Estimation of current population status of the Alcon large blue Phengaris alcon (Denis & Schiffermüller, 1775) (Lepidoptera: Lycaenidae) in Bela krajina (SE Slovenia) based on egg counts

. An estimate of population size of the endangered Alcon large blue (Phengaris alcon) in Bela krajina (SE Slovenia) was performed based on egg counts. The presence of eggs has been confirmed in all seven currently known populations of the larval host plant Gentiana pneumonanthe in the area. Owing to the short distances between the populations it is probable that the species forms a single metapopulation with an estimated population size of about 190 individuals. One local extinction is documented on a site where P. alcon was present in 2008. We propose conservation measures to preserve the existence of P. alcon in Bela krajina.

Caterpillars spend their early larval stages feeding inside the flower buds of their host plants.After reaching their fourth instar, the larva falls to the ground, waiting for a foraging Myrmica ant to adopt it and take it into the nest (Thomas et al. 2010, Czekes et al. 2014).In ant colonies they either prey on ant brood in the case of predatory species (P.arion, P. nausithous and P. teleius) or are fed by nurse ants in the case of the cuckoo species (P.alcon) (Nowicki et al. 2005, Czekes et al. 2014).After spending at least 10-11 months in the ants' nest, gaining about 98% of their final biomass, they pupate during late spring (Als et al. 2002, Czekes et al. 2014).
The genus Phengaris is thought to have evolved in the steppes of central Asia (Sibatani et al. 1994, Als et al. 2002).The European Phengaris species were pre-adapted to survive and disperse in traditional European agricultural landscapes (Als et al. 2002).Today, all the Phengaris species in Europe are threatened and are declining due to habitat fragmentation, habitat loss, habitat degradation and changes in agricultural management (Als et al. 2002, Maes et al. 2004).As a result, all Phenagris species were included in the red lists of most European countries (Czekes et al. 2014).In Slovenia, all four species from the genus Phenagris are on the Red List of Slovenian Butterflies (Ur. l. RS 2002) and are also protected at the national level with the Decree on protected wild animal species (Ur.l.RS2004a).Two species (P.teleius and P. nausithous) are listed on Annex II and Annex IV of The EU Habitats Directive 92/43/EEC, while P. arion is listed in Annex IV.In Slovenia, P. alcon is listed as 'endangered' (EN) species and is one of the fastest declining butterflies in Slovenia (Verovnik et al. 2012).It has disappeared from large parts of its range, becoming possibly extinct in the Koroška region and having only a handful of isolated populations left in the Štajerska, Gorenjska and Bela krajina regions (Verovnik et al. 2012).
P. alcon has two ecotypes based on larval host plant and habitat type (f.alcon uses G. pneumonanthe and f. rebeli uses G. cruciata).Here, we focused on f. alcon which lives on humid grasslands, but its populations are very local and rare (Verovnik et al. 2012, Rebeušek 2006).The only regions where it can be locally abundant are in the karst poljes -flat depressions in the Dinarides, the edges of the Ljubljansko barje south of Ljubljana, and in the eastern part of Goričko (Verovnik 2000, Verovnik et al. 2012).The distribution of P. alcon in Bela krajina (study area) is well known (Verovnik & Škvarč 2002).
The aim of our study was to assess the population status of P. alcon in Bela krajina (SE Slovenia).We counted the eggs of P. alcon and the genets of G. pneumonanthe at locations with known presence of P. alcon or G. pneumonanthe from the literature.In addition, localities with suitable habitat were checked for possible new findings of larval host plant (G.pneumonanthe) or P. alcon.The results of the study are important for the conservation of P. alcon in Bela krajina.

Study area
Our study was conducted in the central part of Bela krajina (SE Slovenia) around the village of Dragatuš with existing old data on occurrence of P. alcon or G. pneumonanthe.The central part is a flat karst plain (150-200 m a.s.l.) bordered to the north by the Gorjanci Hills, westwards by the karst plateau Kočevski rog, while in the south and towards the east it borders to the Kolpa River, which also forms the border with Croatia.
Bela krajina is a region of temperate continental (sub-Pannonian) climate with annual precipitation of 1200-1300 mm.The types of soil are a reflection of an overwhelming carbonate bedrock.The karst landscape of Bela krajina is characterized by scarce superficial water flows and an intense underground water connection.The Lahinja is the largest river in Bela krajina and flows into the Kolpa (Štangelj & Ivanovič 2013).The Lahinja meanders through the plain forming wetlands that are partly conserved, for example Lahinjske luge and Nerajske luge which are complex wetlands consisting of wet meadows, marshes and reed beds.Of the few existing tributaries, it is worth mentioning the Podturščica, which is app.3.5 km long, mainly regulated stream.The only part left with meanders are the first 500 m near the spring.Until the beginning of the 20th century the landscape of Bela krajina was agricultural.After mass human emigration in the 20th century and changes in economic activities, the landscape has become heterogeneous with a substantial proportion of land being overgrown with forest (Paušič & Čarni 2012).Compared with the 20th century, the agriculture is today intensive and present in the flat lowlands.The streams were mainly regulated and the agricultural land drained in the 1980s.
The study area is partly protected by law.Lahinjske luge and Nerajske luge are two natural reserves within Lahinja Regional Park (Ur. l. RS 1998).The Lahinja and Podturščica rivers with the pertaining wet meadows are listed as valuable natural features (Ur. l. RS 2004b).

