Open Access

The genus Sphagnum (L.) belongs to the Bryophyte plant division and includes 150 to 400 species. As all mosses Sphagnum has no roots and can hardly regulate its water uptake. As long as enough water is available Sphagnum can grow nearly unlimited while the lower, older parts die off and may accumulate as peat. Single Sphagnum species are able to build up an acrotelm as a hydrological self-regulating mechanism of a bog, a type of intact peatland (mire) only fed by precipitation. Because Sphagnum dominates nearly half of the peatlands in the world, it is one of the globally most important peat formers. Sphagnum biomass is an important raw material for many valuable products, but in a much larger scale Sphagnum is used in its fossil state – as Sphagnum peat. With a consumption of c. 40 million m³ per year globally, Sphagnum peat is the predominant raw material for horticultural growing media. To get Sphagnum biomass it is currently collected from wild populations, to get Sphagnum peat it is extracted from bogs. By far, more peatlands (including bogs) are subjects to drainage for agri- and silvicultural use since centuries, which harms their ecosystem services, including their typical biodiversity, carbon storage capacity, water regulation function and palaeo-environmental archive. In Europe, c. 25 % of all peatlands are used for agriculture, in Germany more than 80 %. Globally drained peatlands cover 0.4 % of land surface but produce 5 % of all anthropogenic greenhouse gas emissions. Sphagnum farming aims to cultivate Sphagnum biomass on rewetted degraded bogs as a new agricultural crop. Sphagnum farming is paludiculture and contributes to the protection of bogs and their peat by conserving the peat body through rewetting and by offering a climate-friendly alternative to fossil peat in horticulture. Next to climate change mitigation, Sphagnum farming has benefits for nutrient retention and biodiversity conservation. This thesis contributes to the development of Sphagnum farming by studying the conditions under which Sphagnum may reach maximal growth. Under (semi)controlled glasshouse conditions, we tested the effects of different water regimes and fertilisation levels on the productivity of various Sphagnum species. On a 1260 m² large irrigated field on cut-over bog in Lower Saxony (Germany) we studied length increase, biomass productivity and tissue nutrient content of Sphagnum over a period of 10 years. Finally, we reviewed all scientific literature and practical experiences with respect to Sphagnum farming worldwide as a first step towards a science-based implementation manual. The main conclusions of our studies are: 1. It is possible to cultivate Sphagnum on rewetted cut-over bog and on rewetted former bog grassland. 2. The rapid establishment of a closed, highly productive Sphagnum lawn requires the deployment of a loose, >1(–5) cm thick Sphagnum layer (80–100 m³ of Sphagnum founder material per hectare) at the start of the growing season (when long frost periods are no longer probable) and adequate water supply. 3. Water table management must be very precise until a dense, well-growing Sphagnum lawn has established. For highest yields the water table should rise with Sphagnum growth and be kept a few centimetres below the Sphagnum capitula. Water supply via open irrigation ditches seems to function better than via subsurface irrigation pipes. 4. Fertilisation does not increase Sphagnum productivity on sites with high atmospheric nitrogen deposition and irrigation with phosphate-rich surface water from the agricultural surroundings. To avoid growth reduction a balanced stoichiometry is important. 5. From all studied species, Sphagnum fallax has the highest productivity. Its fast decomposition and low water holding capacity, however, may make this species less suitable for use in horticultural substrates. 6. Vascular plant cover on Sphagnum production fields can be kept low (<50 % cover) by regular mowing. Higher covers retard Sphagnum growth and reduce its quality for growing media. 7. Pathogenic fungi occurred far more in the glasshouse than in the field and have to be controlled for highest Sphagnum yields. We found Sphagnum vitality and growth rate to be stimulated by high water levels, where Sphagnum is less vulnerable to fungal or algal infection despite high nutrient loads. 8. The rate of Sphagnum biomass accumulation may remain constant over at least 4–5 years after establishing a Sphagnum production field with sufficient water supply. At dry conditions Sphagnum biomass accumulation is lower as a result of lower biomass productivity and higher decomposition rates.

