Journal of Insect Conservation

, Volume 21, Issue 3, pp 559–571 | Cite as

Isolated Asian steppe element in the Balkans: habitats of Proterebia afra (Lepidoptera: Nymphalidae: Satyrinae) and associated butterfly communities

  • Alena BartoňováEmail author
  • Vojtěch Kolář
  • Jana Marešová
  • Martina Šašić
  • Jana Šlancarová
  • Pavel Sucháček
  • Martin Konvička


A characteristic butterfly of Asian steppes, Proterebia afra (Fabricius 1787), was studied in its two relic areas of occurrence within the Balkans—the Askion Mts (a single mountain massif in NW Greece) and Dalmatia (karstic inland S Croatia)—together with co-occurring butterfly communities during its early spring adult flight period. P. afra adults tolerate harsh continental conditions prevailing at its localities during its flight. Local densities were higher in the Askion Mts (up to 100 individuals per 50 m walk) than in Dalmatia (up to 30 individuals per 50 m). Within both areas of occurrence, it inhabits dry grasslands subject to intermediate grazing pressure that retains enough dry grass litter but still suppresses shrubs and trees. Hotter slopes are preferred in more northerly Dalmatia than in the more southerly Askion Mts. More co-occurring butterfly species and more Mediterranean species were recorded in the Askion Mts (69 vs 63; 23 vs 12), but more butterfly individuals and higher species’ richness per transect were observed in Dalmatia. In ordination analyses, the main gradients organising the grassland butterfly communities distinguished between grassy plains and rocky slopes containing more woody plants. Given the distribution extents, local densities and present land use conditions, the Balkan P. afra is not currently endangered, but this may swiftly change with grazing abandonment, especially in Dalmatia, and some conservation actions should be considered in the near future.


Proterebia afra Proterebia phegea Butterfly communities Mediterranean grasslands The Balkans Steppe 


Proterebia afra (Fabricius 1787) (Nymphalidae: Satyrinae) is a butterfly with presumably relic distribution within Europe, the only representative of its genus worldwide, and deeply phylogenetically distinguished from other European Satyrines (Deodati et al. 2009; Peña et al. 2015). With its dark colour and early spring flight, it resembles the mountain genus Erebia in continental conditions.

Its rather continuous range covers the Pontic steppes and northern parts of the Irano-Turanian biogeographic region: Turkey, Transcaucasia, Northern Iran, Crimea, Northern Caucasus, Volga region, Southern Urals, up to Eastern Kazakhstan. A Few additional isolated populations inhabit the Balkan peninsula; they are classified as P. a. dalmata (Godart 1824), occurring in Dalmatia, South Croatia (from Pag island to the Biokovo Mts.) and reaching Bosnia and Herzegovina (Koren et al. 2010; Koren and Trkov 2011); and P. a. pyramus de Louker and Dils 1987 in the Askion Mts, NW Greece, Kozani district (Pamperis 2009). A small population of the latter inhabits Greek Thrace (Xanthi district) (Pamperis 2011). Whereas the Askion population is limited to a single mountain massif and its surroundings (ca 250 square kilometres), the Croatian distribution covers approximately 5,000 square kilometres (Koren et al. 2010), encompassing altitudes from sea level up to ca 1,400 m a. s. l.

The Balkan steppe-like dry grasslands, exclaves of the Asian steppe biome, exist only locally within the biotically diverse Balkan Peninsula (Polunin 1980; Radovic et al. 2009), part of the Mediterranean global biodiversity hotspot (Myers et al. 2000; Mittermeier et al. 2005). They represent extra-zonal biome patches, presumable remnants of a more arid past. The Mediterranean basin was formed during the Tertiary as a descendant of the Thetys Sea (Dewey et al. 1973; Dercourt et al. 1986). From the Eocene onwards, repeated accumulation of water in glaciers caused reductions of sea (Fairbanks 1989; Ryan et al. 1997), extending the continental conditions in Eurasia (Allen et al. 1999). Under such circumstances, the Balkans acted as both a glacial refugium and land bridge allowing East–West migrations of steppe elements (Van Andel and Tzedakis 1996; Magyari et al. 2008). More humid periods, such as the recent, witnessed area shrinkage and disjunct distribution of the Balkan steppes (cf. Aufgebauer et al. 2012).

As with many South European biotopes, the Balkan steppe-like grasslands may be at risk of successional overgrowth associated with human land use change (Debussche et al. 1999; Mazzoleni et al. 2008; Slancarova et al. 2016). A crucial role of grazing herbivores, either wild or domesticated, for maintenance of steppe-like grasslands near the biome distribution limits is beyond doubt (e.g., Cremene et al. 2005) and grazing reduction could turn the steppe exclaves into a scrubland formation.

The species is not endangered either globally or within Europe, and its listing as Near Threatened for Croatia (Sasic et al. 2013) reflects its biogeographic uniqueness rather than decline. However, it may be viewed as an insect model for other Asian steppes’ relics in Southern Europe, some of them probably more endangered.

The presented paper is a result of two 6-week surveys on the areas of P. afra European occurrence, aiming to collect more detailed data on this charismatic species. Specifically, we (a) describe adult behaviour observed in the two areas; (b) map distribution extent in Greece and Dalmatia; (c) describe and compare habitat conditions in the two areas; and (d) use ordination analyses to study co-occurrence patterns with other butterfly species in the two areas. We used this information to discuss future prospects of P. afra in Europe.

Materials and methods

The species

The single annual P. afra generation occurs from late March to early June (Koren et al. 2010; Pamperis 2009; Tolman and Lewington 2008); the March 3 records in Koren et al. (2010) are erroneous (2016, Verovnik pers. comm.). Festuca ovina L. agg. is the host plant in Europe, but oviposition was also observed on Bromus condensatus Hack. (Mihoci and Šašić 2007). After 2 weeks as an egg, the larva hatches and immediately starts feeding, probably diapauses in summer, overwinters as the 5th instar and the adults appear in spring after a short (20 days) pupal stage (Roos et al. 1984; Jutzeler and Lafranchis 2011). Reported habitats in the Balkans include dry limestone slopes or plains dominated by grasses (Stipa L., Festuca L., Aegilops L.), or Astragalus L. (Hesselbarth et al. 1995; Tolman and Lewington 2008).

The species’ nomenclature has been recently discussed. G. Lamas (2012, by D. Bolt to, unreferenced) argued that the name Proterebia afra (Fabricius 1787), derived as feminine adjective of the original name Papilio afer Esper, 1783, was a junior primary homonym of Papilio afer Drury, 1782. This led Kudrna et al. (2015) and Numa et al. (2016) to revive the synonym Proterebia phegea (Borkhausen 1788). Here, we retain the name Proterebia afra (Fabricius 1787), as the one in common use (e.g., de Jong et al. 2014; Jutzeler and Lafranchis 2011; Tshikolovets et al. 2016).

