Environmental Impacts—Terrestrial Ecosystems

Abstract

The chapter starts with a discussion of general patterns and processes in terrestrial ecosystems, including the impacts of climate change in relation to productivity, phenology, trophic matches and mismatches, range shifts and biodiversity. Climate impacts on specific ecosystem types—forests, grasslands, heathlands, and mires and peatlands—are then discussed in detail. The chapter concludes by discussing links between changes in inland ecosystems and the wider North Sea system. Future climate change is likely to increase net primary productivity in the North Sea region due to warmer conditions and longer growing seasons, at least if summer precipitation does not decrease as strongly as projected in some of the more extreme climate scenarios. The effects of total carbon storage in terrestrial ecosystems are highly uncertain, due to the inherent complexity of the processes involved. For moderate climate change, land use effects are often more important drivers of total ecosystem carbon accumulation than climate change. Across a wide range of organism groups, range expansions to higher latitudes and altitudes and changes in phenology have occurred in response to recent climate change. For the range expansions, some studies suggest substantial differences between organism groups. Habitat specialists with restricted ranges have generally responded very little or even shown range contractions. Many of already threatened species could be particularly vulnerable to climate change. Overall, effects of recent climate change on terrestrial ecosystems within the North Sea region are still limited.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The chapter starts with a discussion of general patterns and processes (Sect. 11.2), such as impacts of climate change on productivity, phenology and biodiversity. Climate impacts on specific ecosystem types, such as forests, grasslands and mires are discussed in more detail in subsequent sections (Sects. 11.3–11.6). The chapter concludes by discussing links between changes in inland ecosystems and the wider North Sea system (Sect. 11.7) and then summarises the main findings of this assessment in the form of a table (Sect. 11.8). The chapter focuses on the direct impacts of climate change; the potential impacts of indirect drivers are beyond the scope of this chapter.

2 General Patterns and Processes

2.1 Vegetation Zone Shifts, Productivity and Carbon Cycling

The terrestrial part of the North Sea region lies mainly in the temperate forest zone, with some boreal elements and treeless tundra at higher altitudes in Scandinavia and Scotland (Fig. 11.1). Below the tree line, significant areas of treeless vegetation would naturally occur only in wetlands (marshes, river floodplains and mires), where soil saturation precludes tree growth.

Fig. 11.1

Potential natural zonal (determined by macro-climate) vegetation types in the North Sea region. Grey areas were not classified (Bohn et al. 2003; simplified by Hickler et al. 2012)

Deforestation and land degradation as a result of grazing and other anthropogenic activities have decreased the natural forest cover over thousands of years (e.g. Simmons 2003; Kaplan et al. 2009; Gaillard et al. 2010). Most of the lowland forests in England, for example, had already been cleared 1000 years ago (Ruddiman 2003). Forest cover over large parts of the UK, the Netherlands, north-western Germany and Denmark is currently less than 15 % (Eurostat 2015; Fig. 11.2).

Fig. 11.2

Current land cover in the North Sea region according to CORINE Land Cover (EUROSTAT 2014)

The current distribution of zonal vegetation types in the North Sea region is influenced climatically mainly by temperature because terrestrial net primary productivity (NPP) is less limited by water supply, which is relatively high during the growing season because this is when most rainfall occurs (see Sect. 1.5). In terms of future changes in climate and weather (see Chap. 5), the warming expected by the end of the century can be expected to lead to a northward shift in zonal vegetation types or up in altitude (Hickler et al. 2012), and an increase in NPP where the warming is not accompanied by substantially drier conditions. Most climate change scenarios project an increase in annual precipitation across the North Sea region by the end of the century, although substantially drier conditions have been projected for summer and in particular for the southern part of the region, where water availability already constrains vegetation productivity (see Chap. 5). Together with the slight projected increase in dry spell length (see Chap. 5 and Jacob et al. 2014), vegetation productivity might, therefore, decrease in the southern North Sea region. However, these projections are based on average results from a number of regional and global climate models (RCMs and GCMs) and because not all models agree in terms of the sign of the change in summer precipitation for different parts of the North Sea region, these projections of future water availability during the main growing season contain uncertainties (see Chap. 5 and Jacob et al. 2014). Furthermore, water availability also controls forest productivity strongly in the south-eastern UK (Broadmeadow et al. 2005), not strictly the southern part of the study region. Here too, increasing drought stress in summer would probably negatively impact NPP. Nevertheless, it should be noted that unchanged precipitation implies less water availability because evapotranspiration will increase with rising temperature. According to the multi-model mean of the CMIP5 models (see Chap. 5), the net outcome of changes in precipitation and evapotranspiration is projected to be an increase in annual run-off in the northern part of the region and a decrease in the south (Collins et al. 2013). These changes in the water balance are particularly important for wetlands (see Sect. 11.6).

The uncertainties in projections of future summer moisture (see Chap. 5) make it difficult to predict the impacts of climate change on terrestrial ecosystems. Morales et al. (2007) simulated the combined effects of climate change and increasing atmospheric carbon dioxide (CO₂) levels on European ecosystems with a dynamic vegetation model, using projections from a variety of combinations of RCMs, bounding GCMs and emission scenarios (Christensen et al. 2007, not accounting for changes in land use). With the exception of north-western France, all simulations indicated increasing NPP in the North Sea region by the end of the century. According to these simulations, the northern part of the study region remains a carbon sink, and the southern part continues to be a small source. However, different climate impact models can yield different results even when driven by the same climate scenario data. Using the SRES high A1Fi scenario (Nakićenović and Swart 2000), a number of dynamic global vegetation models (DGVMs) simulated increasing NPP over most of the North Sea region by the end of the century (Sitch et al. 2008), whereas the Lund-Potsdam-Jena (LPJ) DGVM showed decreased vegetation carbon storage especially in the southern part (Sitch et al. 2008). Most of the models in this study, as well as the model used by Morales et al. (2007), included the potential beneficial plant-physiological effects of increasing atmospheric CO₂ concentrations, but not the constraints on this effect through nutrient limitation.

Increasing levels of atmospheric CO₂ will increase NPP (sometimes referred to as the CO₂ fertilisation effect), and most plants reduce stomatal opening in response to higher CO₂ concentrations (e.g. Ainsworth and Long 2005; Hickler et al. 2015). Reduced stomatal opening leads to lower plant transpiration rates, commonly increasing soil water content and thereby counterbalancing potentially increasing drought stress under climate warming (Arp et al. 1998; Morgan et al. 2004; Körner et al. 2007). Increasing leaf area as a result of higher NPP can counteract this water saving effect (e.g. Gerten et al. 2004), but mostly under conditions of ambient nutrient supply, which enables plants to take advantage of increasing CO₂ and to increase their leaf area (Arp et al. 1998; McCarthy et al. 2006; Norby et al. 2010). According to future simulations with a GCM that includes dynamic vegetation changes, the net outcome of the two effects will be a substantial increase in global run-off (Betts et al. 2007). However, CO₂ enhancement experiments with conifer trees have shown hardly any reduction in stomatal conductance (Körner et al. 2007), implying that the vegetation models probably overestimate the reduction in stomatal conductance and transpiration in conifer forests (Leuzinger and Bader 2012). The magnitude of the CO₂ fertilisation effect on NPP and carbon storage is highly debated (e.g. Körner et al. 2007; Thornton et al. 2007). Although photosynthesis increases under elevated CO₂, this enhancement of carbon assimilation often does not lead to increased biomass as the extra carbon is mainly allocated to below-ground carbon pools with fast turnover (fine roots, root exudates, transfer to mycorrhiza) (Körner et al. 2005; Finzi et al. 2007; Norby et al. 2010; Walker et al. 2014). Nitrogen (N) deposition can also increase NPP, but in the southern North Sea region, N-deposition is already so high that nitrogen is not limiting terrestrial productivity directly (but may decrease productivity through negative side effects such as soil acidification; Bowman et al. 2008; Horswill et al. 2008). N-deposition across the study region is expected to remain at similar levels as today (2014) or to decrease slightly (Tørseth et al. 2012), but N-mineralisation in the soil will probably increase in the northern North Sea region due to warming (Lükewille and Wright 1997; Melillo et al. 2011), which would increase terrestrial productivity particularly in N-limited vegetation on acidic soils (see also Sects. 1.7 and 11.6).

Net primary productivity is an important driver of many ecosystem services, including total carbon storage, but in the North Sea region its dynamics are determined largely by land use, which has not been accounted for in the DGVM study mentioned previously (Sitch et al. 2008). Over most parts of Europe, including the North Sea region, forest carbon stocks, for example, are currently increasing as forests grow older and less timber is harvested than a few decades ago (Janssens et al. 2003; Nabuurs et al. 2003; Ciais et al. 2008).

