Re-establishment of grazing is an effective way to restore bryophyte

Re-establishment of grazing is a promising way to restore vascular plant communities in abandoned mesic semi-natural grasslands (Pykälä 2003) and the same conclusion can be made concerning the restoration of bryophyte communities in this biotope (I). Pykälä (2003, 2005) found that the restored grasslands in the very same study region more often resembled the abandoned than the continuously grazed sites in their vascular plant communities five years from the restart of grazing. In this study, the majority of the restored grasslands were similar to the continuously grazed sites in their bryophyte communities after 10-15 years of restorative grazing (I).

However, the recovery pace of bryophytes and vascular plants cannot be compared in this instance, as we do not know how much the vascular plant communities had recovered in the ten years which separates these two studies.

Despite the predominantly promising results of restoration, high variability in the species richness (alpha and gamma) and cover of bryophytes in the restored grasslands indicates that the recovery of bryophyte communities had not been equally successful at every site (I). Some individual restored grasslands were even similar with the abandoned ones in their bryophyte species richness and cover (I). This was probably due to an insufficient grazing intensity in these sites, maintaining intensive competition in the ground layer and preventing frequent soil disturbances typical of traditionally managed grasslands (I). Pykälä (2003) also assessed that low grazing pressure in some of the restored grasslands hampered the recovery of vascular plant communities. Grazing intensity can be regarded as a key factor in the conservation of grassland biotopes (Stewart & Pullin 2008) and comparatively high intensity seems to be essential for the recovery and maintenance of valuable bryophyte communities in mesic grasslands.

Because of the importance of soil disturbances for bryophytes, grazing may be a generally more suitable tool for the restoration of bryophyte communities than mowing (I, Vanderpoorten et al.

2004). Cattle may be the most suitable grazer in the restoration of both bryophyte and vascular plant communities in mesic grasslands, as it consumes graminoids more effectively than sheep (Haeggström 1990). The abundance of a few competitive graminoid species tends to increase in this biotope after abandonment (Pykälä 2005) and these species may restrict the recovery of bryophyte communities via dense turfs and effective litter production (II). The importance of soil disturbances and dung (in moist and shady biotopes) to bryophyte diversity in semi-natural rural biotopes also indicates that mown substitutive environments, such as road verges, may not compensate the loss of grazed biotopes for bryophytes, while they may be more effective in the sustenance of vascular plant diversity (Auestad at al. 2011).

It has been hypothesized that traditional combination of mowing and grazing could be the most effective management practice in the restoration of vascular plant communities in

abandoned mesic grasslands (Hellström et al. 2003). Mowing is usually a more effective way to remove nutrients from grasslands in comparison to grazing (Pykälä 2001) but mowing alone may be ineffective if the lack of soil disturbances limit plant recruitment (Huhta et al. 2001). Undoubtedly, the combination of mowing and grazing would also restore and sustain characteristic bryophyte communities in mesic grasslands. Including also mowing and hay removal in the management may benefit bryophyte communities by decreasing the production of vascular plant biomass in the long term.

In this study, at least one continuously grazed grassland was always found in an immediate vicinity of each restored grassland and the promising results of restoration may reflect a rich regional species pool in the study region. The limited dispersal capacity of species from neighbouring biotopes and from soil seed bank is often posed as an explanation for the incomplete success of restoration of vascular plant communities in grassland biotopes (Stampfli & Zeiter 1999, Mitlacher et al.

2002, Bakker et al. 2002, Aavik et al. 2008, Hellström et al. 2009, Aavik et al. 2013) and bryophytes may also suffer from dispersal limitation. Many pleurocarpous bryophytes, in particular, are slow in dispersal, even when source populations are found in an immediate vicinity (Van Tooren et al. 1990). Shuttle species may also be slow to colonize new grasslands (During 1992), at least when there is no epizoochorous dispersal by grazing animals (Pauliuk et al. 2011). Instead, colonists are adapted to an effective spreading by their light spores (During 1992, Hutsemekers et al. 2008). The role of soil propagule bank in the establishment of bryophytes is insufficiently known although the high investment of some short-lived acrocarpous species (shuttle species) in large spores implies its importance (During

& Willems 1986, During 1992). However, this issue was not tackled in more detail in this study.

The findings of this study are potentially applicable to the restoration of other open or sparsely wooded mesic semi-natural rural biotopes. Of course, grazing intensity has to be adjusted separately in each biotope and site, keeping in mind that

insufficient grazing intensity can hamper the recovery of both bryophyte (I) and vascular plant communities (Kotiluoto 1998, Pykälä 2003). However, many other species groups demand lower grazing or mowing intensities (Söderström et al. 2001, Kruess & Tscharntke 2002, Vessby et al. 2002, Hoste-Danyłow et al. 2010). Hence, management has to be both continuous in time and variable in space (Vessby et al. 2002, Pöyry et al. 2004, 2006).

With the aid of the field inventories of different taxonomic groups, management efforts and different management practices can ideally be directed to the sites where their effects are optimal (WallisDeWries et al. 2002). The results of this thesis may be less useful in the restoration and management of the moistest and driest ends of semi-natural rural biotopes, as the relationship between bryophyte diversity and management may deviate from what found in this study (e.g. Löbel et al. 2006).

3.3.2 Grassland connection may be more serious threat to