Study areas

The Askion Mts (highest summit: Mt. Siniatsiko 2,111 m a. s. l., prominence 970 m, length × width ca 35 × 25 km), located in Western Macedonia, Greece (Fig. 1a, b) have a notably xeric character owing to the rain shadow produced by surrounding higher ranges—Pindos (2,637 m a. s. l.) in the West, Olympos (2,917 m a. s. l.) in the Southeast, Voras (2,524 m a. s. l.) in the Northeast, and Baba (2,601 m a. s. l.) in the North. The southern parts of the Askion Mts are built of limestone, the northern part of gneisses, amphibolites and slates (Androulakakis 2012). The northern parts are wooded; the southern ones are covered by steppe-like grasslands. Grazing (cattle, sheep, and goats) is practiced at accessible slopes, small crop fields are found in polje depressions.

Fig. 1

a The map of the Balkans showing positions of areas of P. afra distribution. b Detailed map of the Askion Mts, showing positions of 23 transects used for butterfly recording (dark circles), our observations of P. afra outside of the transects (crosses) and major human settlements (black dots). The distribution of the species in Askion Mts mostly follows the calcareous bedrock (hatched). c Detailed map of Dalmatia, again showing the 23 transects (dark circles), our observations outside of the transects (crosses), records from literature (light circles; Mihoci and Šašić 2005; Koren et al. 2010; Koren and Trkov 2011; Tvrtković et al. 2015) and major human settlements (black dots). The map was produced using ArcGIS (ESRI 2011) and GTOPO30 terrain data

(courtesy of the U. S. Geological Survey) as a background map

Continental Dalmatia, southern Croatia, stretches from the Velebit Mts (the highest summit 1,757 m a. s. l.) in the North to the Prevlaka peninsula in the South and to the Bosnia and Herzegovina borders in the East (Fig. 1a, c). The evergreen Mediterranean scrubland typical for the Adriatic Sea coast recedes rapidly within a few kilometres inland, where it is replaced by the arid grasslands on karstic bedrock, which gradually rise towards the Dinaric Alps (1,913 m a. s. l.). The landscape is either abandoned or grazed, the pastures are often managed by fires; crop fields are restricted to villages’ surroundings. The distribution of P. afra (Mihoci and Šašić 2005; Koren et al. 2010; Koren and Trkov 2011; Tvrtković et al. 2015) includes the island of Pag, the southern part of the Velebit Mts, the inland except for the narrow evergreen coastal strip, the W slopes of the Dinarid Mts on the Bosnian border, and the northern slopes of the Biokovo Mts (1,762 m a. s. l.). The distribution area is ca 170 × 30 km. Landscape exploitation in the region is still hindered by the presence of land mines, placed there during the 1990s wars.

Field surveys

We visited the Askion Mts. between mid-April and early June, 2014, and Dalmatia in the same period of, 2015. In each area, we established 23 transects, designed to cover the entire range of natural grassland and scrubland habitats and elevations. The Askion Mts transects were situated within ca 200 square kilometres (Fig. 1b, Online Resource 1). The Dalmatian transects, reflecting the much larger P. afra distribution extent, were situated within ca 2,500 square kilometres, (Fig. 1c).

Each transect was 200 m long, divided into four 50 m segments. Each segment was defined by geographic coordinates and altitude (lat, long and alt) and the following habitat characteristics, recorded from a 50 × 50 m square dissected by the transect path, during the first visit (Table 1): first, cover as separated proportions of rock, stone, scree, bare soil, tree (>1.5 m), shrub, green plants (forbs and grass) and grass litter; furthermore, inclination (scale 0–4, from flat to steep); aspect (ordinal variable expressing thermal conditions of the slope: 1: North, Northeast, 2: Northwest, East, 3: flat, 4: Southeast, West, 5: Southwest, South); and gully (the presence/absence of a sheltered or wetter place, a track). Supply of nectar (1: <10 flowers, 2: <100 flowers, 3: flower rich site), was recorded separately for each visit and used as the mean value in analyses.

Table 1

Descriptive statistics of habitat characteristics recorded on 50 × 50 m study sites of P. afra occurrence in Askion Mts and Dalmatia


Askion Mts.


Habitat characteristic

Mean (±SD)



Mean (±SD)




2.0 (4.78)



8.2 (10.69)




5.7 (6.87)



11.7 (7.49)




11.7 (13.05)



20.8 (18.01)




19.4 (13.07)



13.3 (8.29)




13.9 (17.0)



8.5 (9.15)




10.4 (8.61)



5.4 (6.04)



Green plants

51.3 (21.88)



32.0 (13.66)




19.3 (16.69)



23.6 (13.8)




2.2 (1.06)



0.6 (0.81)




3.3 (1.38)



3.2 (0.97)






23 yes/69 no



12 yes/ 80 no


1.7 (0.50)



2.0 (0.43)



Values for cover (tree to litter) are in percentages (they can overlap); inclination, aspect and nectar are in ranked scales (see text), gully as present or absent

We visited each transect four times (mid-April, late April, early to mid-May and late May/early June) to record P. afra and associated butterflies. The butterflies were recorded by adjusted transect walking (Pollard 1977), in a ≈7 m cube in front of the recorder, with 30 min spent on the whole transect. The walks were restricted to conditions suitable for butterfly activity, between 9:00 and 16:00, local summer time. The timing of walks of individual transects was randomised among the visits. We recorded the abundance of P. afra and all co-occurring butterflies. We also noted down all observations of P. afra adult activity. The co-occurring butterflies were identified in flight, in the net, or, in case of difficult species, in the laboratory. The associated butterflies were classified into Mediterranean or non-Mediterranean species: the former distributed in the Mediterranean region, not crossing the Alps, sometimes reaching the Pannonian and Pontic lowlands. Butterfly nomenclature (Online Resource 2) follows Fauna Europaea (de Jong et al. 2014).

Apart from the transects, all other records of P. afra presence were noted and the coordinates used to produce a map of the overall distribution of the species in the two areas (Fig. 1).

Statistical analyses

All analyses were performed separately for Askion Mts and Dalmatia, and for pooled data from both study areas.

Habitats of P. afra

Presuming that adult abundance provides a good signal of habitat suitability, we modelled effects of environmental predictors recorded along the transects, using the information theory approach, selecting from multiple non-nested a priori hypotheses (van Strien et al. 2011) and relying on the Akaike information criterion (AIC) for model selection. For the whole dataset and for each study area separately, we constructed sets of generalised linear models (GLM) in R 3.2 (R Core Team 2016) with Poisson distribution of errors, for the following hypotheses:

(i) As a xerophilous butterfly, P. afra should react positively to sward openness (predictors: rock, soil, scree, stone). (ii) Likewise, it should be negatively associated with tree and shrub. (iii) Inclination, aspect and gully are directly associated with microclimate. (iv) Nectar is an important adult resource for majority of butterflies. (v) Due to its association with grasses, P. afra abundance should respond to litter and green plants cover.

Prior to assessing the above hypotheses, we constructed P. afra abundance models based solely on latitude, longitude and altitude of transect sections, including interactions. These covariate models were constructed iteratively, using backward and forward procedures, until models with the lowest possible AICs were achieved. Onto this model, i.e., onto the residuals of variation due to geographic position and proximity of transect sections, we one-by-one added predictor of the above hypotheses (i)–(v), searching for predictors combinations best explaining the data.

