The study site consisted of several stream valleys and hill slopes covered with a mosaic of grazed pastures, wet meadows, spruce woodlands and shrublands. The potential vegetation is spruce forest that was cleared for agriculture and turned into mesophilous grasslands. Due to low grazing pressure 5-30% of the pastures’ area was overgrown by Corylus avellana. Around these shrubs, a typical forest-edge vegetation developed with Brachypodium pinnatum and Chamaecytisus triflorus. Mark-recapture studies were carried out at three grazed pastures, where butterfly density was highest. All three sites were located on south-facing slopes and moderately grazed by cattle and sheep. Size of site A was ca. 5 ha, site B 2 ha and site C 1 ha, and they were within 600 metres from each other. Chamaecytisus triflorus covered approximately 10-30% of the area.
The potential vegetation of this north-facing hillside is beech forest, whereas the present vegetation is acidofrequent dry grassland dominated by Agrostis capillaris. Since grazing pressure was relatively high, bush cover was low and Chamaecytisus triflorus was rare (< 5%). Besides, the area was frequently visited by tourists.
The potential vegetation of this south-facing hillside is beech forest, whereas the present vegetation is acidofrequent dry grassland dominated by Brachypodium pinnatum with some elements of forest-steppes, like Aster amellus and Peucedanum sp. Due to cessation of grazing several years ago, high proportion of the area was covered by bushes in 2011. The potential food plant here is Cytisus albus reaching 10-20% cover.
The potential vegetation of this north exposed hillside is Quercus cerris-Quercus robur forest on the slopes, Quercus pubescens scrub at the hilltop and beech forest at lower altitudes. The present vegetation in 2011 was a mosaic of closed rocky grasslands dominated by Carex humilis and species-rich mountain hay meadows at places with thicker soil layer. Bush cover was high, although grazing intensity was relatively high as well. Cytisus albus occurred in closed rocky grasslands and Chamaecytisus triflorus in mountain hay meadows, but their cover was below 5% due to intensive grazing.
The potential vegetation of this hillside is Quercus pubescens scrub and Quercus cerris-Quercus robur forest, whereas by 2011 the vegetation became a mosaic of closed rocky grasslands dominated by Stipa pennata and patches of steppic grassland. Due to cessation of grazing several years ago, high proportion of the area was covered by bushes. Here, the potential food plant was Cytisus albus, but its coverage was very low (< 5%).
The potential vegetation of this southeast-facing steep hillside ia probably Quercus pubescens scrub, whereas in 2011 the vegetation was a closed rocky grassland dominated by Stipa dasyphylla at the hilltop and steppic grasslands on the slopes. Bush cover was high due to abandonment. Both Cytisus albus and Chamaecytisus triflorus occurred, but their cover was only 5-10%.
This site was located on a slope of a 1281 m high mountain. The potential vegetation, i.e. closed rocky grassland dominated by Stipa pennata was still the most widespread vegetation type, and dry grassland was present in place of former beech forests at the hillfoot. Bushes were frequent at more humid places. Both Cytisus albus and Chamaecytisus triflorus occurred with an average cover of 10-20%.
The potential vegetation of this west-facing hillside was beech forest, which was still surrounding the study site in 2011. The vegetation was mesophilous grassland dominated by Brachypodium sylvaticum. The habitat was moderately grazed and bush cover was higher at the foot of the hillside. The potential food plant was Chamaecytisus triflorus covering only 5-10% of the area.
The potential vegetation is beech forest, but the area was covered by mesophilous grasslands in 2011 with extensive clones of Brachypodium sylvaticum. The habitat was lightly grazed and patchily encroached by bushes. The potential food plant was Chamaecytisus triflorus covering less than 10% of the area.
Set 1: direction + grazing animal + grazing intensity + slope steepness
Set 2: direction + grazing animal + grazing intensity + host plant density
Set 3: direction + grazing animal + grazing intensity + bush cover
Set 4: direction + grazing animal + vegetation height + slope steepness
Set 5: direction + grazing animal + vegetation height + bush cover
Set 6: direction + grazing animal + vegetation height + host plant density
Supplementary Table 1
Suplementary Fig. 1
Supplementary Fig. 2
Supplementary Fig. 3
Supplementary Fig. 4
Supplementary Fig 5