859 and P. sphenops, have a wider, yet localised, distribution in the continent.Therefore, further research is needed in order to seek and explore the reasons for the low occurrence frequency of this species in Europe.

Fig. 1 .
Fig. 1.Map indicating the geographical location of the geothermally heated Lake Vouliagmeni

Table 1
Diagnostic characters commonly used to identify molly species