Field work
Field work was carried out on 28. and 29. 7. 2015.Flight period of P. alcon in Slovenia begins in June and lasts until early September; however, it depends on the region and on the season.At our study sites we assumed that the flying period was at its end, although we still saw some egg-laying females (at three locations).The egg count is the standard method for assessing the abundance of Phengaris alcon (Maes et al. 2004, Van Swaay et al. 2012).The white eggs are very conspicuous on the green flower buds of G. pneumonanthe; and most of the (empty) egg shells remain on the host plant until about two weeks after the flight season (Maes et al. 2004), because the larvae do not eat them after eclosion (Ebert & Rennwald 1993, as cited in Verovnik 2002).
Potential habitat patches for P. alcon and G. pneumonanthe were determined by existing data from Verovnik & Škvarč (2002), data in the database of the Centre for cartography of fauna and flora (15.7.2015; unpublished data of A. Škvarč, M. Govedič, B. Frajman, V. Zakšek and R. Verovnik) and data of locations of G. pneumonanthe mentioned by Ivanovič (1983) and Dražumerič (1992).All these locations were subsequently checked for P. alcon and G. pneumonanthe, including the wet meadows around Podturščica where data for the host plant exist but are spatially inaccurate.The locations were first examined for G. pneumonanthe, for which we inspected the meadows and the forest edge (Fig. 1).At locations with small populations of G. pneumonanthe we counted all the genets and the number of eggs on each genet.At locations with bigger populations we divided the meadow into different plots according to clear difference in density of genets or a human factor (one part of the meadow being mowed).Afterwards we delimited the habitat patches according to our field survey in ArcGIS (ver 10.2.2, ESRI 2014).In two of our study sites, where G. pneumonanthe was abundant (Lahinjske luge and Obrh), we estimated the population of genets and eggs using quadrat method.We randomly threw a stick in the meadow where we placed a 5×5 m 2 quadrat.Eggs and genets were then counted in ten randomly selected quadrats at Lahinjske luge and five at Obrh.Because the data of eggs and genets per quadrats were not normally distributed we log transformed the data to calculate the confidence intervals.The total number of genets on sites where their populations were small was probably underestimated, because some genets did not flower and therefore were very difficult to spot among the grass.However, we consider the number of such genets to be relatively small as we thoroughly inspected the meadows.To get an estimate of population size of imagos (EPS) we assumed that: (1) a female lays on average 80 eggs (Maes et al. 2004, Mouquet et al. 2005), (2) the sex ratio in the population is 1:1.The EPS was calculated by: In case of fewer than 40 eggs the number was rounded to 1 individual.Maes et al. (2004) determined the maximum local movement distance based on markrelease-recapture as 500 m.They treated flight areas separated by more than 500 m as different populations.Maes et al. (2004) also determined the maximum observed colonization capacity as 2000 m, based on distances from newly colonized habitat patches to the previously existing ones.Thus we defined distinct populations as habitat patches > 500 m apart.Conversely we treated disjoint habitat patches of < 500 m as one population.

Results
In our field study we investigated a total of 35.7 ha potentially suitable habitat in Bela krajina (Fig. 1). Figure 1 shows the current distribution of P. alcon and G. pneumonanthe based on our study.We recorded 7 populations of P. alcon within the studied area.The discoveries of P. alcon at the locations Dragatuš and Podlog (habitat patches 6 and 2 in Fig. 1, respectively) are new for the area.P. alcon disappeared from the meadow »F« at the location of Nerajske luge (Fig. 3).The adult P. alcon on meadow »8« (Fig. 1) found in 2001 (Verovnik & Škvarč 2002) was probably a dispersing individual, as no G. pneumonanthe were recorded at the location (neither in 2001 or 2015).The lowest number of eggs was at Sela, which held 5 eggs (Tab.1), and the highest at Lahinjske luge with an estimated number of 4,032 eggs (532-4593, p=0.05).The largest number of eggs present on one genet was 289 at the location Podlog.The two locations where we used different methods to estimate the population of eggs and genets are shown in Figs. 2 and 3.The patch »E« (Fig. 3) had 18 genets, but without eggs.The patch »D« (Fig. 3), on the other hand, hosted a population of 322 genets with 514 eggs.The patch »F« (Fig. 3) hosted a population of genets with eggs in 2001 (Verovnik & Škvarč 2002) but is now locally extinct.pneumonanthe); F: pokošeno, brez močvirskega svišča, lokalni izgin vrste.

