Keywords

1 Introduction

Boreal and temperate forests cover 1210 and 680 million ha, respectively (Keenan et al. 2015). In contrast to tropical forests, whose extent is decreasing due to current deforestation activities resulting in huge emissions of greenhouse gases (Roman-Cuesta et al. 2016), the area of boreal forests remained constant, while the area of temperate forests slightly increased in the last 25 years at an average rate of 2.7 million ha a−1 (Keenan et al. 2015). In total, boreal and temperate forests cover approximately 13% of the global terrestrial land surface.

Temperate and boreal forests are known to provide a wide range of ecosystem services (e.g., Gamfeldt et al. 2013), including timber production, water regulation, soil protection and erosion control, support of biodiversity, or recreation. The role of forests in regulating the climate has been also well acknowledged, due to their strong potential for sequestering atmospheric CO2 in its biomass and soils (De Vries et al. 2003; Vesterdal et al. 2008). In contrast to CO2, the role of forests as both significant sinks and sources of other powerful greenhouse gases, i.e., CH4 and N2O has received comparatively little attention. The crucial role played by forests in regulating nutrient cycling is only possible due to the microbial-mediated transformation processes of the soil organic matter, which make nutrients available again for plant metabolism while also resulting in a substantial release of CO2, as well as CH4 and N2O, to the atmosphere.

Temperate and boreal forests represent one of the major global pools of carbon (C) and nitrogen (N), with more than half of the C being stored in soils (Batjes 1996). Pan et al. (2011) estimated that boreal forest ecosystems store approximately 271 ± 22 Pg C, distributed in the living (54 Pg C) and dead biomass (43 Pg C) and in the soils (down to 1 m) (175 Pg C). This estimate excludes some deep organic boreal forest soils, which explains the significant difference with a recent estimate by Bradshaw and Warkentin (2015) (average: 1096 Pg C, range: 367–1715 Pg C), who included peats, or assessments from IPCC (2007) (471 Pg C). However, all estimates agree that 2/3 to ¾ of all C is stored in soils and peats. For temperate forests, Pan et al. (2011) estimated that the living and dead biomass pool is 62 Pg C, approximately equal to the amount of C stored in soils down to 1 m (57 Pg). Global amounts of N in soils down to 1 m are estimated to be 133–140 Pg (Batjes 1996), while only 10 Pg of N is held in the global plant biomass (Davidson 1994). Figure 1 shows the distribution of forests (Fig. 1A), the soil N stocks down to 1 m (Fig. 1B), and the C:N ratio of these soils (Fig. 1C) for the temperate and boreal zones of the northern hemisphere. The areal extent considered is based on the Olson ecoregions (Olson et al. 2001), and the relative forest cover percentage (Fig. 1A) is based on the GlobCover 2009 (Bontemps et al. 2011). Spatially explicit soil C and N stocks were derived from the ISRIC-WISE soil map (Batjes 2012). According to this approach, the total soil N stocks in boreal and temperate forest soils are 10.4 Pg N and 7.2 Pg N, respectively.

Fig. 1
figure 1

Distribution of boreal and temperate forests (Panel A), the total nitrogen in soils (Panel B), and C:N ratio of soils (Panel C). The areal extent considered is based on the Olson ecoregions “temperate broadleaf and mixed forests” and “temperate coniferous forest” (Olson et al. 2001). The relative forest cover percentage is based on GlobCover 2009 v2.3 (Bontemps et al. 2011) including the following classes: “mosaic vegetation,” “closed to open broadleaf (B) evergreen (E) forest,” “closed B deciduous (D) forest,” “open BD forest,” “closed needleleaf (N) E forest,” “open NE/ or BE forest,” “closed to open mixed BD/ND forest,” “mosaic forest,” “closed to open B regularly flooded,” “closed B forest permanently flooded.” Spatially explicit soil C and N stocks were derived from the ISRIC-WISE soil map (Batjes 2012)