Total ecosystem carbon storage is further influenced by soil carbon dynamics. Soil respiration, and thereby carbon losses from the soil, is expected to increase under global warming, but the sensitivity of the soil carbon pool remains uncertain (Davidson and Janssens 2006; Luyssaert et al. 2010), and combined effects of potentially increasing NPP (and carbon inputs into the soil) and increasing soil respiration rates (reducing carbon storage) on total ecosystem carbon storage are very difficult to estimate.

2.2 Changes in Phenology

Changes in the phenology of biota currently provide the most sensitive and compelling evidence of climate warming impacts in the North Sea region and elsewhere in the middle and higher latitudes. At the same time these changes are particularly well documented due to a pan-European network of phenological data collections that has been run continuously since the mid-20th century (e.g. Menzel 2000) as well as long-term data from bird-ringing stations (e.g. Sparks et al. 2005) and butterfly monitoring programmes (Roy and Sparks 2000). Phenological changes that can be attributed to climate change include leaf unfolding, flowering and leaf colouring as well as the arrival dates of migrant birds, dates of egg laying of birds or the timing of the first appearance of butterflies (Parmesan and Yohe 2003; Parmesan 2006).

An analysis of observational data from the International Phenological Gardens in Europe for the 1959–1996 period (Menzel and Fabian 1999; Menzel 2000) revealed that spring events such as leaf unfolding have advanced on average by 6.3 days (−0.21 days year⁻¹), whereas autumn events such as leaf colouring have been delayed on average by 4.5 days (+0.15 days year⁻¹). This trend has resulted in an average extension of the annual growing season by 10.8 days since the early 1960s. This trend is of particular significance for regions bordering the North Sea Basin such as Denmark and northern Germany. Similar results were obtained in a more regional study analysing data from the phenological network of the German Weather Service for the period 1951–1996. In this study, Menzel et al. (2001) found the strongest phenological advances in key indicators of earliest and early spring (−0.18 to −0.23 days year⁻¹) whereas changes in autumn were less pronounced (delay of +0.03 to +0.10 days year⁻¹). Overall, the mean growing season for the period 1974–1996 was up to 5 days longer than for the period 1951–1973. Similar findings were made by van Vliet et al. (2014) in the Netherlands. Using data from the Dutch phenological observation network they found that significant changes in life cycle events started only in the early 1990s. In a large-scale meta-analysis using data from 21 European countries for the period 1971–2000 (Fig. 11.3), Menzel et al. (2006) showed that 78 % of all bud break, flowering and fruiting records advanced, and only 3 % were delayed.

Fig. 11.3

Temperature sensitivity and response across the year. a Maximum correlation coefficients for 254 mean national time series of phenophases in nine European countries with mean temperatures of the previous months. b Regression coefficients against mean temperature of the previous month. F flowering; LU leaf unfolding; LC leaf colouring. The overall dependence of temperature sensitivity and response on mean date is high a R ² = 0.59, p < 0.001; b R ² = 0.47, p < 0.001) (Menzel et al. 2006)

This study clearly demonstrated that phenology is directly linked to the temperature of preceding months with a mean advance of spring/summer by 2.5 days per °C and a mean delay of leaf colouring and leaf fall by 1.0 days per °C. So far, phenological changes of this type are reversible and depend on weather conditions in the year of observation.

In the UK, mean laying dates for the first clutches of 20 bird species advanced on average by 8.8 days between 1971 and 1992 (Crick et al. 1997). Similarly, spawning of two amphibian species (toads) in England advanced by two to three weeks between 1978 and 1994, and the arrival of three newt species in breeding pools advanced by as much as five to seven weeks (Beebee 1995). Based on a composite map of 70,000 records for 1998–2007 for the common frog Rana temporaria, Carroll et al. (2009) found an average advance of first spawning of about 10 days in the UK compared to map-based data 60 years before.

On the island of Heligoland in the south-eastern corner of the North Sea, mean spring passage times for 24 species of migratory birds advanced by 0.05–0.28 days year⁻¹, which in most species correlated strongly with warmer local temperature during the migration period as well as with the strength of the North Atlantic Oscillation (NAO; Hüppop and Hüppop 2003). Almost identical findings were made at a larger spatial scale from several ringing stations by Sparks et al. (2005). At the continental scale, Both et al. (2004) analysed 23 European populations of pied flycatcher Ficedula hypoleuca and found that nine showed an advanced laying date, which were all from those areas with the strongest warming trend and mostly situated at the southern fringe of the North Sea basin. In an area of southern England (Oxfordshire), Cotton (2003) demonstrated that earlier arrival of 20 species of long-distance migratory birds was positively correlated with enhanced air temperatures at wintering grounds in Sub-Saharan Africa.

Climate change also has significant impacts on the winter distribution of migratory birds that fly south to avoid the northern winter. Based on ringing data from the Netherlands, Visser et al. (2009) found that 12 of 24 species studied showed a significant reduction in their migration distance to the south, and that this was strongly correlated with the Dutch winter temperature in the year of recovery. For three common waterfowl species, Lehikoinen et al. (2013) demonstrated that shifts in wintering areas to the northeast correlated with an increase of 3.8 °C in early winter temperature in the north-eastern part of the wintering areas, where bird abundance increased exponentially, corresponding with decreases in abundance at the south-western margin of the wintering ranges. In line with these findings, Maclean et al. (2008) showed that the centres of wintering distribution for five species of wading birds along the north-western European coast flyway shifted 95 km north-eastwards within the period 1981–2000.

For the UK, Roy and Sparks (2000) showed that 26 of 35 species of butterfly exhibited an earlier appearance over the relatively short period 1976–1998 (statistically significant for 13 species). The authors estimated that a warming of 1 °C might advance first and peak appearances of most butterfly species by 2–10 days.

2.3 Matches and Mismatches Across Trophic Levels

Shifts in phenology as a response to climate change differ among species and populations. This has been shown for a range of taxonomic groups (Parmesan and Yohe 2003; Parmesan 2006). If climate responses differ between strongly-interacting species, such differences can have immediate impact on key ecological interactions, such as plant–pollinator, herbivore–plant, host–parasite/parasitoid and predator–prey (Visser and Both 2005; Thackeray et al. 2010). This may lead to a phenological mismatch of evolutionary-synchronised species but also to a phenological match of formerly asynchronised species resulting in so far avoided competition, parasitism or predation (Parmesan 2006).

For the Netherlands (Fig. 11.4), it was demonstrated that earlier bud break in sessile oak Quercus petraea due to climate warming leads to an earlier appearance of caterpillars, thus disrupting food supply during the main hatching period of the migratory pied flycatcher, which did not keep pace in its arrival at breeding grounds with the advance in peak food supply (Both and Visser 2001; Both et al. 2006). As a consequence of this mismatch in timing, pied flycatcher populations breeding in oak forest declined by 90 % between 1987 and 2003 (Both et al. 2006).

Fig. 11.4

Trends in pied flycatcher populations in response to the local date of peak caterpillar abundance (Spearman rank correlation: r _s = 0.80, n = 9, p = 0.013) (a), and the slope of annual median egg laying date on spring (16 April–15 May) temperature (r _s = −0.86, n = 7, p = 0.03) (b). Populations of pied flycatchers with an early food peak and a weak response declined most strongly. Population trend is the slope of the regression of the log number of breeding pairs against year. In ‘b’, the x axis shows the slope of a linear regression of median laying date against mean temperature from 16 April to 15 May. Error bars represent the standard errors of the slopes of the regression lines. All points in ‘b’ are also in ‘a’, except for one point, for which no data regarding the caterpillar peak were available (Both et al. 2006)

Such phenological mismatches may show effects across four trophic levels leading to deterioration in the timing of food demand and availability for passerines and their avian predators (Both et al. 2009). Biotic mismatches are also considered a major cause of the disproportionate decline in long-distance migratory bird species compared to short-distance migrants that are able to react more flexibly to phenological changes in their breeding areas (Møller et al. 2008; Both et al. 2010; Saino et al. 2011). Evidence for this phenomenon is, however, so far mostly correlative. The genetic basis of mechanisms of adaptation to phenological change is still poorly understood. Although the actual consequences of phenological changes on ecosystem functioning are not clear, several studies highlight the potential risk of desynchronising trophic linkages between primary and secondary consumers (Thackeray et al. 2010). This includes, for example, the distortion of entire food webs, in which top predators moving to cooler regions may trigger trophic cascades that lead to local extinctions and altered ecosystem processes (Montoya and Raffaelli 2010). To date, most of these assumptions are theoretical and more or less unsupported by experiments or empirical data.