Abstract Aim Species–area relationships (SARs) are fundamental scaling laws in ecology although their shape is still disputed. At larger areas, power laws best represent SARs. Yet, it remains unclear whether SARs follow other shapes at finer spatial grains in continuous vegetation. We asked which function describes SARs best at small grains and explored how sampling methodology or the environment influence SAR shape. Location Palaearctic grasslands and other non‐forested habitats. Taxa Vascular plants, bryophytes and lichens. Methods We used the GrassPlot database, containing standardized vegetation‐plot data from vascular plants, bryophytes and lichens spanning a wide range of grassland types throughout the Palaearctic and including 2,057 nested‐plot series with at least seven grain sizes ranging from 1 cm2 to 1,024 m2. Using nonlinear regression, we assessed the appropriateness of different SAR functions (power, power quadratic, power breakpoint, logarithmic, Michaelis–Menten). Based on AICc, we tested whether the ranking of functions differed among taxonomic groups, methodological settings, biomes or vegetation types. Results The power function was the most suitable function across the studied taxonomic groups. The superiority of this function increased from lichens to bryophytes to vascular plants to all three taxonomic groups together. The sampling method was highly influential as rooted presence sampling decreased the performance of the power function. By contrast, biome and vegetation type had practically no influence on the superiority of the power law. Main conclusions We conclude that SARs of sessile organisms at smaller spatial grains are best approximated by a power function. This coincides with several other comprehensive studies of SARs at different grain sizes and for different taxa, thus supporting the general appropriateness of the power function for modelling species diversity over a wide range of grain sizes. The poor performance of the Michaelis–Menten function demonstrates that richness within plant communities generally does not approach any saturation, thus calling into question the concept of minimal area.

Abstract Aims Pinus uncinata is the major treeline‐forming species in the Pyrenees. Yet, the role of its reproduction and dispersal as drivers of treeline dynamics remains unknown. Here we quantify seed production, dispersal and germination changes along the elevation gradient to assess whether they may constrain the foreseen treeline advance in the Pyrenees. Location Central Pyrenees, Catalonia, NE Spain. Methods We established four plots along an elevation gradient from the closed subalpine forest to the krummholz zone at five study sites. In each plot, we collected cones from five to six trees, measured their length, and triggered their opening in the laboratory to count the number of empty seeds and the number and weight of full seeds. We used the collected seeds in a germination experiment under controlled conditions in growth chambers. Additionally, we installed seed traps along the forest–alpine grassland transition to measure seed rain for three consecutive years in three of the study sites. Results The number of full seeds per cone decreased along the elevation gradient and was correlated with cone length. However, the proportion of full seeds per cone and their weight did not differ between elevation positions. Seed rain decreased drastically with elevation and no seeds arrived into the alpine grassland traps consistently across study years. Although germination success did not significantly differ between elevation provenances (i.e., elevation position of origin), we found significant differences in germination dynamics between study sites and between elevation provenances within sites. Conclusions Our results indicate that whereas the viability of Pinus uncinata seeds is not limited by elevation, seed production and dispersal are constraining the ongoing rates of treeline advance in the Pyrenees.