Full size image

Globally, both boreal and temperate forest soils have been identified as a source of atmospheric N2O and as a net sink for atmospheric CH4. The IPCC 2001 report listed the source strength of temperate forests for atmospheric N2O with 1.0 Tg N2O-N a−1 (0.1–2.0 Tg N2O-N a−1), while for boreal forest soils, an estimate was missing. More recently, Dalal and Allen (2008) estimated that boreal forests are a weak source for N2O (0.33 ± 0.27 Tg N2O-N a−1) and confirmed earlier estimates for temperate forests (1.05 ± 0.37 Tg N2O-N a−1). With regard to atmospheric CH4, Dutaur and Verchot (2007) estimated the sink strength of boreal and temperate forests to be 3.4 ± 5.0 Tg CH4-C a−1 and 2.5 ± 2.6 Tg CH4-C a−1, respectively. However, since wetland forests were excluded from the study, this estimate is likely biased, because wetlands show net CH4 emissions at the annual scale. Dalal and Allen (2008) estimated a smaller, more variable sink strength of boreal forest soils (2.0 ± 4.0 Tg CH4-C a−1), whereas the estimated CH4 sink strength of temperate forests was with 3.7 ± 0.5 Tg CH4-C a−1 higher and highly constrained.

Climate change refers here to the human-induced alteration of weather patterns, such as temperature and rainfall (amount, frequency, seasonal distribution). Climate change affects soil environmental conditions, as well as landscape hydrology, vegetation cover, and substrate supply. Indirect effects of climate change on land use (e.g., it is expected an agricultural expansion further north (Kicklighter et al. 2014)) are not covered here. Different climate models indicate that the temperate and boreal zones will experience warming in the range of 1.4–5.8 °C by 2100 (Hanewinkel et al. 2013), accompanied by an increase in extreme weather events, which will provoke the shrinkage of permafrost, and the reduction of the snow cover period (IPCC 2013). This will result in changing environmental conditions in both forest canopies and soils, along with shifts of vegetation zones, i.e., upward and northward expansion of the temperate and boreal forest biomes. Alterations of forest species composition, forest growth, and vitality of natural and managed forest landscapes will result in modification of the ecosystem services sustained by forests. While synthesis on the contribution of forests to several ecosystem services is already available (e.g., Millennium Ecosystem Assessment 2005), comprehensive studies linking forests and the exchange of non-CO2 greenhouse gases between forests and the atmosphere are much more scarce. This chapter aims at evaluating the role of temperate and boreal forests as providers of climate regulation services. Special emphasis is given to the production and consumption of non-CO2 greenhouse gases by forest soils under changing environmental conditions. Specifically, this chapter assesses how the changes in climate and associated effects may affect temperate and boreal forest soils N2O and CH4 fluxes, thereby summarizing existing knowledge and identifying research gaps.

2 Governing Processes and Mechanisms of Forest Soil-Atmosphere CH4 and N2O Exchange

Nitrous oxide is mainly produced by the microbial processes of nitrification and denitrification, i.e., an oxidative process converting ammonia/ammonium (NH3/NH4 +, classical nitrification) or organic N (heterotrophic nitrification) to nitrate (NO3 −) and a reductive process, which uses NO3 − as an electron acceptor for C oxidation to finally convert it to N2 (denitrification) (Butterbach-Bahl et al. 2013). In these key microbial processes, N2O is either a facultative (nitrification) or obligate (denitrification) intermediate, which can be released to the soil air, consumed in other parts of the soil profile or finally be emitted to the atmosphere. Although denitrification is considered the most important source of N2O in forest soils at the European level, nitrification activity also drives total soil N2O emissions (Ambus et al. 2006). Other microbial processes such as NO3 − ammonification or physicochemical processes, e.g., chemical decomposition of reactive inorganic N species such as hydroxylamine (NH2OH) or nitrite (NO2 −), can lead to N2O formation too (Butterbach-Bahl et al. 2013). The major controls for N2O production in forests soils are substrate availability, i.e., NH3/NH4 + and/ or NO3 − as well as easily degradable C availability (Butterbach-Bahl et al. 2012), temperature (with sensitivity for N2O emission varying widely) (Brumme 1995; Butterbach-Bahl et al. 1997; Díaz-Pinés et al. 2014; Sitaula and Bakken 1993; Zhang et al. 2016), and soil moisture and soil aeration, as both affect the soil redox potential and thus the preference of reductive processes such as denitrification (Butterbach-Bahl et al. 2013). In addition, soil N2O emissions are indirectly controlled by tree and associated plant species, forest stand characteristics, and their effects on the abovementioned parameters, soil C:N ratios (Klemedtsson et al. 2005) and soil microbial community composition (Philippot et al. 2009). Finally, the occurrence of extreme events, such as wildfires and pronounced freeze-thaw and soil drying-rewetting cycles (Borken and Matzner 2009; Butterbach-Bahl et al. 2013), strongly affects microbial activity and availability of substrates for N2O-producing processes.