Differences in response to climate change among species may also lead to a matching of originally asynchronous species, also with considerable ecological implications (Visser and Both 2005; Parmesan 2006). Case studies documenting this process are rare. Van Nouhuys and Lei (2004) showed that warmer, early spring-temperatures favoured the parasitoid wasp Cotesia melitaearum disproportionally, bringing it into closer synchrony with its host the butterfly Melitaea cinxia. Although the authors found no direct effect of the phenological matching on local host population size, the synchrony is likely to be important for overall host meta-population dynamics via variation in the rate of colonisation by the parasitoid.

In addition to phenological mismatches, trophic interactions can also become disrupted if host plants and species feeding on these host plants shift their ranges asynchronously. For the monophagous butterfly Boloria titania and its larval host plant Polygonum bistorta, Schweiger et al. (2008) showed that climate change may lead to a spatial mismatch of trophically interacting species due to asynchronous range shifts. Schweiger et al. (2012) analysed the potential for such mismatches in the future for 36 European butterfly species by simulating the potential range shifts for butterflies and host plants separately with bioclimatic envelope models, also taking into account land use. They found that those butterflies that are already limited in their distribution by their host plants could suffer most from global climate change, particularly if the host plants have restricted ranges.

2.4 Range Shifts and Biodiversity

Recent climate change has already influenced the distribution of species and their abundance (Walther et al. 2002; Parmesan and Yohe 2003; Parmesan 2006; Lenoir et al. 2008; Chen et al. 2011). According to a recent meta-analysis covering a range of taxonomic groups across the globe, species have on average shifted their ranges 16.9 km to higher latitudes and 11 m up in altitude per decade, more than estimates from earlier studies. The average shifts have been larger in those areas that have experienced the strongest warming and have, on average, been sufficient to track temperature changes, but with large variation between species, and most observations are from the temperate zone and tropical mountains (Chen et al. 2011). More than 20 % of species actually shifted in the opposite direction (in latitude and altitude) to the one expected based on temperature changes over the last few decades. Such changes can be explained by drivers other than temperature, such as habitat destruction and water availability, together with biotic interactions such as dependency on certain host plants and special physiological constraints (e.g. minimum or maximum temperatures during crucial phases of the life cycle) (Chen et al. 2011; Crimmins et al. 2011; Tingley et al. 2012). The differences between species reported here were only poorly related to broad taxonomic groups, such as birds and butterflies; rather the differences within such groups were larger than the differences between groups. In all taxonomic groups, habitat specialists and those with a low dispersal and colonisation capability show the lowest or even negative range shifts towards higher latitudes or altitudes (Warren et al. 2001; Chen et al. 2011). However, a recent global analysis of projected rates of temperature shift across landscapes compared to maximum projected speeds at which species can move across landscapes (from observations and modelling studies) showed that many species will probably be unable to track climate change, particularly for the warmer scenarios, which imply faster warming than in the recent past (Settele et al. 2014). Furthermore, this analysis also suggested large differences between organism groups in terms of their dispersal capacity. Herbaceous plants and trees seem to have particularly low dispersal capacity (Settele et al. 2014). An analysis for the British Isles (Fig. 11.5) also showed substantial differences in range expansion for different taxonomic groups (Hickling et al. 2006).

Fig. 11.5

Latitudinal shifts in northern range margins for 16 taxonomic groups in the British Isles during recent climate warming. Results are given for three levels of data subsampling (grey recorded; white well-recorded; black heavily recorded). Only species occupying more than twenty 10 km grid squares across two time periods (between 1960 and 2000, depending on organism group) are included in the analyses; for several of the species-poor groups, these criteria excluded all species from the analysis of ‘heavily recorded’ squares (Hickling et al. 2006)

In the North Sea region, substantial average northward shifts have been well-documented for birds, butterflies, moths, dragonflies and damselflies, but mostly with large numbers of species also showing no shift or even retreating northern range boundaries (Parmesan 2006). Among the well-studied groups, plant ranges show the smallest responses to recent climate change, at least in lowland areas, probably because of their limited capacity to disperse and colonise new habitats in highly-fragmented landscapes (Honnay et al. 2002; Bertrand et al. 2011; Doxford and Freckleton 2012). Analyses of community composition, however, show substantial increases in warm-adapted vascular plants and epiphytic lichens across the Netherlands, which have probably been partly driven by climate change (van Herk et al. 2002; Tamis et al. 2005). Also, as these changes have clearly been driven by other factors (such as changes in land use, eutrophication and, in the case of lichens, decreasing sulphur emissions) attributing them to climate change is challenging. Seventy-seven new epiphytic lichen species colonised the area between 1979 and 2001, nearly doubling the total number of species (van Herk et al. 2002) and overall vascular plant richness also increased (Tamis et al. 2005).

Average model projections for the migration rates that would be necessary to track climate change in Europe are substantially larger than those historically observed, but the magnitude of the mismatch depends heavily on the climate change scenario (Skov and Svenning 2004; Huntley et al. 2008; Doswald et al. 2009). Simulations with bioclimate envelope models suggest large local (per grid cell) species losses and turnover rates, assuming that species fully track climate change by migration (Thuiller et al. 2005; Pompe et al. 2008). For the SRES A1 scenario (Nakićenović and Swart 2000) and the HadCM3 climate model, for example, Thuiller et al. (2005) estimated an average turnover of 48 % per grid cell for the European plants considered (1350 for all of Europe) in the European Atlantic region by 2080. However, results from bioclimate envelope models should be interpreted more as potential shifts in the climatic window in which species can thrive rather than projections in range shifts. Furthermore, such models may overestimate change because they are developed based on correlations between species ranges and environmental factors. They do not capture the fundamental niche of species and so underestimate the climatic niche when species have not yet reached their distribution in equilibrium with the climate, which appears to be common, at least for trees (Svenning and Skov 2004; Normand et al. 2011). Furthermore, dispersal is rarely simulated explicitly, and dispersal projections are uncertain, for example, because of large uncertainties in projected wind speeds (Bullock et al. 2012). Nevertheless, it could be expected that many mobile, generalist species will continue to shift their distributions northward and up in altitude in response to climate change, although many habitat specialists (often those that are rare and already endangered) will not, and that many cold-adapted species will probably experience range losses at their southern distribution limit or at lower elevations (Hill et al. 2002; Chen et al. 2011; Sandel et al. 2011; Schweiger et al. 2012). As the area south of the North Sea region is generally more species-rich (e.g. Thuiller et al. 2005), biodiversity in the North Sea region could even increase. Negative impacts on cold-adapted species are expected to be most severe in mountain regions, where species have limited possibilities to migrate upwards or northwards, such as on mountains in the British Isles (Berry et al. 2002; Hill et al. 2002). Recent climate change has also affected the community compositions of birds and butterflies in Europe. Analyses of 9490 bird and 2130 butterfly communities in Europe show large changes, equivalent to a 37 and 114 km northward shift in bird and butterfly communities, respectively. However, these analyses suggest an even larger ‘climatic debt’, corresponding with a migration lag of 212 and 135 km for birds and butterflies (Devictor et al. 2012).

Intensification of agricultural activities and increasing anthropogenic nitrogen inputs since the 1950s and 1960s have been major drivers of biodiversity changes in the North Sea region (e.g. Ellenberg and Leuschner 2010). Wesche et al. (2012) found large changes in grassland community composition in five floodplain regions in northern Germany between the 1950s and 2008 and a decline in species richness at the plot level of 30–50 %. The decline was particularly strong among nectar-producing herbs, which is likely to have had negative effects on pollinators (Wesche et al. 2012). An analysis of Ellenberg indicator values for nutrient availability and a qualitative comparison with a protected area in the same region suggests that these changes were largely driven by increased nutrient inputs. Also, for a number of insect groups, a decline in species preferring low-productivity habitats and dry grassland specialists has been recorded in northern Germany (Schuch et al. 2012a, b). Pollinators are generally declining in Europe, and this has been particularly well documented for the Netherlands and the UK (Biesmeijer et al. 2006; Potts et al. 2010). However, the reasons for the decline are unclear. Potential drivers include habitat loss and fragmentation, agrochemicals, pathogens, invasion of non-native species, climate change and the interactions between them (Potts et al. 2010). These changes show the significant role of land use practice for biodiversity in north-western Europe. Further intensification of agricultural practices, possibly driven by an increasing demand for biofuels, is likely to have negative effects on biodiversity even if atmospheric N-deposition does not increase.