Species have to cope with climate change either by migration or by adaptation and acclimatisation. Especially for long-living tree species with a low seed dispersal capacity (e.g. European beech, hereafter called beech), the in situ responses through genetic adaptation and phenotypic plasticity play an important role for their persistence. Beech, the dominant climax tree species in Central Europe, shows a high drought sensitivity and its distribution range is expected to shift northwards. On the other hand, projected northward shifts need to be taken with caution, as some studies suggest a sensitivity of beech to frost events in winter and spring. However, studies on the growth performance of cold-marginal beech populations are still rare. Previous studies on beech populations found local adaptation to drought and phenotypic plasticity in fitness-related traits as well as phenological traits. However, studies on the regeneration of beech under natural conditions are yet missing, although germination and establishment of young trees are a very first selective bottleneck and are crucial for tree population persistence and for successful range shifts. This PhD-thesis aimed to identify the potential of plasticity and local adaptation in the important early life-history traits germination, establishment after the 1st year, and survival after the 2nd year in a reciprocal transplantation experiment at 11 sites across and even beyond the distribution range of beech (Manuscript 1). Moreover, this thesis investigated the climate sensitivity and the adaptation potential of beech populations by conducting dendroecological studies along a large climatic gradient across the distribution range (Manuscript 2) and along a strong winter temperature gradient towards the cold distribution margin in Poland (Manuscript 3). In addition, the impact of local climatic singularities was studied in a local study at the southern margin (Manuscript 4). Warm and dry conditions limited natural regeneration, which was indicated by very low survival of young trees, even though germination rates increased with increasing temperature (Manuscript 1). This was also the case in parts of the distribution centre due to the hot and dry conditions in 2018. Although the transplantation experiment revealed high plasticity in the early life-history traits, this plasticity might thus not buffer against climate change under dry conditions. Local adaptation was not detected for any of these traits along the climatic gradient. In contrast, the results of the dendroecological study across the gradient (Manuscript 2) hint towards an adaptation potential of adult trees to drought at the southern margin. Thus, adult trees seemed to be adapted to drought at the southern margin, whereas tree growth in the distribution centre was sensitive to drought. These results indicate that parts of the centre may become ecologically marginal with increasing drought frequency in times of climate change. Interestingly, Manuscript 4 shows that beech growth was positively influenced by frequent fog immersion at the southern distribution margin in north-eastern Spain. This study underlines the importance of local climatic singularities, as they may allow marginal populations to grow in climate refugia in an otherwise unfavourable climate. At the cold distribution margin, the study in Manuscript 1 found a remarkably higher survival of young trees in Sweden than in Poland. Moreover, the dendroecological studies revealed that beech was hampered by both drought at the cold-dry margin (Manuscript 2) and by winter cold at the cold-wet margin in Poland (Manuscript 3). All these results highlight the importance to study climate sensitivity of adult trees and the response of early life-history traits at the cold margin with a more differentiated view comparing cold-dry against the cold-wet populations and growing conditions. However, the high plasticity of the early life-history traits may allow for an increasing germination rate with climate warming at the northern margin and may thus facilitate natural regeneration there. In contrast, the dendroecological studies suggest that adult trees at the cold distribution margin may suffer either from drought or from winter cold and that the risk for spring frost may increase. Thus, the often-predicted compensation of dry-marginal population decline by a northward range expansion should be discussed more critically. In conclusion, my PhD thesis provides new knowledge about the potential of natural regeneration and about climate sensitivity of adult trees across the distribution range of beech. Moreover, it underlines the importance to study both the young tree stages as well as adult trees to assess the performance and vulnerability of tree species under climate change, as both showed differences in their response to changing environmental conditions.