Methane is predominantly produced in anaerobic, organic-matter-rich microsites of forest soils as a final step of the anaerobic decomposition of organic matter (Conrad 1996). CH4 production has been observed in both the forest floor and the mineral soil (Butterbach-Bahl and Papen 2002). Forest soils can predominantly function as weak sources of CH4 (0–20 kg CH4-C ha−1 a−1) if soils are poorly drained or seasonally flooded due to their topographic position in the landscape, such as many aspen or alder stands (Mander et al. 2015; Matson et al. 2009). In upland soils, CH4 produced at anaerobic microsites or in deeper soil layers is likely to be oxidized while passing through aerobic soil layers. This implies that the observed CH4 flux at the forest soil-atmosphere interface is the net result of simultaneously occurring production and consumption processes (Conrad 1996). Most of the temperate and boreal forest soils are upland soils, which predominantly function at annual scales as weak sinks for atmospheric CH4 (0–5 kg CH4-C ha−1 a−1) (Dutaur and Verchot 2007). However, topographically complex ecosystems may lead to spatial fragmentation at the landscape level, with specific locations being net CH4 sinks while others being strong “hotspot” CH4 emitters (Nykänen et al. 2003). The CH4-oxidizing microbial communities are mostly using O2 – but under certain circumstances also use sulfate or NO3 − – as electron acceptors (Conrad 2009). High-affinity methanotrophic bacteria found in most forest soils are capable to gain energy from soil atmosphere CH4 concentrations lower than 1.7 ppmv. Climate change interacts in several ways with CH4 production and consumption processes in soils. On the one hand, climate change directly affects soil environmental conditions, namely, moisture and temperature, and by this the balance between oxidative and reductive processes, e.g., temperature increases, will – as long as water availability is not limiting – likely result in an increase in aerobic respiration, thus decreasing soil oxygen (O2) availability and the CH4 oxidizing capacity of upland soils. On the other hand, global change and increases in atmospheric CO2 concentration affect plant biomass production and its aboveground-to-belowground ratio, root exudation, and litter quality. All these changes finally modify ecosystem CH4 exchange, with results being different across different ecosystem types and climatic zones. Finally, climate change also affects regional water balances and thus landscape groundwater levels. This will ultimately control the future distribution of wetlands and emission magnitudes of CH4 at the landscape scale (Jungkunst and Fiedler 2007).

3 Forest Composition and N2O and CH4 Fluxes

Forest tree species composition and tree species richness are of high significance with regard to the provision of economic and ecological services by forests (Gamfeldt et al. 2013). While extensive research has been conducted to elucidate the effect of tree species on biomass production (De Vries et al. 2003), biodiversity (Barbier et al. 2008), water regulation (Ewers et al. 2002), or soil C sequestration (Vesterdal et al. 2008; Díaz-Pinés et al. 2014), our knowledge is rather limited with regard to the relationship between forest composition and its importance for the function of forests as climate regulators, specifically in view of the importance of forest soils as sink or sources of non-CO2 greenhouse gases.

Individual trees strongly interact with the surrounding environment by, e.g., reducing the amount of light reaching the soil surface, intercepting water in their canopies, taking up water and nutrients from the soil, and returning organic matter back to the soil. Specific tree species usually behave differently (due to, e.g., different growth rates, water or nutritional requirements, or canopy and root system structure) and therefore create distinct ecological conditions and biogeochemical characteristics in both the canopy (radiation levels, microclimate) and the soil (moisture, pH value, or availability of nutrients). Consequently, microbial processes responsible for production and consumption of N2O and CH4 in both the forest floor and the mineral soil are usually tree-species-dependent (e.g., Borken et al. 2003; Butterbach-Bahl et al. 2002; Díaz-Pinés et al. 2014).

Litter is an inherent part of nutrient and C cycling in forest ecosystems. Aboveground litter regulates microclimatic conditions by forming a protective layer on the soil surface (Sayer 2006). Litter material from conifers contains high amounts of lignin and tannins, which are mainly decomposed by fungi (Dix and Webster 1995), as opposed to litter originated from deciduous trees (e.g., beech). The latter has simpler chemical structures and can be decomposed by broader spectra of soil microorganisms. This usually provokes that coniferous forests develop a thicker forest floor, which both produces and consumes CH4 (Butterbach-Bahl and Papen 2002) and probably limits the transport of atmospheric CH4 into the mineral soil (Borken and Beese 2006; Borken et al. 2003). At the same time, the usually compact and moist litter layer developed under deciduous forest can lead to high N2O production rates (Pilegaard et al. 2006).

