Lycaena helle (BINK, 1992)

Habel, J., Ivinskis, P. & Schmitt, T., 2010, On The Limit Of Altitudinal Range Shifts - Population Genetics Of Relict Butterfly Populations, Acta Zoologica Academiae Scientiarum Hungaricae 56 (4), pp. 383-393 : 386-388

EXAMPLE: LYCAENA HELLE

Genetic data for L. helle based on five highly polymorphic microsatellite loci ( HABEL et al. 2008 b) give a clear resolution and strong evidence on a regional ( FINGER et al. 2009) as well as on the continental level ( HABEL et al. 2010 a, b, c), and are suitable to demonstrate past range shifts (going back at least to the last glacial period) (cf. KIMBERLY & SELKOE 2006).

Postglacial colonisations of Lycaena helle

In a neighbour-joining dendrogram based on genetic distances sensu CAVALLI- SFORZA and EDWARDS (1967) all Fenoscandian samples build one branch most closely related to eastern European populations ( Fig. 1 ). While the Finnish populations are genetically closer to the eastern European ones than the Swedish populations, these similarities mirror the feasibility of this postglacial expansion route (eastern Europe– Finland – Sweden), also reflected in losses of alleles during this colonisation process. This colonisation pathway of L. helle from north-eastern Europe via Finland to Sweden mostly coincides with that of Trollius europaeus ( DESPRES et al. 2002) and maybe Ranunculus glacialis ( SCHÖNSWETTER et al. 2003) .

Evolution on mountain archipelagos in Lycaena helle

The low-altitude mountains of western Europe provided exclaves of suitable climatic conditions for cold-adapted species during the postglacial warming. These mountain areas are geographically isolated from each other. After colonisation, individual exchanges among these areas have been restricted, as shown, for example, by genetic analyses of L. helle : The studied populations of the Massif Central, Madeleine Mountains, Vosges, Ardennes, Eifel and Westerwald represent strongly differentiated and distinct gene pools affected by the strong isolation over several thousands of years ( HABEL et al. 2010 a, c). They are also characterised by private alleles; more than 11% of the total number of alleles analysed are endemic to a single mountain area ( Fig. 2 ). In addition to this genetic uniqueness in microsatellite alleles, morphological characters distinguish the populations of each of these mountain areas so that L. helle was split into nine subspecies ( MEYER 1982). The combination of distinct genetic and morphological characters may be used to define them as evolutionarily significant units sensu MORITZ (1994) to underpin the high evolutionary value of these relict populations.

recalculated from HABEL et al. (2010 c)

and HABEL et al. (2010 a, b)

These genetic results also underline that L. helle is suffering from anthropogenic habitat deterioration and rising isolation; particularly the populations scattered over the species’ western distribution range suffer from the low or even missing exchange rates of individuals ( HABEL et al. 2010 a). Even within the analysed mountain areas, isolation-by-distance systems support the geography-dependent interconnectivity of local populations ( FINGER et al. 2009).

Predictions for the future distribution of Lycaena helle

A Climate Envelope Model (CEM) (PHILIPS et al. 2006) identified and quantified the climatically suitable habitats for L. helle and possible connections between them (for details see HABEL et al. 2010 c). These projections of the climatic envelope into the geographic space allow to assess the potential distribution of the butterfly under the current climate (cf. ARAUJO & WHITTACKER 2005). The resulting areas with a predicted suitability of>75% are restricted to higher elevations and the North, and are all separated from each other by unsuitable areas ( HABEL et al. 2010 c); this models mostly match with the actual distribution of the species ( KUDRNA 2002). These results coincide with the genetic picture of isolated remnant population groups aggravated by the low dispersal power of L. helle ( BINK 1992) . This pattern is corroborated by strong genetic differentiation between neighbouring mountain areas like the Ardennes/Eifel complex and the Westerwald ( FINGER et al. 2009) or even neighbouring populations within such mountain areas ( HABEL et al. 2010 a, b).

Applying different scenarios of climate warming, the climate envelope models (CEMs) suggest a strong decline of potentially suitable habitats, especially at the western edge of the species’ distribution. Most of the recent areas of predicted suitability may disappear. Areas with a predicted suitability of 75% remain exclusively over areas of the Jura Mts and the Alps. Areas with a predicted suitability of at least 50% remain in parts of the Massif Central, the Pyrenees and the Vosges. Even over major parts of the Alps, potentially suitable habitats may largely disappear ( HABEL et al. 2010 b).

Figures

Fig. 1. Neighbour-joining tree based on genetic distances (CAVALLI-SFORZA & EDWARDS 1967) performed on five microsatellite loci, representing the analysed populations of Scandinavia, Finland, eastern Europe (Poland, Lithuania, Romania) and the Vosges. Genetic distances are projected on a map. Solid lines display the genetic distance, arrows show the locations of the sampling sites. Data

Fig. 2. Allele frequency distributions of L. helle populations of the Pyrenees and the western low-altitude mountains (Massif Central, Vosges, Ardennes). The colours in the pie charts indicate the distribution of alleles (white: occurring in several mountain areas, black: exclusive to a single mountain area, grey: exclusively occurring in the respective population).Data taken from FINGER et al. (2009)