, 2005). Overall, birch accounted for 56% of regenerating saplings in our study. The density of birch regeneration on clearfelled upland moorland on our study sites is similar to that recorded in a storm damaged lowland conifer site in Britain (Harmer and Morgan, 2009) and to clearfelled upland conifer sites in Scotland (Wallace, 1998). Despite the presence of mature individuals of ash, beech, juniper and hazel adjacent to clearfelled sites only a handful of saplings of these species were noted. see more Overall we found that pioneer, shade-intolerant species such

as birch, rowan and willow regenerated more frequently than shade-tolerant species such as beech and holly (Brzeiziecki and Kienast, 1994). We explored the role of distance from seed source on regeneration density for distances up to 100 m from the source. The regeneration of the small-seeded and wind-dispersed alder and birch species were found to be strongly dependent on the distance from parent trees. The majority of the saplings were

found within 20 m of a parent tree, although for birch there was a long tail, limited in our study to the width of the clearfelled site. The patchy distribution which results from this clumping around seed sources is not necessarily a disadvantage for establishment of natural woodland. Rodwell and Patterson (1994) suggest that 20–50% of woodland sites should be retained as open ground to enhance structural diversity and wildlife this website value. The fluctuations in sapling density may result in a more natural woodland structure to that produced through planting. The shoulder of the regeneration curve at distances less than 10 m from the woodland edge could be attributable to an edge effect – root competition

or light and rain interception from the mature trees counteracting the increased regeneration caused by the rise in seed density as you approach the edge. The seed dispersion curve for a point source (Harper, 1977 and Nathan et al., 2001) is similarly shaped to the regeneration curves for solitary trees in having a peak in seed fall density a short distance from the parent tree. Regeneration of oak Methamphetamine and rowan was found to be significantly clumped although not significantly dependent on distance from the seed source. Rowan is primarily dispersed through ingestion by birds, particularly various thrush species (Raspe et al., 2000), while oak relies on hoarding by both birds and mammals but especially Garrulus glandarius (jay) and Apodemus sylvaticus (wood mouse) ( Forget et al., 2004), both of which occur at the study sites. The distribution of regenerating saplings will therefore be partly controlled by the behaviour of the dispersing animal. Previous work in central Europe has demonstrated that the majority of oak regeneration occurs within 100 m of a seed source and declines rapidly at greater distances ( Mirschel et al., 2011).