Belowground, rhizodeposition and root decay supply soil microorganisms with C to sustain further microbial decomposition (Cheng and Kuzyakov 2005). This, along with root respiration and water and nutrient uptake, significantly alters important biochemical properties (i.e., soil moisture, pH, O2 and CO2 concentrations, and labile C and N concentrations) in the rhizosphere. In a rhizotron experiment, it has been recently found that roots from different tree species affect soil microorganisms and C dynamics in different ways, with Fraxinus excelsior showing a higher CH4 sink and a lower N2O source strength compared with Fagus sylvatica or root-free soil (Fender et al. 2013), underpinning the possible tree-species-dependent root effects on trace gas production in soils.

In addition to the inherent variation of greenhouse gas fluxes along the landscape due to changing environmental conditions, trees can also pose a strong effect on the spatial pattern of N2O and CH4 exchange between the soil and the atmosphere (Butterbach-Bahl et al. 2002). It has been observed that fundamental soil properties (e.g., C and N contents, O2 availability, microbial activity, moisture) strongly vary with distance from the stem (Chang and Matzner 2000) or from the canopy edge (Simón et al. 2013), and this pattern has been found to be tree-species-dependent (Butterbach-Bahl et al. 2002; Van Haren et al. 2010). Further, tree stems can be major conduits for soil-produced CH4 and N2O into the atmosphere. The transport may take place through aerenchymous tissues (extra-large intercellular spaces intended to facilitate aeration in the root system) as has been described for alder trees (Rusch and Rennenberg 1998) but also as dissolved gases in the water stream of the xylem. The contribution of tree trunks and tree leaves to the total ecosystem release of N2O has been estimated to range from 1% to 3% in temperate beech forests (Díaz-Pinés et al. 2015) to 8% in boreal pine forests (Machacova et al. 2016). With regard to CH4, Alnus glutinosa and Betula pubescens trees were found to contribute up to 27% of the ecosystem flux of temperate forested wetlands (Pangala et al. 2015). To our knowledge, information on the contribution of trees to the release of N2O and CH4 in boreal forested peatlands is not available.

Coniferous forests are predominant in temperate and boreal biomes (Douglas et al. 2014). In the frame of a changing climate, tree species better adapted to warmer temperatures and more tolerant to summer drought are supposed to have an adaptive advantage compared with other more water-sensitive tree species. Drought-induced decline of Scots pine (Pinus sylvestris) stands has been observed already in the Alps (Rebetez and Dobbertin 2004) or in the Pyrenees (Galiano et al. 2010), at the extent of Quercus species . In the Rocky Mountains, drought-induced mortality of Abies and Picea species has also been observed (Bigler et al. 2007). In addition to the vegetation succession in view of changing environmental conditions, forest managers have promoted mixed forests or broadleaf species in the last decades, under the belief of having higher stability against disturbances (Jandl et al. 2007). It has been predicted the areal contribution of coniferous forests will shrink at the extent of broadleaf forests (Hanewinkel et al. 2013), even if we lack a clear understanding of how drought- and heat-induced tree mortality will impact the composition of most forests (Anderegg et al. 2013).

Following the change of forest composition, environmental conditions and soil microbial communities are expected to change, along with the organic matter transformation processes, ultimately leading to different end-, co-, and by-products during microbial N turnover (e.g., N2O, NO, N2). There is a substantial number of publications showing higher N2O emissions in soils under deciduous forests than under coniferous ones (Ambus et al. 2006; Brumme et al. 1999; Butterbach-Bahl et al. 2002; Díaz-Pinés et al. 2014). This has been associated to larger NO losses in conifer forests (Butterbach-Bahl et al. 1997), resulting in higher N2O:NO ratios (Papen et al. 2003). Recent results from forest floor incubations in European forests support higher NO emissions from coniferous forests but also a N2O sink potential in the forest floor of deciduous species (Gritsch et al. 2016). Other authors have associated the lower N2O emissions under conifer forests to a decoupling between N2O production and reduction processes, resulting in decreased N2O:N2 ratios (Menyailo and Hungate 2005). With regard to CH4, broadleaf forests usually show higher atmospheric CH4 oxidation rates compared with coniferous ones (Butterbach-Bahl et al. 2002; Maurer et al. 2008), probably due to the distinct CH4 diffusivity of the forest floor developed under each type of forest (Borken and Beese 2006; Borken et al. 2003). However, direct tree species effects on CH4 fluxes can interact with soil moisture effects (Menyailo and Hungate 2005).