3 Forests

Forests are currently considered the most important carbon sink in Europe (Janssens et al. 2003). Due to the relatively low proportion of forests in the present land cover for countries bordering the North Sea—except Norway—the regional significance of this area as a carbon sink is relatively small or even negative compared to other European regions with higher forest cover (Janssens et al. 2005).

3.1 Climate Impacts on Productivity and Carbon Stocks

Although estimates of the mean long-term carbon forest sink (net biome production, NBP) are more reliable than those from grasslands (Janssens et al. 2003), the role of wood harvests, forest fires, losses to lakes and rivers and heterotrophic respiration remains uncertain and difficult to predict. Almost one third of the NBP is sequestered in the forest soil, but large uncertainty remains concerning the drivers and future of the soil organic carbon pool under climate change (Luyssaert et al. 2010). Nevertheless, increasing temperatures, longer growing seasons, higher atmospheric CO₂ concentrations, and in the north, increasing N-mineralisation, are likely to increase the potential forest productivity where summer precipitation does not decline (Lindner et al. 2010). Moreover, it is uncertain to what extent this potential can be realised as forests will increasingly face a climate to which the planted species or provenances are not adapted, which might increase their susceptibility to pests and pathogens, such as bark beetle (Scolytinae) outbreaks, which can lead to major forest die-back events particularly in Norway spruce Picea abies stands (Schlyter et al. 2006; Bolte et al. 2010). Furthermore, warmer and longer vegetation periods will accelerate the development of bark beetles, in some regions allowing for additional generations within a growing season (Jönsson et al. 2009). Other insect herbivores will also benefit from warmer conditions (Lindner et al. 2010). In a climate manipulation experiment in a Norwegian boreal forest, raised temperature and CO₂-level stimulated the outbreak of heather beetle Lochmaea suturalis and led to a shift in the ground vegetation from common heather Calluna vulgaris to blueberry Vaccinium myrtillus and cowberry Vaccinium vitis-idaea (van Breemen et al. 1998).

The complex interplay between climatic stress, pests and pathogens, and further disturbance such as windfall is hardly captured in the forest models used to project potential future impacts of climate change (e.g. Kirilenko and Sedjo 2007). As a result, it is highly uncertain whether climate change will lead to higher standing biomass in forests.

3.2 Shifts in Communities and Species Distribution

Projections of potential climate-driven transient shifts in broadly-defined forest types suggest only moderate changes in the North Sea region by 2100 (Hickler et al. 2012). The most significant changes projected are the spread of broadleaved and hemi-boreal mixed forests northward in southern Sweden and Norway as well as an upwards shift of the tree-line in the southern Scandes, which is already taking place (Kullman 2002).

Long-term equilibrium of shifts in forest type could be much more substantial, with thermophilous forests dominating in the south-western UK and temperate broadleaved forest along most of the Norwegian coast (Hickler et al. 2012). Recent range shifts northward (Fig. 11.6) have already been observed for cold-hardy, broadleaved, evergreen species such as holly Ilex aquifolium at their northern distributional limit in Europe (Walther et al. 2005; Berger et al. 2007).

Fig. 11.6

Distribution of holly Ilex aquifolium and the 0  °C-January isoline at different times. a Former range of I. aquifolium based on Enquist (1924) and Meusel et al. (1965), isoline based on Walter and Straka (1970), symbols based on Iversen (1944); circles: I. aquifolium within or at the border of the station area; circles: with cross I. aquifolium strayed into woods from gardens; stars: Ilex area lies immediately outside the station area; crosses: I. aquifolium missing in the station area. b Modelled range of I. aquifolium in the recent past (1931–1960), isoline as in ‘a’. c Former range of I. aquifolium as in ‘a’; isoline updated for 1981–2000 based on Mitchell et al. (2004), symbols as for ‘a’. d Former range of I. aquifolium complemented by the simulated species distribution under a moderate climate change based on 1981–2000 climate data, isoline as in ‘c’; triangles represent locations with new observations of I. aquifolium (Walther et al. 2005)

Many European tree species have not yet filled their potential climatic niche in Europe because of dispersal limitations (Svenning and Skov 2004; Normand et al. 2011). Thus dispersal-limited species may be unable to track future climate change, unless foresters assist migration.

In contrast, there is almost no evidence of range shifts in herbaceous forest plants. Unlike mountain forest with short migration distances, there is some evidence that in lowland forests plant distribution changes will lag behind climate warming (Bertrand et al. 2011). Observational (Honnay et al. 2002) and modelling studies (Skov and Svenning 2004) suggest that this is probably due to dispersal limitation resulting from forest habitat fragmentation in lowlands. Where significant range shifts of forest herbs northward and eastward have been documented, such as for the oceanic annual woodland herb climbing corydalis Ceratocapnos claviculata, it is questionable whether this is due to climate change or to other drivers such as eutrophication or assisted migration through the international timber trade (Voss et al. 2012).

Among the major forest tree species, beech Fagus sylvatica is expected to extend its range northward in Britain and southern Scandinavia (Kramer et al. 2010; Hickler et al. 2012), whereas environmental conditions for the commercially-important Norway spruce will become less favourable (Pretzsch and Dursky 2002; Schlyter et al. 2006; Hanewinkel et al. 2013). Although beech is often considered to be very sensitive to drought, several studies (Lebourgeois et al. 2005; Meier and Leuschner 2008; Mölder et al. 2011) showed considerable phenotypic plasticity in response to drought stress (e.g. Bolte et al. 2007). The same is true for sessile oak, which proved to be highly resilient even to extreme drought (Leuschner et al. 2001; Lebourgeois et al. 2004; Friedrichs et al. 2009; Merian et al. 2011; Härdtle et al. 2013). Given the generally damp climatic conditions of north-western Europe, major broadleaved forest trees such as beech and sessile oak are probably not constrained by the projected climate change, which is in line with predictions of vegetation models (Kramer et al. 2010; Hickler et al. 2012). However, using older climate projections with lower projected rainfall than the latest average projections (see Chap. 5), simulations with a forest tree suitability model based on climatic and edaphic factors suggested that the majority of native broadleaved species would become unsuitable for commercial timber production in southern England due to increasing drought severity (Broadmeadow et al. 2005).

Forest management includes a wide range of measures to mitigate climate change effects, such as the selection and planting of species and provenances adapted to future climate (Isaac-Renton et al. 2014); a reduction in rotation cycles to accelerate the evolution and establishment of better adapted genotypes (Alberto et al. 2013); and the use of mixtures of high genetic variation across an array of environmental conditions (Hemery 2008; Köhl et al. 2010). Scientifically-sound implementation of such adaptation measures requires a wide range of research and monitoring activities such as testing of the suitability of new tree species and provenances, a regional risk analysis based on retrospective performance as well as the analysis of climate envelope and climate matching under potential future climates (Hulme 2005; Bolte et al. 2009; Hemery et al. 2010).

4 Grasslands

After cropland, grasslands are the dominant land use type in the North Sea catchment area. In the UK, grasslands comprise more than 40 % of land cover (EUROSTAT 2015). Due to conversion into cropland, and the cessation and intensification of agricultural practices, grasslands underwent fundamental change during the 20th century (Bullock et al. 2011). Changes in management practice and eutrophication are currently the major drivers of ecological change in grasslands. At the same time, grasslands are of major significance for biodiversity and nature conservation in north-western Europe.

4.1 Climate Impacts on Carbon Stocks and Cycling

Unlike forests, carbon accumulation in grassland ecosystems occurs mostly below ground. As fluxes of greenhouse gases in grasslands are intimately linked to management and site conditions, grasslands can be either a sink or a source of greenhouse gases. Although many studies consider temperate grassland to be a carbon sink (Soussana et al. 2004), there is still high uncertainty about their current and future net global warming potential (in terms of CO₂ equivalents) at both a regional and continental scale (Janssens et al. 2003, 2005; Smith et al. 2005).

For Britain and Wales, both dominated by grasslands, Bellamy et al. (2005) suggested that significant losses of soil organic carbon (SOC) between 1978 and 2003 must be attributed to climate change because they occurred across all types of land use. However, as shown by Smith et al. (2007), this assumption was precarious and lacked clear empirical evidence. At a global scale (Guo and Giffort 2002), current SOC losses and gains in grasslands can be predominantly attributed to changes in land cover and management, whereas the role of climate change remains uncertain but is predicted to increase over the coming century (Smith et al. 2005). These considerable uncertainties are because grassland ecosystems are particularly complex and difficult to study owing to the wide range in management and environmental conditions to which they are exposed. As a result, studies on the effects of climate change on grasslands are often affected by this variability as well as by other confounding effects such as eutrophication and changes in management practice, which cause difficulties for observational studies and modelling (Soussana et al. 2004).