Discussion
P. alcon often occurs in small and isolated, but stable habitat patches, supported by a few hectares of suitable land (Nowicki et al. 2005).This might also be the case in Bela krajina.
If we strictly follow the 500 m parameter for determining different populations, we have seven separate populations of P. alcon in Bela krajina (Lahinjske luge, Nerajske luge, Obrh, Dragatuš, Podlog, Podturnščica and Sela).However, because the distances between adjacent populations are never more than the colonization capacity of 2000 m, we can expect occasional migrations between populations.This implies a classic metapopulation (sensu Hanski 1999) of P. alcon in Bela krajina.In its classic form (Hanski 1999), metapopulations experience frequent extinctions and colonisation events.In Bela krajina four populations (Obrh, Dragatuš, Podturnščica and Sela) have less than 5 EPS and relatively small G. pneumonanthe population (< 110, except Obrh).This makes them highly vulnerable to extinction but also to re-colonization from bigger adjacent populations.P. alcon densities predominantly depend on density of G. pneumonanthe (Nowicki 2007) and Obrh is the only location in our study that stands out in this view (Tab.1).The population of G. pneumonanthe at Obrh is extensive with more than 2853 genets (without counting the mown half of the meadow, see Fig. 2) but the EPS is only 3.Although we have certainly missed some eggs using the quadrat method on the patch »B« (Fig. 3) we could expect a bigger population of P. alcon.The location was known for G. pneumonanthe but not P. alcon in 2008 (Zakšek, CKFF database), although eggs could had been overlooked.The meadows were certainly mown in the last years which might account for the low local population.Even during our study, half of the suitable habitat with G. pneumonanthe was mown (Fig. 2).Early mowing (June and July) is not detrimental to G. pneumonanthe but makes it impossible for the larvae of P. alcon to survive or eliminates stems suitable for egg laying.It will depend on the mowing regime in the years to come whether this population might increase or go extinct.Local extinction occurred in 2001 at location Nerajske luge habitat patch »F« (Fig. 3) where no G. pneumonanthe has been found in 2015.The cause might be excessive mowing and fertilization (Oostermeijer 1994).It is interesting that the small and isolated location Sela hosted 5 eggs and 18 genets (2015), which is very similar to 6 eggs and 20 genets in 2001 (Verovnik & Škvarč 2002).G. pneumonanthe can reach a high age of more than 30 years (Oostermeijer 1994), so it is possible these 18 plants are the same as in 2001.It is doubtful that this is a self-sustained population, and it probably depends on frequent recolonizations from the nearby bigger populations.
We can determine three main populations for P. alcon in Bela krajina: Lahinjske luge, Podlog and Nerajske luge with 101, 66 and 13 EPS, respectively.We estimate that Obrh is also an important local population based on arguments already stated and a high potential for increase of the population (which presumably depends on the mowing regime).

Implications for conservation
The current populations of G. pneumonanthe in Bela krajina are small and localized and thus prone to extinctions.All of them host eggs, implying that the host plant is the main limiting factor for P. alcon.The summed number of imagos in the whole metapopulation is 190.Nowicki (2007) surveyed a population of slightly more than 500 individuals of P. alcon in Poland by the method of egg count and the population remained stable over the course of two years.The metapopulation in Bela krajina might thus be stable even though the number of individuals seems low.Monitoring in the next years would show the trend for P. alcon in Bela krajina and discoveries of new habitat patches with P. alcon are also possible.
At the time of our field work some patches with G. pneumonanthe were already mown.
The biggest patch already mown was at Obrh (Fig. 1).We assume that the frequent mowing at this patch might be the cause for the low number of individuals of P. alcon.Although mowing is used to promote grassland G. pneumonanthe populations (Oostermeijer et al. 1994), it must be done in late September when P. alcon larvae have already left G. pneumonanthe (Mouquet et al. 2005).Mowing in July and August should be avoided.
Another unfavourable management practice is applying fertilizers, since this leads to eutrophication which deteriorates the living conditions for the G. pneumonanthe and should be avoided in meadows where the plant grows.
All the core populations except Lahinjske luge are part of valuable natural features (Ur.l.RS2004bc).