Tree species composition may change naturally in the course of ecological succession, but human interventions often also actively modulate stand composition and structures. Thus, forest management plays an active role for determining forest composition, which affects the benefits provided by forests in terms of ecosystem services, including its importance as climate regulators. However, the processes responsible for emitting or taking up N2O or CH4 are highly dynamic, and they are the result of complex biogeochemical processes and feedbacks and usually show a high temporal and spatial variability (e.g., Brumme et al. 1999; Butterbach-Bahl et al. 2002). Further, the forest composition is strongly influenced by the topography and landscape configuration, which in turn impacts the net exchange of CH4 and N2O. Finally, the relevant role of other parameters such as soil texture, precipitation (Borken and Beese 2005), or N limitation (Pilegaard et al. 2006), which may largely overwhelm the direct tree species effects on the net soil-atmosphere N2O and CH4 exchange, appeals for more comprehensive studies including not only different tree species but also soil types, N deposition rates, and climatic regions.

4 Effects of Tree Line and Forest-Tundra Ecotone Shifts on N2O and CH4 Fluxes

Effects of climate change (increasing temperatures, changing precipitation patterns) are particularly intense in temperate and boreal ecosystems compared with global averages, and it has been anticipated that the trend will continue in the coming years (Callaghan et al. 2005). Climatic changes are already affecting the location of the contact line between forest and grasslands or shrublands formation, the so-called tree line ecotone (Wieser 2010). Furthermore, the tree species composition of the forest is changing because of the natural adaptation to new environmental conditions, in addition to the forest management efforts to increase the forest resilience in the frame of global change and the succession after forest disturbances (e.g., wildfires, windthrow). Finally, the spatial configuration and boundaries of temperate and boreal landscape units will probably change (Fig. 2), and the areal extent of the main land cover types (e.g., evergreen and deciduous forests, shrublands, woody wetlands) will vary following changing environmental conditions.

Fig. 2
figure 2

Cross section representing the predominant upland and lowland landscape forms in boreal ecosystems. Size of vertical arrows is proportional to the magnitude of the fluxes (Adapted after Douglas et al. (2014) and Jorgenson et al. (2010)). Vegetation vectors designed by Natasha Sinegina and Bob Comix (creative commons license)

Full size image

Both an upward shift of the tree line in mountainous areas (e.g., Kammer et al. 2009) and a northward migration of the forest-tundra ecotone in boreal latitudes (Serreze et al. 2000) have been detected, indicating an encroachment of forest areas into herbaceous and shrub communities. To our knowledge, there is no information available on the consequences of the movement of the mountainous tree line on the soil N2O and CH4 emissions. With regard to the forest encroachment in boreal latitudes, available studies are limited to individual case studies. When investigating the arctic tree line in Canada, Rouse et al. (2002) found significant releases of CH4 in a fen compared with the negligible CH4 fluxes in the forest (ca. 50 vs 8 kg CH4-C ha−1 a−1). Tupek et al. (2015) found higher CH4 uptake in the upland forest than in the forest-mire transition, whereas no N2O fluxes were observed in any of the ecosystems. In another tundra-forest comparison (Takakai et al. 2008), the forest was a small CH4 sink compared with neutral or CH4 sources from different grasslands. In the same study, the forest was a modest emitter of N2O, whereas the magnitude of N2O emitted by grasslands was highly dependent on water content. On the other hand, others have found no support for the hypothesis that conversion of tundra to forest or vice versa would result in a systematic change of net CH4 fluxes in well-drained soils (Sjögersten and Wookey 2002). As it is usual that both CH4 and N2O emissions from tundra soils largely vary in space and time depending on local hydrological regimes (Zhu et al. 2014), with CH4 annual fluxes from the same site ranging from 10 to 250 kg CH4 ha−1 a−1 (Nykänen et al. 2003), prediction and assessment of CH4 and N2O emissions at the landscape or regional scale are highly challenging.