4.2 Climate Impacts on Plant Communities

Few experimental studies attempt to isolate the effects of climate change from other confounding effects. A study simulating warming and extended summer drought in calcareous grasslands at Buxton and Wytham in northern England (Grime et al. 2000) covered two different types of grassland with contrasting effects: after five years of climate manipulation, the more fertile, early successional grassland at Wytham showed significant changes in species composition and aboveground productivity especially for the combination of winter warming and summer drought. In contrast, an oligotrophic and more traditional calcareous grassland at Buxton exhibited almost no response to warming and drought treatments. This was still true even after 13 years of climate modification (Fig. 11.7; Grime et al. 2008). One reason for the high resistance of this infertile grassland may be the small-scale spatial heterogeneity in soil depth allowing the coexistence of drought-tolerant and more mesic species at small spatial scales (Fridley et al. 2011). As indicated by seed addition experiments, the minor changes in species composition, even after long-term climate treatments, are significantly affected by dispersal limitation rather than just biotic resistance (Moser et al. 2011). The prominent role of dispersal limitation as a cause of delayed response to climate effects has also been highlighted in several other studies (Buckland et al. 2001; Stampfli and Zeiter 2004; Zeiter et al. 2006).

Fig. 11.7

Mean dissimilarity of treatment and control species composition for each year of the Buxton climate change experiment conducted in a nutrient-poor calcareous grassland. Dissimilarity was measured by Sørensen distance estimated separately within each replicate (n = 5 per treatment year). Dashed lines indicate mean dissimilarity in year 1 of the experiment. Error bars indicate one standard error of the mean. Statistics indicate whether treatment dissimilarity progressively increased over time based on linear autoregressive models (Grime et al. 2008)

Overall, the results of the Wytham and Buxton experiments suggest that more productive grasslands, strongly altered by human activities, might respond more to effects of climate change than infertile and more traditionally managed grasslands with rich species pools that can buffer climate effects (Grime et al. 2000, 2008). However, infertile traditional grasslands show low resilience towards eutrophication and changes in land management, which are currently more important drivers of ecological change in grasslands than climate change. Conversely, future warming potentially in association with increased drought risk could supersede eutrophication as the main driver of change, and in so doing potentially favour the persistence or even spread of dry and infertile grassland types (Buckland et al. 1997).

Observational studies also suggest that changes in the grasslands of north-western Europe can be attributed to recent regional climate change (e.g. Gaudnik et al. 2011), but these are mostly of high uncertainty due to strong confounding effects (McGovern et al. 2011). However, flowering phenology of many typical grassland plants in the UK does reveal significant effects of climate warming. Of 385 plant species, 16 % flowered significantly earlier in the early 1990s compared to previous decades, and earlier onset of flowering was most significant in annual species (Fitter and Fitter 2002). Williams and Abberton (2004) confirmed a significant trend of earlier flowering within different agricultural varieties of the common grassland legume white clover Trifolium repens.

4.3 Climate Impacts on Animal Communities

More convincing evidence of climate change effects in grasslands comes from animal groups typical of grassland habitats such as butterflies and grasshoppers; several have extended their range northwards significantly over past decades (Parmesan et al. 1999; Hill et al. 2002; Hickling et al. 2006). Unlike most vascular plants, which are often chronically persistent and immobile, highly-mobile animal species can often quickly respond to changing climate by significant range extensions. In north-western Germany, for example, the Roesel’s busch-cricket Metrioptera roeselii, a typical grassland species, has been rapidly extending its range northward since the early 1990s, which was probably helped by increased rates of macroptery in this normally short-winged species, as a sign of density stress at the range margin (Hochkirch and Damerau 2009; Poniatowski and Fartmann 2011; Poniatowski et al. 2012).

Significant northward range expansions have also been documented for many typical grassland butterflies in the UK (Hill et al. 1999, 2002). However, only particularly mobile habitat generalists can fully exploit the emerging potential offered by climate warming (Fig. 11.8), whereas less mobile species and habitat specialists still suffer from habitat fragmentation and deterioration of habitat quality (Hill et al. 1999; Warren et al. 2001). Thus, range expansion is in many cases significantly lagging behind the current climate and can be reduced or even reversed by non-climatic drivers (Hulme 2005; Oliver et al. 2012).

Fig. 11.8

The degree to which three butterfly species have changed their ranges (a–c, without subsampling) and are lagging behind current climate in Britain (d–f; 10-km grid resolution). a+d, silver-studded blue Plebejus argus; b+e, speckled wood Pararge aegeria; c+f, comma Polygonia c-album. For maps a–c, black circles show butterfly records for both 1970–1982 and 1995–1999; green circles show apparent extinction (recorded 1970–1982; not 1995–1999); pink circles show apparent colonisation (no record 1970–1982; record 1995–1999). For maps d–f, black circles (climate suitable, butterfly recorded) and grey circles (climate unsuitable, butterfly not recorded) show where observed 1995–1999 and simulated distributions agree; red circles (climate predicted suitable, butterfly not recorded) and blue circles (climate deemed unsuitable, butterfly recorded) show mismatches (Warren et al. 2001)

5 Heathlands

In countries bordering the North Sea basin, heathlands dominated by shrubs of the ericaceous family still cover extensive areas especially in highlands of the UK and the southern Scandes (Webb 1998; van der Wal et al. 2011). Due to conversion into cropland and afforestation, heathlands have declined dramatically in the UK-lowlands and in the southern part of the North Sea region (Denmark, Germany, Netherlands). In these regions, they are at the edge of extinction in many sites and have become a major object of biodiversity and nature conservation efforts. Eutrophication and acidification through atmospheric inputs and changes in land management are currently the major drivers of change in these ecosystems (Härdtle et al. 2006), which makes the identification of climate change impacts difficult.

5.1 Climate Impacts on Ecosystem Processes

Effects of climate change on ecosystem processes in European heathlands were specifically addressed within the framework of two EU-projects simulating raised temperatures and drought (Wessel et al. 2004). These studies included sites in the UK, Denmark and the Netherlands. Experimental warming of 1 °C induced a significant increase in total above-ground plant biomass growth of 15 % in the most temperature-limited site in the UK, whereas drought treatments led only to a slight decline (Peñuelas et al. 2004). Drought decreased flowering (by up to 24 % in the UK). Warming and drought decreased litterfall in the Netherlands (by 33 and 37 %, respectively). Tissue concentrations of phosphorus (P) generally decreased and the N:P ratio increased with warming and drought except at the UK site, indicating the progressive importance of P-limitation as a consequence of warming and drought.

Owing to their richness in soil organic matter, mature heathlands may become important sources of C and N-release triggered by increasing temperatures and more frequent periods of drought. For the same experiments as above, Schmidt et al. (2004) found mostly weak and insignificant effects of warming and drought treatments on nitrogen and carbon budgets in the soil solution. Only at a strongly N-saturated site with high atmospheric N-deposition in the Netherlands, did warming trigger a significant increase in N-leaching. Similarly, in the same warming and drought experiments, Jensen et al. (2003) and Emmett et al. (2004) found largely weak or inconsistent responses in major soil processes such as decomposition, respiration and N-mineralisation. The latter turned out to be predominantly controlled by soil moisture. The response of soil-related processes to warming and drought treatments was generally found to be strongly dependent on local site conditions (Emmett et al. 2004). At mesic sites in the Netherlands and Denmark, soil respiration decreased in response to drought but recovered quickly to pre-drought levels after re-wetting in the following winter. In contrast, repeated drought treatments at a particularly damp site in the UK, which was particularly rich in organic matter, caused a disturbance in soil structure and a persistent reduction in soil moisture, which induced increased and continuing carbon losses through soil respiration (Sowerby et al. 2008).

The heathland studies conducted within the framework of the CLIMOOR and VULCAN projects (Peñuelas et al. 2007) show that the magnitude of the response to warming and drought was dependent on differences between sites, years, and plant species. Thus there are complex interactions between other environmental factors that condition plant and ecosystem performance, which makes it extremely difficult to predict net responses.

5.2 Climate Impacts on Plant Communities

Observational and experimental evidence for floristic changes in heathlands that can be clearly attributed to climate change is weak; for example, Werkman et al. (1996) found some indications that climate warming in combination with N-deposition might enhance the spread of the noxious weed bracken Pteridium aquilinum into heathlands in the UK. However, declines in arctic-alpine and boreal-montane lichen species in the heathlands of north-western Europe can be attributed to changes in traditional management practices and acidification, and are probably not directly connected to climate change (Hauck 2009).