Co-occurring butterfly communities and their habitats

To investigate P. afra co-occurrence patterns with other butterflies, we used ordination analyses computed in CANOCO v. 5.0 (Ter Braak and Šmilauer 2012). In all analyses, we used per-segment abundance records for all butterfly species, P. afra included. Four species recorded as single specimens in the whole study were excluded: Zerynthia cerisyi, Erynnis marloyi and Argynnis pandora (all from Askion) and Ochlodes sylvanus (Dalmatia). The species abundances were log-transformed and the function downweighting rare species was used.

Unconstrained ordinations of species indicated gradient lengths =3.0 (pooled data), 2.6 (Askion Mts) and 2.5 (Dalmatia). Since the longest gradient was ≈3.0 and the data were compositional, we used unimodal ordinations (Ter Braak and Šmilauer 2012). First, we used the canonical correspondence analysis, CCA, a constrained ordination relating the composition of samples to external predictors, to construct a model based on latitude, longitude, and altitude of transect sections, including interactions. The CANOCO forward selection procedure and Monte-Carlo permutation tests (999 runs, p < 0.01) were used to define a significantly fitting combination of these predictors. This combination was used as a covariate model in a partial detrended correspondence analysis (DCA). This ordination depicted the co-occurrence patterns among the butterflies after removing the patterns attributable to geographic position and proximity of transect sections. Third, to describe the co-occurrence patterns, another CCA using the forward selection procedure was adopted to pick up the combination of habitat characteristics, with covariate model as in the previous step.


Adult behaviour

Patrolling of males over grasslands was the most frequent activity. Females were also frequently seen in flight, being more sedentary than males but still not hiding as was observed in other species, probably seeking ovipositing possibilities. We repeatedly observed oviposition on Festuca ovina in the Askion Mts; this activity consisted of short settling on the grass tussock, dropping a single pinkish egg and flying towards another tussock. Nectaring was extremely rare, always in early afternoons (1 p.m.) on Alyssum L. sp. (n = 2) and an unidentified yellow-blooming Ranunculus L. sp. (n = 1) in the Askion Mts, and on Thymus L. sp. (n = 2), Globularia cordifolia L. (n = 1), and Crepis rubra L. (n = 5) in Dalmatia. Under less favourable weather, P. afra was often the only butterfly detectable, taking off from the grass when closely approached. In the Askion Mts, we observed a few individuals patrolling over rapidly melting snow after a snowstorm in mid-April.

Distribution in the Askion Mountains

The Askion Mts distribution follows the calcareous bedrock (Fig. 1b; for GPS see Online Resource 1). Besides the Askion massif, we recorded the butterfly also in an exclave SW of Lake Kastoria and on northern calcareous foothills of the Vourinos Mts. The species inhabited all types of grasslands existing on the mountains and surrounding: rocky southern slopes above Siatista, half-shaded oak savannas on eastern slopes, grassy plains at the foothills and on the karstic plateaux, mountains summits heavily grazed by cattle, and Stipa grasslands with Juniperus bushes in the Kozani surroundings (Fig. 2).

Fig. 2

Types of biotopes hosting P. afra. Askion Mts: a foothills near Lygeri village, altitude 700 m, a site without shrubs or trees and with high dry grass litter accumulation; b central karstic plateau with some shrubs above Galatini village, altitude 1100 m, Mt. Siniatsiko summit in the background; c rocky slopes above Siatista village, altitude 1300 m; besides P. afra inhabited by Euchloe penia and Pseudophilotes bavius. Dalmatia: d foothills of Svilaja Mt, altitude 500 m, lightly grazed grasslands with shrubby patches; e mainly abandoned landscape in the Knin surroundings, altitude 300 m; e heavily grazed cattle pasture on northern Svilaja Mt slopes, altitude 800 m. At sites a and d, we recorded the highest abundances in Greece and Croatia, respectively

On the foothills (<800 m a. s. l.), the flight period started prior to our arrival, peaked during the first visit (13 April), and heavily worn individuals still occurred on 9 May. On summits (ca 1,400 m a. s. l.), the first individuals appeared on 17 April and heavily worn individuals were still present in early June. The peak abundances were up to 100 individuals per 50 m transect segment. Mean abundances per segment were 9 (median 5, SD 13.9, range 0–100) for the first visit, 15 (median 11, SD 13.6, range 0–70) for the second visit, and 13 (median 6, SD 16.0, range 1–100) for the third visit; the species was present only on the two highest altitude transects during the fourth visit. The altitudinal range of records was 691–1,556 m.

Distribution in Dalmatia

Our findings corroborated those of Koren et al. (2010) (Fig. 1c). The habitats on Pag Island were heavily grazed stony plains with scarce vegetation (Fig. 2). Near Benkovac, the butterfly occurred in abandoned rural landscapes with tall grasslands, small crop fields, sheep and goat pastures, motorway embankments. Near Knin and near the Vransko Lake, where the abandoned landscapes were mostly covered by scrub, P. afra was limited to the remnant open patches. The localities near the village of Unesic were sheep and goat pastures, sometimes dominated by sedges. Svilaja Mt was mostly grazed by cattle; newly established pine plantations were present in some areas. On Svilaja Mt northern foothills (Štikovo village), open woodlands developing at sites of abandoned gardens and crop fields are used as sheep and goat pastures. The shrubs overgrowing the steppe-like pastures in Dalmatia included Juniperus spp. and Pauliurus spina-christi Mill.

The first individuals were observed on 12 April. With the progressing season, the butterfly occurred on almost all grasslands in the landscape, but only as a few individuals per segment. The abundance peaked in early May. Mean abundances per segment were 0.6 (median 0, SD 1.1, range 0–5) for the first visit, 4 (median 2, SD 5.8, range 0–30) for the second visit, and 3 (median 2, SD 2.8, range 1–15) for the third visit; no butterfly was recorded during the fourth visit. The altitudinal range recorded was from sea level to 1,056 m.

Habitat requirements

For pooled data from both areas as well as for the two areas analysed separately, covariate model (geography) fitted over 30% of variation in the data (Table 2). On residuals of these the covariate models, all groups of predictors decreased the deviance of pooled data, supporting our hypotheses regarding P. afra abundance. The best supported hypotheses were (v), concerning herbaceous cover and indicating the highest P. afra abundance at transect sections with intermediate grass litter; (i), indicating that P. afra avoided high cover of bare ground; and (iii), indicating that P. afra reached higher abundance at flat sites with low values of aspect (i.e., north- or east facing), but also avoided humid gullies. Taking this altogether, the best P. afra sites would be situated on relatively flat and not too hot surfaces, which would contain an intermediate amount of grass litter.

Table 2

The set of a priori GLM models of habitat selection of adults of P. afra within its Balkan localities


Pooled data




Askion Mts.