Due to the lack of a standardized terminology and methodology to locate, characterize, and observe changes, a general shift of the tree-grassland boundary was so far not detected at regional scales for northern latitudes (Callaghan et al. 2005). In addition, while climate change may promote forest encroachment through higher temperatures, human interventions lead to forest degradation and loss of forest-covered area. However, even in the scenario of no alteration of boundaries between ecosystem types, the soil-atmosphere balance of CH4 and N2O in the tree line ecotone is very sensitive to changing environmental conditions. Warming across an extensive northern gradient along the tundra-forest ecotone in Scandinavia consistently enhanced the soil CH4 sink strength (Sjögersten and Wookey 2002), and in northern China, soil warming experiments in tundra resulted in increased uptake rates of both CH4 and N2O (Zhou et al. 2016). On the contrary, Karbin et al. (2015) have recently shown that soil warming does not affect soil CH4 fluxes at the Alpine forest tree line, likely due to the negligible effect of warming on soil water contents.

5 Extreme Events and Forest Soil N2O and CH4 Fluxes

Most climate projections show that temperature change will be highest in northern latitudes, accompanied by significant increases in precipitation (IPCC 2013). As much of the precipitation increase will be in the form of rain, including rain-on-snow, an accelerating rate of snowmelt can be expected (Callaghan et al. 2005), and a higher frequency of freeze-thaw events is forecasted in temperate and boreal regions (Fig. 2). Freeze-thaw events have been shown to result in major pulses of soil N2O emissions, not only from agricultural but specifically from forest soils (De Bruijn et al. 2009). Luo et al. (2012) reported for the Höglwald Forest, a temperate forest in Southern Germany, freeze-thaw events in 5 out of 14 observational years, and that in those years, annual soil N2O emissions were at least a factor of two higher than in other years due to the strong pulse of N2O emissions associated with freeze-thaw events. In contrast, comparable effects on soil CH4 fluxes were not observed.

Permafrost is perennially frozen ground, which underlies 20–25% of the exposed land surface in cold climates, and many boreal forests grow on permafrost (Serreze et al. 2000). Strong arctic warming forecasted in climate scenarios will provoke a strong collapse and shrinkage of permafrost (Callaghan et al. 2005). There is a growing body of research showing that permafrost thawing will yield increased C emissions due to the enhancement of both CO2 and CH4 emissions (e.g., Flessa et al. 2008; Schaefer et al. 2014; Schuur et al. 2015), as well as accelerated N losses following dramatic increases in soil N2O fluxes (Abbott and Jones 2015).

Prolonged drought and following soil-rewetting events are also likely to exert a significant effect on annual soil N2O and CH4 fluxes of forest soils. For example, Borken et al. (2000) showed for a Norway spruce stand in Germany that a simulated drought period of 3–5 months resulted in an increase of the net sink capacity of the soil for atmospheric CH4 by 40–100%. On the other hand, rewetting of soils after drought has been shown to result in pulses of soil N mineralization and soil N2O emissions as a result of the release of accessible substrate from accumulated microbial and plant necromass, the lysis of live microbial cells, and the disruption of previously protected organic matter, allowing for accelerated microbial N turnover processes (Borken and Matzner 2009). However, pulse emissions due to soil rewetting may compensate or not for reductions in soil N2O emissions during the drought period compared with constantly wet scenarios, depending on the frequency of the drying-wetting cycles and the length of the drought periods and the magnitude of the priming effects following the wetting pulse (Borken and Matzner 2009). Moreover, it has been found that under drought conditions, soils might turn from a source to a sink of N2O (Goldberg and Gebauer 2009), so that the expected future increases in drought periods may finally result in a decreasing atmospheric N2O source strength of temperate and boreal forests.