5.3 Climate Impacts on Animal Communities

A significant decline has been observed over recent decades in artic-alpine bird species inhabiting mountain heathlands in the north of the UK such as ptarmigan Lagopus mutus, dotterel Charadrius morinellus and snow bunting Plectrophenax nivalis. In contrast the thermophilous, submediterranean Dartford warbler Sylvia undata has increased its population and spread into southern England, probably due to warmer winters (van der Wal et al. 2011). There is currently almost no empirical evidence of climate change impacts in other heathland-specific animal groups. However, modelling approaches suggest (Thomas et al. 1999; Berry et al. 2002) that eco-thermic animal species in heathlands may benefit from climate warming at their northern range margin.

6 Mires and Peatlands

Peatlands store significant quantities of carbon, nitrogen and other elements in their soils (e.g. Limpens et al. 2008; Yu et al. 2010) and are widespread in the North Sea region (Montanarella et al. 2006, Fig. 11.9). According to Joosten and Clarke (2002), peatlands are areas with a naturally accumulated peat layer at the surface, whereas mires are peatlands where peat is currently being formed. While about 80 % of the peatlands in Sweden and Norway are still mires, the contribution of mires to the total peatland area in the other, more southern North Sea-bordering states is only 1–7 % due to widespread drainage and intensive land use (Joosten and Clarke 2002).

Fig. 11.9

Relative cover of peatlands and peat-topped soils in the North Sea catchment area. Maps based on a the soil mapping units of the European Soil Database (King et al. 1994, 1995) and b the NUTS (Nomenclature of Territorial Units for Statistics) level 2 administrative regions (R. Hiederer non-published data)

Peatlands exchange C- and N-containing gases with the atmosphere, particularly the greenhouse gases CO₂, methane (CH₄) and nitrous oxide (N₂O) and thus influence climate (Blodau 2002; Frolking and Roulet 2007; Limpens et al. 2008; Finkelstein and Cowling 2011). Peatlands also exert strong influences on aquatic ecosystems by lateral waterborne export fluxes of elements, especially as particulate and dissolved organic matter (Urban et al. 1989; Freeman et al. 2001; Worrall et al. 2002; Billett et al. 2004; Dinsmore et al. 2010). Importantly, carbon accumulation, and vertical land-atmosphere and lateral waterborne bio-geochemical fluxes of peatlands are affected by climate change, and at the same time by changes in atmospheric chemistry and land use (e.g. Bragg 2002; Belyea and Malmer 2004; Dise 2009; Billett et al. 2010; Charman et al. 2013).

6.1 Climatic Impacts on Abiotic Conditions

Climatic change will have direct effects on the energy and water budgets of peatlands. Changing quantities and temporal patterns of precipitation will affect the water table in peatlands; with drought lowering and increased precipitation raising peatland water levels (e.g. Sottocornola and Kiely 2010). Higher temperatures, which are projected for the North Sea region in the future (Chap. 5) will increase evapotranspiration through a larger atmospheric water demand (Kellner 2001; Wu et al. 2010; Peichl et al. 2013). Other important variables influencing the energy and water budget of peatlands are net radiation and incoming short-wave radiation (Moore et al. 2013; Runkle et al. 2014). A continuation of the ‘brightening period’ through reduced aerosol loading (Wild et al. 2005) in Europe could increase evapotranspiration (Oliveira et al. 2011) leading to lower peatland water tables. However, most atmospheric models simulate a future decrease in shortwave radiation in the northern North Sea region and an increase in short-wave radiation in the south (Chap. 5). A long-term lowering of the water table due to increased evapotranspiration is expected to be modulated by changes in leaf area and the distribution of plant functional groups leading to increased surface resistance, reduced evapotranspiration and an attenuated fall in water tables (Bridgham et al. 1999; Moore et al. 2013). Desiccation of the moss layer during summer droughts can also lead to reduced evapotranspiration (Sottocornola and Kiely 2010). Higher winter precipitation as projected for the North Sea region (Chap. 5) would lead to larger winter discharge from peatlands and probably to a larger lateral export of dissolved organic matter and nutrients (e.g. Tranvik and Jansson 2002; Worrall et al. 2002, 2003; Pastor et al. 2003; Holden 2005). Increasing summer drought and winter rainfall would enhance peatland erosion and export of dissolved organic carbon (DOC) and particulate organic carbon (POC) in susceptible areas, particularly in the upland blanket bogs of the UK and Norway (e.g. Bower 1960, 1961; Francis 1990; Evans et al. 2006b; Evans and Warburton 2010). Drier mire surfaces in summer would enhance the risk of peatland fires, which lead to strong local emissions of CO₂ (Davies et al. 2013) and waterborne DOC (Holden et al. 2007; Clutterbuck and Yallop 2010). At bare peat sites (under peat extraction or crop cultivation), higher wind speeds and rainfall intensities can lead to strong aeolian or water erosion of peat (Warburton 2003).

6.2 Climatic Impacts on Biotic Interactions

Climate change is expected to have large impacts on biotic processes in peatlands (e.g. Heijmans et al. 2008; Charman et al. 2013). Higher temperatures will generally increase microbial peat decomposition and carbon mobilisation (Ise et al. 2008), leading to greater concentrations of DOC in peatland surface and soil pore waters (Freeman et al. 2001, Fig. 11.10). The mobilised DOC can be mineralised and released to the atmosphere as CO₂ (e.g. Silvola et al. 1996; Lafleur et al. 2005) or CH₄ (e.g. Dunfield et al. 1993; Moore and Dalva 1993; Daulat and Clymo 1998; Hargreaves and Fowler 1998; Fig. 11.11) depending on the reduction-oxidation status of the organic soils. It can also be laterally exported with the peatland discharge into limnic systems (Tipping et al. 1999; Worrall et al. 2003). Warming-induced increases in soil carbon emissions result partly from the direct temperature effect on soil microbial physiological processes and growth and partly from better substrate availability in the soil pore water caused by higher plant productivity (Mikkelä et al. 1995; Joabsson et al. 1999; Van den Pol-van Dasselaar et al. 1999).

Fig. 11.10

Laboratory observations of increased concentrations of dissolved organic carbon (DOC) and phenolic compounds in peat soil in response to rising temperature (Freeman et al. 2001)

Fig. 11.11

Effects of temperature on methane (CH₄) efflux from 30 cm diameter Sphagnum papillosum mire cores. Flux measurements were performed in dark conditions (after Daulat and Clymo 1998)

These findings from field and laboratory studies have been incorporated in local, regional and global soil process models (e.g. Walter et al. 2001; van Huissteden and van den Bos 2006; Bohn et al. 2007; Meng et al. 2012). The models predict considerable increases in CH₄ emissions from peatlands over the coming century due to warming as long as wetland area and soil moisture conditions remain unchanged. However, several global models tested within the Wetland and Wetland CH₄ Intercomparison of Models Project (WETCHIMP) predict a decrease in wetland area in the North Sea region in response to higher temperatures, which is likely to lead to lower CH₄ emissions (Melton et al. 2013); however, it should be noted that these models show a very wide range of responses.

Whether increased peat decomposition and carbon mobilisation due to higher temperatures leads to lower net ecosystem productivity and to higher net carbon emissions, will depend on the land use of peatlands. For the period 1978–2003, Bellamy et al. (2005) reported carbon losses from all soil types across England and Wales, with particularly strong losses from peat soils. However, it is not clear from such data whether carbon was lost due to climate change or to concomitant changes in atmospheric chemistry and land use (see Sect. 11.4.1 and 11.6.3). Likewise, mapping of peat soils in the Dutch province of Drenthe showed that 42 % of the area of peat soils was converted to mineral soils in the last 30–40 years by carbon loss due to drainage and agricultural management; on average 1 cm peat thickness was lost per year (De Vries et al. 2008). In near-natural peatlands with typical mire vegetation, peat accumulation is expected to increase in response to higher mean annual temperatures because the benefit to primary productivity will be higher than for ecosystem respiration (Loisel et al. 2012; Charman et al. 2013). Longer growing seasons (higher winter temperatures, shorter snow-cover duration) allows the vegetation to take up more photons over the year leading to higher plant productivity and peat accumulation. This increase in net carbon uptake under a warming climate—a negative feedback on climate warming—will be modulated by cloud cover and levels of photosynthetically active radiation (Yu et al. 2010; Loisel et al. 2012; Charman et al. 2013).