Null model













Covariate model






8, 83




11, 80




−Rock −soil

−Scree −stone

15, 168



−Rock −soil

−Scree −stone

12, 79




12, 79




−Tree −shrub

13, 170




9, 82






−/− Inclination

−/−Aspect −gully

16, 167




−Aspect −gully

11, 80



+Inclination +aspect −gully

14, 77





12, 171








+Green plants


14, 169



+Green plants +/−litter

11, 80





Covariate models: Pooled data: ~ + long + poly(lat, 2) + poly(alt, 2) + long * poly(alt, 2) + poly(alt, 2) * poly(lat, 2). Askion Mts: ~ + poly(alt,2) + poly(lat,2) + poly(alt,2)*poly(lat,2). Dalmatia: ~ + poly(alt, 2) + long + poly(lat, 2) + poly(alt,2) * poly(lat, 2) + long * poly(lat, 2)

Symbols + and − indicate the direction of a linear relationship, +/− or −/− the form and direction of a polynomial response. Only significant variables are shown

The picture was practically identical for the Askion data separately. Interestingly, nectar supply had no effect on P. afra abundance. For Dalmatia, only two hypotheses, (i) and (iii), gained statistical support, but the patterns were reverse of those in Greece: P. afra abundance increased, rather than decreased, with rock presence, and increased on steeper and sun-exposed surfaces.

Co-occurring butterfly communities

We recorded 83 butterfly species and 8,260 individuals: 69 from the Askion Mts and 63 from Dalmatia (Online Resource 2). P. afra was the dominant butterfly in both areas, with stronger dominance in Askion (≈70% of 4,987 butterfly individuals recorded) than in Dalmatia (≈20% of 3,273 individuals recorded).

There were fewer species per segment in Askion (mean 9, SD 4.4, range 2–20) than in Dalmatia (mean 12, SD 3.8, range 5–23) (t (DF 182) = −4.63, p < 0.001). The difference became less prominent per whole transects (Askion 20 (SD 6.7, range 10–36), Dalmatia 23 (SD 3.9, range 16–31), t (DF 35.5) = −1.86, p = 0.07). Both areas shared 49 species, 20 species were exclusive to the Askion Mts. and 14 to Dalmatia. A total of 25 Mediterranean species were detected (30% of the total), 23 in Askion (33%) and 12 in Dalmatia (19%). Ten of them occurred in both areas (besides P. afra, these were Euchloe ausonia, Melanargia larissa, Melitaea ornata, Libythea celtis, Pieris ergane, P. mannii, Polyommatus escheri, Pyrgus sidae, Zerynthia polyxena); 13 in Askion Mts only (Anthocharis gruneri, A. pandora, Cupido osiris, Euchloe penia, Erynnis marloyi, Hipparchia senthes, H. volgensis, Melitaea trivia, Leptidea duponcheli, Plebejus sephirus, Polyommatus anteros, Pseudophilotes bavius, Z. cerisyi); two in Dalmatia only (Iolana iolas, Limenitis reducta; the latter, absent on our transects, occurred elsewhere in the Askion Mts).

For pooled data, the selected covariate model was: species composition ~ long 2  + alt*lat + alt + long + alt*long (eigenvalues: 0.32, 0.10, 0.08, 0.04; explained variation 17.5%; first axis: F = 20.3, p = 0.001; all axes: F = 9.5, p = 0.001). The partial DCA ordination (eigenvalues: 0.15, 0.13, 0.11, 0.09; explained 18.9% of the variation; Fig. 3a) containing the above covariates was difficult to interpret, as species usually associated with identical habitat types ended up in both positive and negative extremes of the main gradient (horizontal axis: negative values, e.g., Euchloe penia—scree, Melanargia larissa—grasslands, Zerynthia polyxena—scrub; positive values, e.g. Polyommatus bellargus—rocks, Cyaniris semiargus—grasslands, Aporia crataegi—scrub). The same applied for positive values of the second (vertical) gradient.

Fig. 3

Partial ordinations (controlled for geographic covariates) relating Proterebia afra co-occurrence with other butterflies to habitat conditions of study sites. Scientific names are simplified to species names. a Partial DCA, pooled data, showing mutual positions of 50 best-fitting species in our records. b Partial CCA, pooled data, showing relations of 40 best-fitting butterfly species to habitat conditions. c Partial CCA, Askion Mts, 40 species. Covariate model: species composition ~ long 2  + lat + alt*lat (eigenvalues: 0.16, 0.14, 0.06, 0.18; 21.0% variation explained; first axis: F = 6.0, p = 0.001; all axes: F = 4.7, p = 0.001). Partial DCA: eigenvalues: 0.18, 0.13, 0.11, 0.08; explained variation 22.8%. Predictors (partial CCA): species composition ~ covariate model + inclination + aspect + tree + soil + shrub (eigenvalues: 0.11, 0.06, 0.06, 0.04; explained variation 12.6%; first axis: F = 4.5, p = 0.001; all axes: F = 2.7, p = 0.001). d Partial CCA, Dalmatia, 40 species. Covariate model: species composition ~ alt*lat + alt 2  + alt + lat 2 (eigenvalues: 0.16, 0.07, 0.05, 0.16; 20.5% explained; first axis: F = 6.9, p = 0.001; all axes: F = 4.4, p = 0.001). Partial DCA: eigenvalues: 0.16, 0.14, 0.09, 0.07; 25.2% explained. Predictors (CCA): species composition ~ covariate model + green plants + inclination + scree + tree + litter (eigenvalues: 0.08, 0.06, 0.05, 0.04; 12.5% explained; first axis: F = 4.0, p = 0.001; all axes: F = 2.6, p = 0.001)

The following habitat characteristics were selected in a partial CCA model (Fig. 3b): species composition ~ covariate model + green plants + aspect + litter + gully + rock + scree + inclination + tree + nectar + shrub + soil (eigenvalues: 0.07, 0.05, 0.05, 0.04; explained 8.36% of the variation; first axis: F = 4.8, p = 0.001, all axes: F = 2.4, p = 0.001). Except for stone, all the environmental predictors thus influenced the composition of butterfly assemblages after discounting the variation due to geography. The main canonical gradient distinguished between steep, scree-covered, sometimes shrubby or tree-covered slopes (negative values), and flat surfaces with high cover of herbaceous vegetation (positive values). The second gradient went from sun—exposed rocky sites (positive values) to cooler sites with exposed bare soil and abundant nectar (negative values). The third gradient distinguished shrubby or tree covered sites (positive values) from sites with a high accumulation of dry litter (negative values).

The separate CCA models for the two areas differed slightly in significant variables (Fig. 3c, d) but the overall picture remained consistent. For the Askion Mts, the main gradient distinguished treeless plains (negative values) from slopes containing trees (positive values), and the second gradient was affected by aspect, from the northern-cold ones in positive values to southern—warm ones in negative values. For Dalmatia, the first gradient differentiated steep slopes (negative values) from grassy plains with accumulated dry litter (positive values) and the second gradient differentiated grassy and dry litter covered slopes from sites containing some trees and scree.


Proterebia afra, an Asian steppe element with relic distribution in the Balkans, was studied in two areas hosting its largest populations, characterised by karstic bedrock and climates with cold winters and springs and arid summers—the Askion Mts in Greece and Dalmatia in Croatia. Its early spring flying adults experience episodes of harsh, swiftly changing weather. Within both study areas, the species inhabits a wide range of natural grassland types.

In both areas pooled, P. afra prefers more sun-exposed sites with a high accumulation of dry litter, few shrubs and trees, and low cover of rocks, stones, scree or bare soil, suggesting a preference for neither advanced succession nor sparse vegetation and exposed bedrock. A quite similar picture emerged in separate analyses for the Askion Mts, whereas only a preference for hotter slopes and rocks was apparent for Dalmatia. The lack of a negative shrubs or trees effect for Dalmatia could be due a more advanced successional state of Dalmatian vegetation, with some trees and bushes present at all transect sections (Table 1). The contrasting effects of inclination and aspect revealed a preference for warmer places in Dalmatia than in the Askion Mts, and it is well established that insects often require hotter microclimates near their cooler (i.e., northern) range limits (Thomas et al. 1999).