Over thousands of years, peatlands around the world have accumulated carbon (C) stocks of global importance. Drainage for agriculture, forestry and peat extraction has transformed many peatlands from long-term sinks into strong sources of carbon dioxide (CO2). Peat extraction is worldwide responsible for about ten percent of drained peatlands and is mainly carried out in northern countries and Eastern Europe. In Belarus, 0.3 Mha of peatlands are drained for peat extraction, which is twelve percent of the country's peatland area. From 2006 to 2013, 21,333 ha of this area have been rewetted to protect these peatlands from fire and further degradation, reduce their greenhouse gas (GHG) emissions, turn them back into C sinks and promote biodiversity. A further 260,000 ha are no longer used for peat extraction and their rewetting would be a great benefit for nature conservation and climate protection. Rewetting of abandoned peat extraction areas usually leads to inundation of large areas where not adapted plants die and new species establish, depending on water level and nutrient conditions. Beavers, of which there are many in Belarus, also play an important role in the rewetting of peatlands. They dam up ditches in drained and rewetted peatlands, thus contributing to water level increases and vegetation changes. The aim of this PhD thesis was to investigate the impact of inundation on vegetation and GHG emissions in formerly extracted fens in Belarus, to determine the role of water level in this process, and to study whether such fens develop back into C sinks with an almost neutral GHG balance within one or two decades after rewetting (Papers II and III). Also the potential of beaver activities for peatland restoration was assessed (Paper III). Two very different fens, rewetted after peat extraction, were chosen as study areas. The first one, Giel'cykaŭ Kašyl, is a former flood mire and was rewetted with water from the Jasiel'da River in 1985. During the study period 2010–2012 this site was a shallow lake (~ 1 m deep) dominated by very productive, tall reed. Shallower areas along the edges had a partly floating vegetation cover of cattail (Typha latifolia, T. angustifolia) and sedges (Carex elata, C. vesicaria). The second fen, Barcianicha, is fed by groundwater. Rewetting from 1995 onwards resulted in water levels at or slightly above surface and a lower nutrient availability compared to Giel'cykaŭ Kašyl'. This was reflected in the establishment of mesotrophic communities of Eriophorum angustifolium and Carex rostrata. Phragmites australis stands, which were also dominant here, were shorter and less productive than in Giel'cykaŭ Kašyl'. The southern area of Barcianicha was not used for peat extraction and has not been rewetted. Until 2009 vegetation of this part was characterized by forbs (Urtica dioica) and wet meadows (Agrostis stolonifera). From autumn 2009, a beaver dam in the main drainage ditch caused flooding of these areas and led to diverging vegetation development depending on water levels. Within the framework of this doctoral thesis annual fluxes of CO2, methane (CH4) and nitrous oxide (N2O) and the development of water levels and vegetation were monitored for two years at nine sites and evaluated (Papers II and III). Three of the sites, respectively, were located (a) in Giel’cykaŭ Kašyl’, flooded in 1985, (b) in the central area of Barcianicha, which was rewetted in 1995, and (c) in the southern part of Barcianicha, which was flooded by beavers end of 2009. GHG measurements were carried out with manual chambers from August 2010 to September 2012. Annual net CO2 exchange rates (NEE) were modeled based on light response curves of gross primary production (GPP) and on temperature response curves of ecosystem respiration (Reco), which were determined every third to fourth week by alternating measurements with transparent (cooled) and opaque chambers (both with fan) along the daily amplitude of photosynthetically active radiation (PAR) and temperature. Annual CH4 emissions were calculated mainly based on the temperature response of CH4 fluxes over the course of the year, based on biweekly (in summer) to monthly (in winter) repeated single measurements with opaque chambers (without fan). This was done, although all longer rewetted sites were dominated by aerenchymatic plants whose gas transport during the vegetation period may change over the course of the day and can be influenced by shading. This might apply to the six longer rewetted sites, two of which were dominated by Phragmites australis, and the others by Typha latifolia, Carex elata, Carex rostrata or Eriophorum angustifolium. For these six sites therefore studies on the daily course of CH4 release and the influence of chamber shade were conducted, covering 8–24 hours and lasting at least from sunrise to afternoon. Also the extent to which flux rates were affected by a lack of chamber headspace mixing by fans was investigated in the mentioned studies (Papers I and II). The daytime course of CH4 emissions showed a pronounced dynamic for Phragmites australis in both fens, with minimum release during the night and maximum during the day (Paper I). The other sites in contrast did not show a significant diurnal