If peatland water tables become significantly lower due to increased evapotranspiration and/or decreased summer precipitation, peat decomposition and the release of CO₂ will be enhanced (e.g. Silvola et al. 1996; Laine et al. 2009). Higher N₂O emissions can also be expected with lower peatland water levels (Martikainen et al. 1993; Regina et al. 1996; Goldberg et al. 2010). On the other hand, CH₄ production and emission will decrease, while CH₄ oxidation will be enhanced (e.g. Daulat and Clymo 1998; Hargreaves and Fowler 1998; Nykänen et al. 1998; Le Mer and Roger 2001; Laine et al. 2007, 2009).

The impact of water table drawdown on net ecosystem productivity depends on the response of peatland vegetation, which is difficult to predict and can vary strongly with the micro-topography of mires (Malmer et al. 1994; Strack and Waddington 2007; Lindsay 2010). Since Sphagnum mosses have neither roots nor vessels to transport water from deeper soil layers, they rely on high water tables. Water that is lost at the soil surface through evaporation can be replaced by precipitation and capillary rise in Sphagnum peat and vegetation. But—depending on the Sphagnum species present and the degree of peat decomposition—capillary rise is only efficient in this regard if water tables are not lower than 0.5 m below the land surface (Clymo 1984). Thus, long-lasting drought may damage the vitality of Sphagnum mosses (Gerdol et al. 1996; Bragazza 2008; Breeuwer et al. 2009; Robroek et al. 2009).

Medium- to long-term changes in climatic variables such as temperature and precipitation have complex effects on mire vegetation composition that interact with changes in CO₂ concentration and nutrient availability (e.g. Heijmans et al. 2008; Breeuwer et al. 2010, see also Sect. 11.6.3). In a Sphagnum-dominated bog in southern Sweden, significant shifts in vegetation composition since the 1950s indicate higher nutrient availability, higher productivity, and drier and shadier conditions due to enhanced tree and shrub growth, which have probably all been triggered by warming (Kapfer et al. 2011). Projected climate change, particularly higher summer temperatures and lower summer precipitation may reduce the bio-climatologically suitable space for blanket bogs in the UK (Clark et al. 2010; Gallego-Sala et al. 2010; Gallego-Sala and Prentice 2013, Fig. 11.12). Berry and Butt (2002) suggested that the dominant species in lowland raised bogs of Scotland would find suitable conditions, but that some rare species will probably lose suitable climate space under the expected future climatic conditions. Since lower summer precipitation and higher summer temperatures are projected for the entire North Sea region (Chap. 5), mires on the European continent—especially ombrogenous mires—are likely to experience falling summer water levels with negative impacts on typical mire plant communities. However, Lindsay (2010) cautioned that the climate envelope approaches applied might underestimate the importance of local atmospheric humidity, cloud cover and mist frequency for mire occurrence. Other studies even propose—on the basis of biogeographical, bioclimatological and ecophysiological reasoning—that northern oceanic peatlands will expand owing to the increased oceanicity of the future climate (Crawford 2000; Crawford et al. 2003). Lindsay (2010) also stated that mires with a well-developed acrotelm and microtopography might be able to react to climate change without losing their ability to grow and sequester carbon. Mires that are already degraded will have much less resilience to the projected changes in climate (Lindsay 2010).

Fig. 11.12

Area covered by the bioclimatic envelope of blanket peatlands as predicted by the model PeatStash using the bioclimatic thresholds associated with the 1961–1990 baseline climate for the UKCIP02 high and low emissions scenarios (‘High’ and ‘Low’, respectively) for three time periods: 2020s, 2050s and 2080s. An Ordnance Survey/EDINA supplied service (© Crown Copyright/database 2009) and Met Office/UKCIP gridded climate data (UKCIP02 © Crown Copyright 2002) (Gallego-Sala et al. 2010)

6.3 Competing Effects of Climate Change and Other Influences

In addition to the changes in climate, there have also been changes in the intensity of land use and atmospheric chemistry (such as CO₂ concentration, and atmospheric deposition of nitrogen and sulphur) over recent decades, with strong impacts on near-natural and degraded peatlands. These may intensify, mask or reverse the effects of climate change on peatlands.

6.3.1 Atmospheric Chemistry

Research in Sphagnum bogs found that elevated atmospheric CO₂ concentrations did not affect NPP due to the strong N-limitation of these ecosystems (Berendse et al. 2001). However, CH₄ emission (Dacey et al. 1994; Hutchin et al. 1995) and DOC export (Freeman et al. 2004; Van Groenigen et al. 2011) from peatlands were enhanced under elevated CO₂ concentrations. Increased N-deposition promotes microbial peat decomposition and thus CO₂ and CH₄ emissions (Aerts et al. 1992; Aerts and de Caluwe 1999) and DOC export fluxes (Bragazza et al. 2006). Large shifts in mire vegetation composition may also occur in response to elevated CO₂ concentrations and increased N-deposition (Van der Heijden et al. 2000; Berendse et al. 2001; Fenner et al. 2007; Heijmans et al. 2008). Deposition of nitrogen and sulphur leads to acidification of top soils and thus changes the solubility and mobilisation of dissolved organic matter. The rise in DOC concentrations in limnic water bodies observed in the latter half of the 20th century in Great Britain and Sweden seems to have been mainly driven by decreasing S-deposition with the warming effect of minor importance (Freeman et al. 2001; Worrall et al. 2002; Evans et al. 2005, 2006a, 2007; Monteith et al. 2007; Erlandsson et al. 2008). However, S-deposition has also been shown to suppress CH₄ emissions because sulphate reduction is energetically favourable compared to methanogenesis (Dise and Verry 2001). Heavy air pollution can even lead to die-off of Sphagnum mosses, triggering for example peat erosion in blanket mires (e.g. Tallis 1985).

6.3.2 Land Use

Drained and degraded peatlands are hotspots of greenhouse gas emissions (e.g. Oleszczuk et al. 2008; Couwenberg 2011; Joosten et al. 2012). Mineralisation of peat organic matter and the respective CO₂ emissions are strongly related to drainage depth and management intensity (Aerts and Ludwig 1997; Dirks et al. 2000; Beetz et al. 2013; Leiber-Sauheitl et al. 2013; Schrier-Uijl et al. 2014). A project on peatlands in Germany has shown that the annual greenhouse gas balance of managed peatland areas can be estimated well from two predictor variables—mean annual water level and carbon exported by harvest—which together can be used as a proxy for management intensity (Drösler et al. 2011, 2012). Within the set of managed peatlands, greenhouse gas emissions from deeply-drained peatlands such as cropland or intensively-used pastureland are especially large (Veenendaal et al. 2007; Drösler et al. 2011; Elsgaard et al. 2012; Leiber-Sauheitl et al. 2013). Greenhouse gas emissions can stay high at such sites even when management intensity is moderated by nature conservancy measures (Best and Jacobs 1997; Schrier-Uijl et al. 2010; Hahn-Schöfl et al. 2011). Intensively-used peatlands are more common in lowlands than in uplands due to better accessibility and suitability for high-intensity agriculture. Burning peatland vegetation as a management practice in the UK may strongly affect carbon sequestration and dissolved organic matter export (Garnett et al. 2000; Clutterbuck and Yallop 2010; Yallop et al. 2010) although the evidence is not conclusive (cf. Worrall et al. 2007; Clay et al. 2009; Allen et al. 2013). It should be stressed that change in land use is the primary driver of changes in peatland hydrology and biogeochemistry, and probably has a stronger impact than climate change. However, some climate change effects will exacerbate the impact of human activities such as drainage, grazing, burning and peat mining (e.g. Petrescu et al. 2009). On the other hand, land use-related effects on peatlands often make them more vulnerable to climate change impacts (e.g. Parish et al. 2008).

7 Inland Ecosystems and the Wider North Sea System

Inland ecosystems have important functions within the coupled land-ocean-atmosphere system of the North Sea region. Major functions of inland ecosystems are freshwater storage and transmission, carbon storage, carbon sequestration, greenhouse gas emission and the export of dissolved and particulate organic matter to aquatic systems. While forests in the North Sea region currently sequester carbon and act as greenhouse gas sinks (Ciais et al. 2008; Luyssaert et al. 2010), agricultural systems are greenhouse gas sources through CO₂ and N₂O emissions from soils and CH₄ emissions from enteric fermentation of livestock and manure management (Schulze et al. 2009). Greenhouse gas emissions from degraded and agriculturally-used peatlands are significant in several countries of the North Sea region when compared to their total national greenhouse gas emissions, with contributions of about 5 % in Germany and Denmark, 2–3 % in the Netherlands and about 1 % in the UK (Cannell et al. 1999; Van den Bos 2003; Drösler et al. 2008, 2011; Verhagen et al. 2009; Joosten 2010; Worrall et al. 2011; Nielsen et al. 2013). On the other hand, CO₂ uptake by the few remaining near-natural peatlands in the North Sea region is negligible compared to CO₂ release by degraded peatlands or CO₂ uptake by forests. This means that reducing emissions from reclaimed peatlands is more important than the possible contribution of natural peatland to carbon sequestration.