Grass litter cover is a proxy for grazing pressure (Vermeire and Gillen 2001). Heavy grazing prevents litter accumulation, causing first prevalence of forbs and ultimately, exposing barren bedrock. Under low grazing, a site ultimately succumbs to shrubs and trees. P. afra may prefer sites with grass litter accumulation not only because its larvae develop on grasses, but also because grass litter insulates overwintering larvae (cf. Örvössy et al. 2013; Stuhldreher and Fartmann 2014), buffers the climatic conditions, shelters against some predators (Möllenbeck et al. 2009), and may protect resting adults in early spring, when fresh green vegetation has not yet developed. The preference indicates the need for intermediate levels of grazing, not suppressing the grasses but keeping in check shrubs and trees. In both areas, the speed of woody regrowth is slowed down by the relative continentality of the climate, but shrubby or woody formations would eventually prevail even there, except perhaps on some stony slopes, in the absence of grazing. Slopes inaccessible to animals are covered by sparse oak savanna in the Askion Mts, while sites removed from grazing due to land mines from the 1990s wars are turning into scrub in Dalmatia (e.g., the Knin surroundings). Grazers, either wild roaming or domesticated, have arguably acted as ecosystem engineers throughout Eurasian steppe biome (Johnson 2009; Sandom et al. 2014) and their role may be even more critical in extrazonal insular steppe patches (e.g., Cremene et al. 2005; Pokluda et al. 2012).

Notably, a requirement for intermediate grazing (or mowing) is frequently reported for grassland butterflies from more northerly areas of Europe (e.g., Kruess and Tscharntke 2002; Eichel and Fartmann 2008; Loeffler et al. 2013; Stuhldreher and Fartmann 2014), whereas species associated with other biotopes, such as woodlands, typically prosper in mosaics of various successional stages (e.g., Slamova et al. 2011). On the relic steppe-like grasslands of the Balkans, the patchy presence of accumulated grass litter seems to indicate a grazing pressure not yet suppressing species sensitive to overgrazing.

In both areas, P. afra was the dominant butterfly of the spring phenological aspect. This was more prominent in the Askion Mts, where P. afra reached higher per-segment abundance, combined with lower per-segment butterfly species richness. The higher local abundance in the Greek Askion Mts. might be a case of a negative density-area relationship (Hambäck et al. 2007), a little understood ecological phenomenon, as the Greek area of occurrence is considerably smaller than the Dalmatian one. Alternatively, the lower density in Dalmatia could be due to its existence near (northern) range margins (e.g. Lawton 1993), but such reasoning is hardly applicable for the disjunct range in the Balkans (Sagarin and Gaines 2002). An ecological explanation might be that most of the Dalmatian area of occurrence is an abandoned rural landscape, formed by a mosaic of grasslands, shrubs and trees, whereas the Askion area of occurrence is mainly continuous grassland. Finally, the lower density in Dalmatia might be a result of a more advanced succession due to faster woody plants growth in the more northerly region, perhaps hastened by grazing declines during and after the 1990s wars. We cannot decide between these explanations, as our results originated from two areas only. Data from more easterly parts of P. afra range could further elucidate the species’ requirements.

Regarding associated butterflies, our study covered the spring aspect only (April to June), so parts of the regional species pools remained unrecorded. Still, it is evident that the Askion Mts host a higher proportion of Mediterranean species, as it is situated more southerly, within one of the most biotically diverse regions in Europe–the Greek–Macedonian borderland (cf. Kudrna et al. 2011 for butterflies; Kryštufek and Griffiths 2002 for mammals). Judging from their total distribution ranges, several species co-occurring with P. afra probably represent relics of more easterly distribution: Anthocharis grunneri (ranging from Southern Balkans to Iran); E. penia (reaching Turkey, Syria and Northern Iraq); E. marloyi (from the S Balkans through the Middle East to Transcaucasia); or Pseudophilotes bavius (steppes of NE Balkans, Crimea, Southern Urals) (cf. Pamperis 2009). In Dalmatia, a northerly projection of the Mediterranean biome where the Mediterranean and Euro-Siberian faunas intersect, there were fewer Mediterranean species, but higher total species count and more species per transect. The occurrence of such distinctly Euro-Siberian elements (cf. Kudrna et al. 2011, 2015) as Argynnis niobe, Euphydryas aurinia, or Pararge aegeria directly on P. afra sites was likely facilitated by higher representation of shrubs or trees. These “northern” species occur in mainland Greece as well, but prefer other habitats than the arid Askion Mts grasslands (cf. Pamperis 2009).

The observed situation was supported by ordination analyses (Fig. 3). The main gradients structuring butterfly communities in both pooled and separate data sets distinguished steep slopes with trees from grassland-covered plains. Arguably, the heterogeneous landscapes of the Balkans allow coexistence of open forest and grassland species on very small scales (e.g., the woodland species Satyrium spini and Z. polyxena, together with the grassland species M. larissa and Plebejus idas). The commonly used classifications of habitat specificity of European butterflies (Shreeve et al. 2001; Nylin and Bergström 2009) tend to reflect the situations in more northerly regions, where successional processes are faster and the biota homogenised by intensive land use (Ekroos et al. 2010). In the traditionally used areas of southern Europe, even species with contrasting requirements coexist at very minute scales (Loos et al. 2014; Slancarova et al. 2015).

Conservation message

The areas inhabited by P. afra in both the Askion Mts and Dalmatia are fairly large, the species dominates spring faunal aspects and inhabits a wide range of grassland types in both areas. The population densities seem to be lower in Dalmatia, but this is compensated by considerably larger distribution extent. It therefore seems that the species is not facing imminent threat in any of the regions. There are, however, warning signals, such as the total cessation of grazing and entire land-use abandonment in extensive parts of Dalmatia, the pine planting in Dalmatian karstic areas, or too heavy cattle grazing at mountain tops (both areas). There was a higher representation of trees and shrubs on Dalmatian transects, although our data do not allow deciding whether this is due to recently accelerating succession, or just due to a cooler and wetter climate in the more northerly region. In more northerly Europe, rapid crashes of seemingly common butterflies across large parts of ranges have occurred due to land use change (e.g., Kadlec et al. 2010), and such crashes should be avoided in the biotically richer European South. Furthermore, both areas are hosting other butterfly species with either restricted relic distribution (mainly the Askion Mts), or occurring near their northern range limits in the Balkans (Dalmatia). Some of these species exist there in much lower densities than P. afra, and may display more exacting biotope requirements and might be more susceptible to changing conditions.

