Gap regeneration patterns in Carpathian old-growth mixed beech forests - Interactive effects of spruce bark beetle canopy disturbance and deer herbivory

The combined effects of various disturbance types have lasting consequences on forest ecosystems and the services they provide. We examined inter-specific differences in the responses of European beech, silver fir and Norway spruce to canopy disturbance across four life stages (small seedlings, tall seedlings, saplings and pole-sized trees) in two Carpathian old-growth mixed beech forests located in the Gorce National Park (GNP) and Babia Gora National Park (BGNP), Poland. Both study sites were recently affected by spruce bark beetle outbreak. We compared the composition of tree regeneration between forest plots, canopy gaps and expanded areas (i.e. areas located under tree crowns). Moreover, we studied how various gap properties (size, expanded gap to canopy gap size ratio, shape) and other factors (browsing, plant and advanced regeneration cover) shaped regeneration patterns. Inter-specific differences in species abundance relative to gap properties and intra-specific differences in species response across various life stages to gap properties were found. Gap properties had more pronounced effects on saplings and pole-sized trees than on small and tall seedlings. Gap size had the most noticeable effect on beech across all regeneration classes. However, its effect varied from negative to positive depending on life stage. Pole-sized trees of all species responded positively to gap size. Thus, it seems that large gaps provided an opportunity for all species to recruit to canopy. Fir and spruce responded contrastingly to expanded gap to canopy gap size ratio, which reflects differences in their ability to adapt to high-intensity solar radiation. With the exception of beech, the 'forest-gap' gradient only affected trees in the advanced life stages (saplings and pole-sized trees). Beech and fir sapling densities and fir pole-sized tree densities were higher when under forest canopy, but spruce pole-sized trees were slightly more abundant in expanded areas. Browsing rates differed between the two study sites; they were more severe in BGNP, where deer densities in the last twenty years have been significantly higher. This intense deer browsing kept silver fir regeneration below the height of 0.5 m and prevented any recruitment to sapling stage. We conclude that even though the recent spruce mortality caused by bark beetle outbreak improved light conditions in both study areas, thus providing good regeneration opportunities to both beech and fir, deer browsing may have long-lasting effects on successional patterns by hindering fir recruitment in BGNP.