The most important natural disturbances impacting the functioning of temperate and boreal forests are wildfires, windthrows, and insect and disease outbreaks. In a changing climate, the intensity and duration of extreme weather events will increase, with unexpected consequences for disaster, i.e., widespread damage and severe alteration in the normal functioning of ecosystems (IPCC 2012). As a result, ecosystem services provided by forests, including climate regulation, will be also affected. Studies have found a climate signal in the increased wildfire activity in North America (Gillett et al. 2004; Westerling et al. 2006). Enhanced drought in boreal forests is associated with higher frequency of defoliator outbreaks, while late-spring frost has a role in terminating outbreaks (Volney and Fleming 2000). Forest damages due to windthrows have also increased in Europe during the last century (Schelhaas et al. 2003). All these disturbances result in significant reductions or even complete losses of the canopy cover and an episodic, over-proportionated incorporation of plant litter and residues (e.g., ashes) on the forest floor. Reduced water interception and transpiration after forest disturbance has been shown to increase soil moisture levels (Huber et al. 2004), whereas reduced canopy cover leads to enhanced soil temperature amplitudes. Similar to clear-cutting effects, such disturbances usually result in strongly reduced CH4 uptake activities (Strömgren et al. 2016; Wu et al. 2011) and in pulses of N2O emissions during the growing season (e.g., Mcvicar and Kellman 2014; Zerva and Mencuccini 2005) as well as during freeze-thaw periods, due to accelerated organic matter mineralization, which provides substrate for soil microbial processes (Rosenkranz et al. 2010). However, the excess availability of N and the associated pulses seem to be limited to at most a few years, as N losses also increase along hydrological (NO3 − leaching; e.g., Huber et al. 2004, 2010) and gaseous pathways (N2 emissions due to denitrification) and as the regrowing vegetation becomes quickly a strong competitor for the available soil N (Rosenkranz et al. 2010).

6 Synthesis

Temperate and boreal forests play a pivotal role in climate regulation, a critical ecosystem service. Dynamic changes in environmental conditions in the coming decades will likely affect the net forest-atmosphere balance of CH4 and N2O fluxes at different temporal and spatial scales, and these changes will feedback on chemical composition of the atmosphere and, thus, on the global climate.

Increasing temperature and changes in precipitation patterns are specifically expected for northern latitudes, with changes being more pronounced compared with tropical or temperate regions (IPCC 2013). This will affect the production and consumption of N2O and CH4 directly by changing the environmental conditions for microbial processes and indirectly through modifications of the landscape configuration, territory land use, and forest vegetation. A conversion of coniferous to broadleaf and mixed forests in response to climate change could lead to enhanced CH4 uptake rates, whereas the overall effect on N2O emissions is unclear if the role of increasing likeliness of summer droughts is taken into account. There is still insufficient empirical evidence of a direct and consistent link between tree species and the soil-atmosphere CH4 and N2O exchange rates , due to the complex interactions between tree species composition and site/soil characteristics. Soil drainage and more pronounced summer drought will increase the CH4 oxidation strength of upland mineral soils, but this effect could be highly overwhelmed by the anticipated massive collapse of permafrost (Douglas et al. 2014), which is likely to result in the mobilization of currently frozen C and N stocks and years of high CH4 and N2O emissions. Furthermore, the lowering of the water table level will likely enhance soil N2O emissions. Since shorter periods of snow-covered soil are expected, the occurrence and intensity of freeze-thaw events will increase, potentially leading to more frequent pulses of N2O. Finally, acceleration of natural disturbances such as wildfires and pests will decrease the soil’s capacity to uptake atmospheric CH4 and will likely provoke further releases of N2O as ecosystem N stocks are mobilized.

We should mention that there is already a reasonable understanding on how single climate change parameters such as temperature or moisture affect soil environmental conditions, C and N cycling, and associated soil N2O and CH4 emissions. However, we still do not understand how interacting effects of environmental condition changes, i.e., combined changes in tree species composition, atmospheric chemistry (i.e., CO2, ozone, or reactive N concentrations), or enhanced and more severe extreme weather events will jointly affect the soil-atmosphere CH4 and N2O flux. Only a few experiments have tried to tackle this problem comprehensively (e.g., Carter et al. 2012). Even though the fact that gaining this knowledge is fundamental for predicting future feedbacks of soil N2O and CH4 fluxes from temperate and boreal forest soils in response to environmental changes, we are lacking suitable experimental approaches. Moreover, research has been mainly focused on plot to ecosystem scale, without confronting the issue of how landscape fluxes may change due to specific changes in environmental drivers. For example, changes in groundwater table levels, lateral water flows, and flooding events will occur at landscape scales and provoke huge changes in hydrological regimes. Those alterations may exert for many landscapes a more pronounced effect specifically on forest soil N2O fluxes as, e.g., changes in temperature (Butterbach-Bahl and Dannenmann 2011). Here, manipulation experiments of groundwater table levels in combination with other climate change factors and a better fine-scale mapping of the affected territory offer a way forward to get a better insight in forest soil greenhouse gas fluxes in the frame of global change.