Peatland restoration: A mossy affair? : The similarities of the vegetation communitiesbetween restored bogs and natural bogs.

Abstract: Peatlands function as important global carbon sinks as peatlands store 30% to 40% of the total soil carbon even though peatlands only cover 3% of the Earth's terrestrial surface. However, drainage activities have damaged the ecosystem and caused many peatlands to function as carbon sources instead. Restoration projects aim to recover the hydrology and Sphagnum moss populations in peatlands by rewetting and removing trees, respectively. However, long-term studies on the recovery of restored bogs (a type of peatland) are scarce, and more so from a functional point of view. Here, the similarities of the environmental characteristics, and vegetation communities in terms of species composition and functionality (i.e., plant functional traits and functional diversity) between two restored bogs and the adjacent reference bogs (the two study sites ‘Anderstorp’ and ‘Store Mosse’) in southern Sweden were compared, eight years post-restoration. The pH significantly differed between areas (i.e., restored or reference) in both sites but the differences were small. The electrical conductivity (EC, i.e. concentration of ions in the water) was higher in the restored areas than in the reference areas, and Anderstorp had higher EC levels in general than Store Mosse. The effect of restoration on the water table depth (WTD) between the areas depended on the site. At Store Mosse, the WTD was closer to the surface in the restored area than the reference area. For both sites, the species composition was significantly different between areas and the Sphagnum moss coverages were higher in the natural areas than in the restored areas. In addition, no environmental variable (pH, EC or WTD) seemed to act as the dominant driver for the vegetation composition. The plant trait analyses were based on vascular bog plants and both sites showed similar trends between areas for 7 out of 8 traits, with a larger effect size at Store Mosse.  For the traits seed mass, leaf dry matter content (LDMC), leaf nitrogen (N) content and leaf phosphorus (P) content, significant differences were found between areas in which the restored areas had a lower mean than the reference areas. In contrast, for the traits specific leaf area (SLA) and onset of flowering, significant differences were found between areas in which the restored areas had a higher mean than the reference areas. For the trait canopy height, no significant differences were found between area, sites or their interaction. Furthermore, the natural bogs of the two sites had similar means for all traits, except leaf N content, which was higher at Store Mosse. Only for the trait leaf N:P ratio, the effect of restoration depended on the site in which the restored areas had a lower mean than the natural areas at Store Mosse. The functional diversity (i.e., functional richness and functional evenness) did not significantly differ between areas, sites or their interaction for functional richness, whereas the functional evenness significantly differed between areas at Store Mosse. In conclusion, the restored areas are partly similar to the reference areas in both sites in terms of environmental characteristics, but the vegetation composition in the restored areas are not similar to the reference areas at both sites from a species abundance point of view nor a functional point of view, albeit this only reflects on the vascular plants and the included traits.