CH4 flux dynamic (Paper II). Lack of headspace mixing by fans as compared to chambers with fan resulted in a slight underestimation of CH4 emissions at very high chambers (220 and 250 cm), as used for Phragmites australis in Giel'cykaŭ Kašyl', while there was no difference at lower chambers (≤185 cm), as used for the other sites. Opaque chambers resulted for sites dominated by Typha latifolia and Carex elata in significantly (1.2 times and 1.1 times, respectively) lower CH4 fluxes compared to transparent chambers. For the other sites, opaque chambers did not significantly reduce CH4 emissions. This result was unexpected, especially for Phragmites australis, as PAR out of all parameters tested had the strongest influence on CH4 emissions from both reed sites, and clouds directly led to reduction of their emissions. Presumably the gas flow in the reed shoots located within opaque chambers was maintained by shoots outside the chamber that were connected to the enclosed shoots by rhizomes (Paper I). The investigations showed that single measurements between 9 a.m. and 6 p.m. with opaque chambers without fan, as performed for the determination of annual CH4 fluxes, resulted for Carex rostrata and Eriophorum angustifolium in estimates similar to the daily mean, but for Phragmites australis in estimates that were rather above the daily mean. Annual CH4 fluxes from Phragmites australis could therefore be slightly overestimated. CH4 fluxes from Typha latifolia and Carex elata during the vegetation period were corrected by a factor of 1.2, although darkness inside of opaque chambers matters only at day, not at night. Daily and annual CH4 fluxes from these sites have been therefore most likely slightly overestimated, too. Water saturation and the establishment of adapted vegetation were the most important conditions for the restoration of C sinks (gaseous CO2 and CH4 fluxes) in the investigated peatlands. The only site with falling water levels in summer and thus temporarily aerated peat was the beaver flooded forbs (Urtica dioica) site at Barcianicha. This site was a very strong CO2 emitter and the only significant N2O source of the entire study (Paper III). All other sites were permanently wet, had much lower CO2 emissions or were even net C sinks (Papers II and III). Establishment of adapted vegetation depended on inundation depth and time since rewetting. For example, within one year the meadow site in Barcianicha shallowly flooded by beaver was colonized by Carex rostrata and other adapted helophytes and developed into a CO2 sink, while the deeper flooded site at the same meadow initially attracted only Chara and some individuals of Alisma plantago aquatica and remained a moderate CO2 source. However, the results of the longer rewetted sites show, that also deeply (~ 1 m) flooded fen areas can become densely populated with mire plants in the course of 25 years and develop into net C sinks. Highest annual C uptake in both fens was achieved by the reed sites. Eriophorum angustifolium and Carex rostrata in mesotrophic Barcianicha were smaller C sinks. Typha latifolia and Carex elata in the eutrophic Giel'cykaŭ Kašyl', on the other hand, released CO2, presumably because the high and fluctuating water levels imposed stress to the plants, and because the large supply of nutrients and dead plant material allowed for strong heterotrophic respiration (Paper II). The simultaneously high CH4 emissions made Typha latifolia and Carex elata major sources of GHG. CH4 emissions from Phragmites australis in Giel'cykaŭ Kašyl' were even higher, but due to extremely high CO2 uptake the site was only a small net GHG source. CH4 emissions in Barcianicha were much lower and comparable to undisturbed sedge fens. The difference between Giel'cykaŭ Kašyl' and Barcianicha was mainly due to the different nutrient supply and the related productivity of the plants. Important conclusions are that stable inundation is an appropriate measure for restoration of the C sink of formerly extracted fens, but nutrient input with water needs to be stopped or reduced in order to decrease CH4 production. If this is not possible, establishment of Phragmites australis and other strong C sinks could help to compensate for the climate impact of high CH4 emissions from eutrophic sites. The effect of the beaver dam on the development of the southern part of Barcianicha depended not only on the initial situation but mainly on the water level. Under optimal conditions, it led to the rapid establishment of adapted mire plants, the restoration of a C sink and a significant reduction of GHG emissions. However, this situation in the shallowly flooded meadow was achieved by chance. In comparison to planned rewetting measures, which aim to raise the water level evenly over the entire peatland, beavers dam ditches in order to improve their immediate habitat, thus influencing water levels only up to a certain distance, but rarely over the entire peatland. Nevertheless, beaver activity is of high value both for mire conservation projects, where existing dams are supplemented by beaver dams, and for abandoned, drained peatlands, like former peat extraction areas in Belarus, many of which at least partially have been rewetted by beavers.