Therefore, we propose to monitor the situation of P. afra in both areas closely. It is vital that current grazing levels at the Askion Mts are maintained, and conservation efforts in Dalmatia should lead to grazing re-establishment for the localities where shrub and tree succession has proceeded too far. Financial support for herdsmen would enhance attractiveness of traditional land use, which could, however, lead to overgrazing if not applied carefully. Direct shrub removal could be practised. Pasture management by fire could be used as a single-term method of grassland re-establishment, but this should be applied cautiously as repeated fires affect the biotic communities (Milberg et al. 2014). Deliberate afforestation should be avoided within P. afra’s range. Rewilding (i.e., reintroduction of wild roaming grazers) as a cost-efficient alternative to domesticated animals (e.g., Svenning et al. 2015), is already taking place in some parts of the Balkans (Jepson and Schepers 2016) and could be, in the future implemented in vast areas of P. afra’s Dalmatian distribution.



This work was supported by The Explorers Club Exploration Fund, FAN(B), the University of South Bohemia in Ceske Budejovice (04-168/2013/P and 152/2016/P), and by the Czech Science Foundation (GACR 14-33733S). We would like to thank the Greek Ministry of Environment, Energy and Climate Change (research permit No. 170916/1344), Croatian Ministry of Environmental and Nature Protection (517-07-1-1-1-15-5) and our friends and colleagues who offered their assistance: Z. Faltýnek Fric, Z. Kolev, M. Krausová, R. Kučmerčík, R. Verovnik and M. Zapletal.

Compliance with ethical standards

Conflict of interest

The authors declare that they have no conflict of interest.

Supplementary material

10841_2017_9995_MOESM1_ESM.docx (79 kb)
Supplementary material 1 (DOCX 79 KB)