The export of dissolved and particulate organic matter from inland ecosystems has important effects on the biogeochemistry and ecology of the receiving aquatic systems (i.e. lakes, rivers, estuaries and the North Sea) and supplies them with inputs of carbon, nitrogen, phosphorus and other important nutrient elements (e.g. Evans et al. 2005). Because export of DOC and POC is controlled by many interacting factors (e.g. temperature, nutrient supply, precipitation, evapotranspiration, run-off), its future behaviour is difficult to predict. Run-off is projected to increase in the northern part of the North Sea region and to decrease in the south (Alcamo et al. 2007). However, due to the projected warming and higher frequency of heavy rain events in the North Sea region (Chap. 5), enhanced mobilisation of soil organic matter and transport of terrestrial DOC to the limnic ecosystems and the North Sea are likely. Dissolved organic matter affects ecosystem nutrient availability (Carpenter et al. 2005), acidification of limnic systems (Oliver et al. 1983) and solubility, transport and toxicity of heavy metals and organic pollutants (Carter and Suffet 1982; Pokrovsky et al. 2005). It also regulates the photochemistry of natural waters (Zafiriou et al. 1984) and influences aquatic production of algae and bacteria (Wetzel 1992; Carpenter and Pace 1997). The export of organic matter into limnic systems can affect human health adversely since these organic substances support bacterial proliferation and lead to the formation of carcinogens when they react with disinfectants (such as chlorine) during water treatment (Nokes et al. 1999; Sadiq and Rodriguez 2004). The magnitude of DOC fluxes in rivers correlates with organic matter storage in the soils of their catchments (e.g. Hope et al. 1997). Riverine organic matter is modified strongly and largely removed through mineralisation and sedimentation during transport in rivers and estuaries (e.g. Raymond and Bauer 2000; Wiegner and Seitzinger 2001; Abril et al. 2002; Raymond et al. 2013). Thomas et al. (2005) estimated that about one million tons of DOC and POC are transported into the North Sea by rivers each year. Only 10 % of the riverine input of organic carbon is probably buried in the shelf sediments (Hedges et al. 1997; Schlünz and Schneider 2000), with the rest incorporated in the food webs of coastal seas.

Flood risk mitigation is an important issue in coastal and fluvial lowlands bordering the North Sea, especially given the projected acceleration in sea-level rise in the future due to climate change (Chap. 5). Peat soil degradation causes land subsidence by a combination of peat oxidation and compaction after drainage (Schothorst 1977). Historical subsidence—caused by drainage since medieval times—often combined with peat extraction for fuel, in coastal peatlands of the Netherlands, Germany and eastern Britain may have resulted in up to several metres of subsidence (Godwin 1978; Borger 1992; Verhoeven 1992; Hoogland et al. 2012). In the eastern British fenlands, compaction and peat oxidation has resulted in up to 4 m of subsidence in 150 years (Godwin 1978). In Dutch managed peatlands, subsidence is ongoing at up to one centimetre per year (Hoogland et al. 2012, and references therein). Under a warmer climate, peat decomposition would be even faster, particularly in drained peatlands. This would increase flood risk, induce costs for creating and managing flood protection systems and ever deeper drainage, and threaten the economic viability of agriculture. Subsidence also influences peatland hydrology and hydrochemistry. The need for increasingly deeper drainage enhances the upwelling of sulphate-rich brackish or salt water (Hoogland et al. 2012). This in turn may enhance peat decomposition by sulphate reduction, with adverse impacts on water quality by increasing dissolved and particulate organic matter and nutrient mobilisation (Smolders et al. 2006). Replenishing surface water with alkaline river water in agriculturally managed peatlands in dry periods may have a similar effect on peat decomposition.

8 Summary

The expected future impacts of climate change on terrestrial ecosystems are summarised in Table 11.1.

Table 11.1 Climate change impacts on terrestrial ecosystems of the North Sea region

Future climate change is likely to increase NPP in the North Sea region due to warmer conditions and longer growing seasons, at least if future climate change is moderate and summer precipitation does not decrease as strongly as projected in some of the more extreme climate scenarios. The physiological effects of increasing atmospheric CO₂ levels and increasing N-mineralisation in the soil may also play a significant role, but to an as yet uncertain extent.

The effects of total carbon storage in terrestrial ecosystems are highly uncertain, due to the inherent complexity of the processes involved. For example, water table effects in mires, large uncertainties in soil carbon modelling, the unknown fate of additional carbon taken up through CO₂ fertilisation, and other important drivers, such as changes in land use (e.g. forest harvest and wetland drainage). For moderate climate change, land use effects are often more important drivers of total ecosystem carbon accumulation than climate change.

Across a wide range of organism groups, range expansions to higher latitudes and altitudes, changes in phenology, and in the case of butterflies and birds, population increases in warm-adapted species and decreases in cold-adapted species have occurred in response to recent climate change. Regarding range expansions, some studies suggest substantial differences between organism groups; for example, herbaceous plants show only small or no responses while variability within other groups is large. Habitat specialists with restricted ranges have generally responded very little or even shown range contractions. Many of these often already threatened species could therefore be particularly vulnerable to climate change. Cold-adapted mountain top species are at particular risk because they have very limited habitat space in which to track climate change.

Overall effects of recent climate change on forest ecosystems within the region are limited, and major impacts on forest type distribution and forest functioning are unlikely if future warming is moderate and summer precipitation does not decrease as much as is projected in some of the more extreme climate scenarios. However, current models simulating potential impacts of climate change on forests rarely include a number of drivers of potentially rapid changes in forest functioning, such as forest pests and diseases (e.g. Kirilenko and Sedjo 2007; Jönsson et al. 2009). As a result, projections of climate-driven changes in future forest productivity, biomass and carbon storage are highly uncertain.

For grasslands, significant range expansion of thermophilous animal species (e.g. Parmesan et al. 1999), changes in flowering phenology (e.g. Fitter and Fitter 2002), and population increases (e.g. Poniatowski et al. 2012) are currently more obvious signs of climate change, than changes in plant community composition and ecosystem processes. Even in experimental studies simulating drought and warming, responses to treatments were modest (Bates et al. 2005; Grime et al. 2008; Kreyling 2010). Evidence from observational and correlative studies is weak and speculative due to many confounding effects such as eutrophication and changes in management practice (e.g. Gaudnik et al. 2011; McGovern et al. 2011).

For heathlands, overall evidence for effects of recent climate change from experimental warming and drought treatments is also weak, variable and inconsistent, suggesting that now and in the near future, climate warming is of low significance compared to other predominant drivers of ecological change in heathland ecosystems such as eutrophication, acidification and altered management practices (e.g. Härdtle et al. 2006). For more extreme climate scenarios, however, substantial effects could be expected in heathlands. Projections of the exact nature of these future effects are highly uncertain.

The projected climatic changes for the North Sea region are likely to have significant impacts on abiotic and biotic processes in mires and drained peatlands. However, the consequences will vary widely between mires and drained peatlands. Higher temperatures and longer growing seasons will increase NPP, but also ecosystem respiration, CH₄ emission and DOC export in mires and drained peatlands. The net effect is expected to result in increased peat accumulation in mires but accelerating peat decay in drained peatlands. In mires, lower water tables due to less summer precipitation and/or higher evapotranspiration will enhance NPP but also—and to a much greater degree—ecosystem respiration, leading to a net loss of peat organic matter and the release of CO₂. On the other hand, CH₄ emission will also be reduced, while effects on DOC export are less clear. In drained peatlands, climatic changes will have less effect on the water budget and biogeochemical fluxes since water tables are regulated.

Low summer precipitation and/or high evapotranspiration can make conditions unsuitable for some mire types. However, well-developed natural mires may have considerable resilience to climate change. The status of peatlands, namely the level of drainage and soil degradation will determine whether peatlands mitigate or exacerbate climate change.

Besides their function as a sink for atmospheric carbon, the export of dissolved and particulate organic carbon and nutrients from terrestrial ecosystems is probably the most significant process directly affecting the North Sea system. Because this export is controlled by many interrelating factors (temperature, precipitation, evapotranspiration, run-off, human impact), its future development is very uncertain and therefore difficult to predict.