  1. Allen JRM, Brandt U, Brauer A, Hubberten H-W, Huntley B, Keller J, Kraml M, Mackensen A, Mingram J, Negendank JFW, Nowaczyk NR, Oberhansli H, Watts WA, Wulf S, Zolitschka B (1999) Rapid environmental changes in southern Europe during the last glacial period. Nature 400:740–743. doi: 10.1038/23432 CrossRefGoogle Scholar
  2. Androulakakis N (2012) Geological Map of Greece [Internet]. c2012–2016. Using: ArcGIS [GIS software]. Version 10.0. Redlands (CA): Environmental Systems Research Institute. Accessed 10 Nov 2016
  3. Aufgebauer A, Panagiotopoulos K, Wagner B, Schaebitz F, Viehberg FA, Vogel H, Zanchetta G, Sulpizio R, Leng MJ, Damaschke M (2012) Climate and environmental change in the Balkans over the last 17 ka recorded in sediments from Lake Prespa (Albania/F.Y.R. of Macedonia/Greece). Quatern Int 274:122–135. doi: 10.1016/j.quaint.2012.02.015 CrossRefGoogle Scholar
  4. Cremene C, Groza G, Rakosy L, Schileyko AA, Baur A, Erhardt A, Baur B (2005) Alterations of steppe-like grasslands in Eastern Europe: a threat to regional biodiversity hotspots. Conserv Biol 19:1606–1618. doi: 10.1111/j.1523-1739.2005.00084.x CrossRefGoogle Scholar
  5. de Jong Y, Verbeek M, Michelsen V, Bjørn P de P, Los W, Steeman F, Bailly N, Basire C, Chylarecki P, Stloukal E et al (2014) Fauna Europaea—all European animal species on the web. Biodivers Data J 2:e4034. doi: 10.3897/BDJ.2.e4034 CrossRefGoogle Scholar
  6. Debussche M, Lepart J, Dervieux A (1999) Mediterranean landscape changes: evidence from old postcards. Glob Ecol Biogeogr 8:3–15. doi: 10.1046/j.1365-2699.1999.00316.x CrossRefGoogle Scholar
  7. Deodati T, Cesaroni D, Sbordoni V (2009) Molecular phylogeny, classification, and biogeographic origin of Callerebia and other related Sino-Himalayan genera (Insecta: Lepidoptera: Nymphalidae: Satyrinae). In: Hartmann M, Weipert J (eds) Biodiversität und Naturausstattung im Himalaya III (Biodiversity and natural heritage of the Himalaya III). Verein der Freunde und Förderer des Naturkundemuseums, Erfurt, pp 107–114Google Scholar
  8. Dercourt J, Zonenshain LP, Ricou L-E, Kazmin VG, Le Pichon X, Knipper AL, Grandjacquet C, Sbortshikov IM, Geyssant J, Lepvrier C, Pechersky DH, Boulin J, Sibuet JC, Savostin LA, Sorokhtin O, Westphal M, Bazhenov ML, Lauer JP, Biju-Duval B (1986) Geological evolution of the tethys belt from the atlantic to the pamirs since the LIAS. Tectonophysics 123:241–315. doi: 10.1016/0040-1951(86)90199-X CrossRefGoogle Scholar
  9. Dewey JF, Pitman WC, Ryan WBF, Bonnin J (1973) Plate tectonics and the evolution of the alpine system. Geol Soc Am Bull 84:3137–3180. doi: 10.1130/0016-7606(1973)84<3137:PTATEO>2.0.CO;2 CrossRefGoogle Scholar
  10. Eichel S, Fartmann T (2008) Management of calcareous grasslands for Nickerl’s fritillary (Melitaea aurelia) has to consider habitat requirements of the immature stages, isolation, and patch area. J Insect Conserv 12:677–688. doi: 10.1007/s10841-007-9110-9 CrossRefGoogle Scholar
  11. Ekroos J, Heliölä J, Kuussaari M (2010) Homogenization of lepidopteran communities in intensively cultivated agricultural landscapes. J Appl Ecol 47:459–467. doi: 10.1111/j.1365-2664.2009.01767.x CrossRefGoogle Scholar
  12. ESRI (2011) ArcGIS Desktop: Release 10. Environmental Systems Research Institute, RedlandsGoogle Scholar
  13. Fairbanks RG (1989) A 17,000-year glacio-eustatic sea level record: influence of glacial melting rates on the Younger Dryas event and deep-ocean circulation. Nature 342:637–642. doi: 10.1038/342637a0 CrossRefGoogle Scholar
  14. Hambäck PA, Summerville KS, Steffan-Dewenter I, Krauss J, Englund G, Crist TO (2007) Habitat specialization, body size, and family identity explain lepidopteran density–area relationships in a cross-continental comparison. Proc Natl Acad Sci USA 104:8368–8373. doi: 10.1073/pnas.0611462104 CrossRefPubMedPubMedCentralGoogle Scholar
  15. Hesselbarth G, Oorschot H van, Wagener S (1995) Die Tagfalter der Türkei unter Berücksichtigung der angrenzenden Länder. Goecke & Evers, KeltnerGoogle Scholar
  16. Jepson P, Schepers F (2016) Making space for rewilding: Creating and enabling policy environment. University of Oxford, Oxford. doi: 10.13140/RG.2.1.1783.1287 Google Scholar
  17. Johnson CN (2009) Ecological consequences of Late Quaternary extinctions of megafauna. Proc R Soc B 276:2509–2519. doi: 10.1098/rspb.2008.1921 CrossRefPubMedPubMedCentralGoogle Scholar
  18. Jutzeler D, Lafranchis T (2011) Die Larvalstadien von Proterebia afra pyramus (de Louker) aus dem Norden Griechenlands. Larvalentwicklung der dalmatinischen P. afra dalmata (Godart, 1824) im Vergleich und zur Geschichte des Namens unserer Art (Lepidoptera Nymphalidae, Satyrinae). Entomol Romanica 16:5–18Google Scholar
  19. Kadlec T, Vrba P, Kepka P, Schmitt T, Konvicka M (2010) Tracking the decline of the once-common butterfly: delayed oviposition, demography and population genetics in the hermit Chazara briseis. Anim Conserv 13:172–183. doi: 10.1111/j.1469-1795.2009.00318.x CrossRefGoogle Scholar
  20. Koren T, Trkov D (2011) Proterebia afra damlata (Godart, 1824) (Lepidoptera, Satyrinae) recorded for the first time in Bosnia and Hercegovina. Nat Slov 13:57–58Google Scholar
  21. Koren T, Burić I, Štih A, Zakšek V, Verovnik (2010) New data about the distribution and altitudinal span of the Dalmatian Ringlet, Proterebia afra dalmata (Godart, [1824]) (Lepidoptera: Satyrinae) in Croatia = Novi podatki o razširjenosti in razponu nadmorskih višin pojavljanja dalmatinskega rjavčka, Proterebia afra dalmata (Godart, [1824]) (Lepidoptera: Satyrinae) na Hrvaškem. Acta Entomol Slov 18:143–150Google Scholar
  22. Kruess A, Tscharntke T (2002) Grazing intensity and the diversity of grasshoppers, butterflies, and trap-nesting bees and wasps. Conserv Biol 16:1570–1580. doi: 10.1046/j.1523-1739.2002.01334.x CrossRefGoogle Scholar
  23. Kryštufek B, Griffiths HI (2002) Species richness and rarity in european rodents. Ecography 25:120–128. doi: 10.1034/j.1600-0587.2002.250114.x CrossRefGoogle Scholar
  24. Kudrna O, Harpke A, Lux K, Pennerstorfer J, Schweiger O, Settele J, Wiemers M (2011) Distribution Atlas of Butterflies in Europe. Gesellschaft für Schmetterlingsschutz, HalleGoogle Scholar
  25. Kudrna O, Pennerstorfer J, Lux K (2015) Distribution atlas of European butterflies and skippers. Wiss. Verl. Peks., SchwanfeldGoogle Scholar
  26. Lawton JH (1993) Range, population abundance and conservation. Trends Ecol Evol 8:409–413. doi: 10.1016/0169-5347(93)90043-O CrossRefPubMedGoogle Scholar
  27. Loeffler F, Stuhldreher G, Fartmann T (2013) How much care does a shrub-feeding hairstreak butterfly, Satyrium spini (Lepidoptera: Lycaenidae), need in calcareous grasslands? Eur J Entomol 110:145–152. doi: 10.14411/eje.2013.020 CrossRefGoogle Scholar
  28. Loos J, Dorresteijn I, Hanspach J, Fust P, Rakosy L, Fischer J (2014) Low-intensity agricultural landscapes in transylvania support high butterfly diversity: implications for conservation. PLoS ONE 9:e103256. doi: 10.1371/journal.pone.0103256 CrossRefPubMedPubMedCentralGoogle Scholar
  29. Magyari EK, Chapman JC, Gaydarska B, Marinova E, Deli T, Huntley JP, Allen JRM, Huntley B (2008) The “oriental” component of the Balkan flora: evidence of presence on the Thracian Plain during the Weichselian late-glacial. J Biogeogr 35:865–883. doi: 10.1111/j.1365-2699.2007.01849.x CrossRefGoogle Scholar
  30. Mazzoleni S, Pasquale G di, Mulligan M, Martino P di, Rego F (2008) Recent dynamics of the mediterranean vegetation and landscape. Wiley, ChichesterGoogle Scholar
  31. Mihoci I, Šašić M (2005) New findings of the butterfly Dalmatian Ringlet, Proterebia afra dalmata (Godart, [1824]) (Lepidoptera, Satyrinae) in Croatia. Nat Croat 14:121–129Google Scholar
  32. Mihoci I, Šašić M. (2007) New distribution data on the endemic butterfly Proterebia afra dalmata (Godart, [1824]) (Nymphalidae, Satyrinae) in Croatia. Nat Croat 16:205–210Google Scholar
  33. Milberg P, Akoto B, Bergman K-O, Fogelfors H, Paltto H, Talle M (2014) Is spring burning a viable management tool for species rich grasslands? Appl Veg Sci 17:429–441. doi: 10.1111/avsc.12091 CrossRefGoogle Scholar
  34. Mittermeier RA, Gil PR, Hoffman M, Pilgrim J, Brooks T, Mittermeier CG, Lamoreux J, Fonseca GAB da, Seligmann PA, Ford H (2005) Hotspots revisited: Earth’s biologically richest and most endangered terrestrial ecoregions, 1st edn. Conservation International, Mexico CityGoogle Scholar
  35. Möllenbeck V, Hermann G, Fartmann T (2009) Does prescribed burning mean a threat to the rare satyrine butterfly Hipparchia fagi? Larval-habitat preferences give the answer. J Insect Conserv 13:77–87. doi: 10.1007/s10841-007-9128-z CrossRefGoogle Scholar
  36. Myers N, Mittermeier RA, Mittermeier CG, da Fonseca GAB, Kent J (2000) Biodiversity hotspots for conservation priorities. Nature 403:853–858. doi: 10.1038/35002501 CrossRefPubMedGoogle Scholar
  37. Numa C, van Swaay C, Wynhoff I, Wiemers M, Barrios V, Allen D, Sayer C, Lopéz Munguira M, Balletto E, Benyamini D, Beshkov S, Bonelli S, Caruana R, Dapporto L, Franeta F, Garcia-Pereira P, Karacetin E, Katbeh-Bader A, Maes D, Micevski N, Miller R, Monteiro E, Moulai R, Nieto A, Pamperis L, Peer G, Power A, Sasic M, Thompson K, Tzirkalli E, Verovnik R, Warren M, Welch H (2016) The status and distribution of Mediterranean butterflies. IUCN, MalagaCrossRefGoogle Scholar
  38. Nylin S, Bergström A (2009) Threat status in butterflies and its ecological correlates: how far can we generalize? Biodivers Conserv 18:3243. doi: 10.1007/s10531-009-9640-4 CrossRefGoogle Scholar
  39. Örvössy N, Kőrösi Á, Batáry P, Vozár Á, Peregovits L (2013) Potential metapopulation structure and the effects of habitat quality on population size of the endangered False Ringlet butterfly. J Insect Conserv 17:537–547. doi: 10.1007/s10841-012-9538-4 CrossRefGoogle Scholar
  40. Pamperis LN (2009) The butterflies of Greece. Editions Pamperis, AthensGoogle Scholar
  41. Pamperis LN (2011) The presence of Proterebia afra (Fabricius, 1787) (Lepidoptera: Satyridae) in the Rhodope Mts, NE. Greece. Entomol Gaz 62.Google Scholar
  42. Peña C, Witthauer H, Klečková I, Fric Z, Wahlberg N (2015) Adaptive radiations in butterflies: evolutionary history of the genus Erebia (Nymphalidae: Satyrinae). Biol J Linn Soc 116:449–467. doi: 10.1111/bij.12597 CrossRefGoogle Scholar
  43. Pokluda P, Hauck D, Cizek L (2012) Importance of marginal habitats for grassland diversity: fallows and overgrown tall-grass steppe as key habitats of endangered ground-beetle Carabus hungaricus. Insect Conserv Diver 5: 27–36. doi: 10.1111/j.1752-4598.2011.00146.x CrossRefGoogle Scholar
  44. Pollard E (1977) A method for assessing changes in the abundance of butterflies. Biol Conserv 12:115–134. doi: 10.1016/0006-3207(77)90065-9 CrossRefGoogle Scholar
  45. Polunin O (1980) Flowers of Greece and the Balkans: a field guide. Oxford University Press, OxfordGoogle Scholar
  46. R Core Team (2016) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna.
  47. Radovic J, Civic K, Topic R, Posavec Vukelic V (eds) (2009) Biodiversity of Croatia. State institute for nature protection. Ministry of Culture—Republic of Croatia, ZagrebGoogle Scholar
  48. Roos P, Arnscheid W, Stangelmaier G, Beil B (1984) Praimaginale Merkmale in der Gattung Proterebia Roos & Arnscheid: Beweise fur die phylogenetische Distanz zur Gattung Erebia Dalman (Satyridae). Nota Lepidopterol 7:361–374Google Scholar
  49. Ryan WBF, Pitman WC, Major CO, Shimkus K, Moskalenko V, Jones GA, Dimitrov P, Gorür N, Sakinç M, Yüce H (1997) An abrupt drowning of the Black Sea shelf. Mar Geol 138:119–126. doi: 10.1016/S0025-3227(97)00007-8 CrossRefGoogle Scholar
  50. Sagarin RD, Gaines SD (2002) The “abundant centre” distribution: to what extent is it a biogeographical rule? Ecol Lett 5:137–147. doi: 10.1046/j.1461-0248.2002.00297.x CrossRefGoogle Scholar
  51. Sandom CJ, Ejrnæs R, Hansen MDD, Svenning JC (2014) High herbivore density associated with vegetation diversity in interglacial ecosystems. Proc Natl Acad Sci USA 111:4162–4167. doi: 10.1073/pnas.1311014111 CrossRefPubMedPubMedCentralGoogle Scholar
  52. Sasic M, Mihoci I, Kucinic M (2013) Crveni popis danjih leptira Hrvatske. Drzavni zavod za zastitu prirode, Ministarstvo zastite okolisa i prirode, Zagreb. Accessed 6 Feb 2017
  53. Shreeve TG, Dennis RLH, Roy DB, Moss D (2001) An ecological classification of British butterflies: ecological attributes and biotope occupancy. J Insect Conserv 5:145–161. doi: 10.1023/A:1017556113534 CrossRefGoogle Scholar
  54. Slamova I, Klecka J, Konvicka M (2011) Diurnal behavior and habitat preferences of Erebia aethiops, an Aberrant Lowland Species of a Mountain Butterfly Clade. J Insect Behav 24:230–246. doi: 10.1007/s10905-010-9250-8 CrossRefGoogle Scholar
  55. Slancarova J, Vrba P, Platek M, Zapletal M, Spitzer L, Konvicka M (2015) Co-occurrence of three Aristolochia-feeding Papilionids (Archon apollinus, Zerynthia polyxena and Zerynthia cerisy) in Greek Thrace. J Nat Hist 49:1825–1848. doi: 10.1080/00222933.2015.1006281 CrossRefGoogle Scholar
  56. Slancarova J, Bartonova A, Zapletal M, Kotilinek M, Fric ZF, Micevski N, Kati V, Konvicka M (2016) Life history traits reflect changes in mediterranean butterfly communities due to forest encroachment. PLoS ONE 11:e0152026. doi: 10.1371/journal.pone.0152026 CrossRefPubMedPubMedCentralGoogle Scholar
  57. Stuhldreher G, Fartmann T (2014) When habitat management can be a bad thing: effects of habitat quality, isolation and climate on a declining grassland butterfly. J Insect Conserv 18:965–979. doi: 10.1007/s10841-014-9704-y CrossRefGoogle Scholar
  58. Svenning J-C, Pedersen PBM, Donlan J, Ejrnaesd R, Faurbya S, Galettie M, Hansen DM, Sandel B, Sandom CJ, Terborgh JW, Vera FWM (2015) Science for a wilder Anthropocene: synthesis and future directions for trophic rewilding research. Proc Natl Acad Sci USA 113:898–906. doi: 10.1073/pnas.1502556112 CrossRefPubMedPubMedCentralGoogle Scholar
  59. Ter Braak CJF, Šmilauer P (2012) CANOCO Reference Manual and User’s Guide: Software for Ordination (version 5.0). Biometris, WageningenGoogle Scholar
  60. Thomas JA, Rose RJ, Clarke RT, Thomas CD, Webb NR (1999) Intraspecific variation in habitat availability among ectothermic animals near their climatic limits and their centres of range. Funct Ecol 13:55–64. doi: 10.1046/j.1365-2435.1999.00008.x CrossRefGoogle Scholar
  61. Tolman T, Lewington R (2008) Collins butterfly guide: the Most complete field guide to the butterflies of Britain and Europe. Collins, LondonGoogle Scholar
  62. Tshikolovets VV, Kosterin O, Gorbunov P, Yakovlev R (2016) The butterflies of Kazakhstan. Tshikolovets, KievGoogle Scholar
  63. Tvrtković N, Verovnik R, Lovrenčić L, Vuković M, Šašić M (2015) New contributions to the butterfly fauna of Mt Velebit and the neighbouring area of Lika (Croatia). Nat Croat 24:281–292.CrossRefGoogle Scholar
  64. Van Andel TH, Tzedakis PC (1996) Palaeolithic landscapes of Europe and environs, 150,000–25,000 years ago: an overview. Quat Sci Rev 15:481–500. doi: 10.1016/0277-3791(96)00028-5 CrossRefGoogle Scholar
  65. van Strien AJ, van Swaay CAM, Kery M (2011) Metapopulation dynamics in the butterfly Hipparchia semele changed decades before occupancy declined in The Netherlands. Ecol Appl 21:2510–2520. doi: 10.2307/41416675 CrossRefPubMedGoogle Scholar
  66. Vermeire LT, Gillen RL (2001) Estimating herbage standing crop with visual obstruction in tallgrass prairie. J Range Manag 54:57–60. doi: 10.2458/azu_jrm_v54i1_vermeire CrossRefGoogle Scholar

Copyright information

© Springer International Publishing Switzerland 2017

Authors and Affiliations

  • Alena Bartoňová
    • 1
    • 2
    Email author
  • Vojtěch Kolář
    • 1
    • 2
    • 3
  • Jana Marešová
    • 1
    • 2
  • Martina Šašić
    • 4
  • Jana Šlancarová
    • 1
    • 2
  • Pavel Sucháček
    • 3
  • Martin Konvička
    • 1
    • 2
  1. 1.Faculty of ScienceUniversity of South BohemiaCeske BudejoviceCzech Republic
  2. 2.Institute of EntomologyBiology Centre CASCeske BudejoviceCzech Republic
  3. 3.Faculty of AgricultureCeske BudejoviceCzech Republic
  4. 4.Croatian Natural History MuseumZagrebCroatia

Personalised recommendations