Thermal springs, Arctic steppe, and Lime-rich fen margin are Critically Endangered CR ecosystem types on Svalbard. The remainder of the ecosystem types are assessed as being of Least Concern LC, with the exception of four Near Threatened NT ecosystem types. In general, it is climate change that is the most significant impact factor.

High Arctic areas are to be found in Norway on the Svalbard archipelago (including Bjørnøya island), as well as on the volcanic island of Jan Mayen. The latter is not included in the assessment due to deficiencies in knowledge and data.

Description of High Arctic terrestrial areas

Svalbard consists of a few large islands and a number of lesser ones, including Edgeøya (5 073 km2), Barentsøya (1 288 km2) and Nordaustlandet (14 443 km2). The total terrestrial area of the Svalbard archipelago is 61 022 km2.

Climate

The climate on Svalbard is strongly influenced by the North Atlantic Current (an extension of the Gulf Stream). Overall, the southern and western areas on the archipelago are less cold than the northern and eastern areas. However, regional gradients are often more dominant and influence the terrestrial environment to a significantly greater degree than large-scale conditions. For example, the degree to which the climate is oceanic is extremely important. The coastal areas on the west side of Spitsbergen are probably situated in a weakly oceanic section, while the inland valleys have weakly continental conditions and a few locations which are in a clearly continental section (Elvebakk and Nilsen 2002). These are the only places in Norway where a clearly continental section occurs in an area that is free from ice. In a Norwegian context, Spitsbergen has the broadest gradient, from weakly oceanic to clearly continental, and with locations within the shortest distance from each other.

On the west side of Spitsbergen and on the island of Prins Karls Forland, the oceanic climate combined with the North Atlantic Current, means that the growing season can be somewhat longer than in the more continental fjord areas, and this compensates for lower average temperatures in the months when growth occurs (Joly et al. 2015). The degree to which the climate is oceanic declines generally in a north-east direction, because the pack ice is more stable in these areas and reduces the effect of the factors that generate an oceanic climate.

The Middle Arctic Tundra zone is the most hospitable climate zone. It has a restricted distribution in the lowlands of the inner fjord areas of Spitsbergen, as well as on Bjørnøya (Bear Island). The North Arctic Tundra zone has a broad distribution at the elevation above the Middle Arctic Tundra zone. Outside the inner fjord areas on Spitsbergen this zone dominates in the lowlands, probably also on the poorly investigated islands of Edgeøya and Barentsøya. The Polar Desert zone is the northernmost bioclimatic zone. It has a large distribution, at higher altitudes, throughout the entire archipelago, dropping down to sea level in the north-eastern areas of Spitsbergen, and it is the dominant zone on Nordaustlandet.

Bedrock and sediment conditions

Bedrock geology and sediment conditions also play a large role in the development of ecosystem types on Svalbard. Ecosystem types where the sediment or bedrock are exposed are significantly more common than in boreal environments. Sedimentary rocks predominate on the archipelago and date from a number of geological eras. Sandstone, siltstone and shale dominate on Spitsbergen, from Reinsdyrflya in the north, to Sørkapp (South Cape) in the south. The sedimentary rocks are in general very exposed to frost weathering. Most of the valley sides are covered by extremely active landslide slopes and in more gradual terrain, at higher elevations, there is often a thick layer of fine-grained weathered material from the autochthonous rocks. Weathered material is also found in sediment-filled basins in the valley corridors. Frost processes in the active layer (the layer of the permafrost that melts in the course of the summer) work upon, and to a large degree sort, the fine material from the sedimentary rocks, and give rise to structures such as polygon fields and frost basins. The continuous mixing and crushing of the sediments also causes the dissolution of minerals in the rocks, and usually leads to the pH in the soil being relatively high. This is described as "intermediately lime-rich" in the NiN classification system. In areas with sedimentary rocks, from Sassendalen in the centre of Svalbard and northwestward towards Billefjorden and Dicksonfjorden, there are large occurrences of limestone and evaporites (gypsum), which provide extremely calcareous (lime-rich) conditions in these areas. Carbonate rocks are also dispersed throughout the archipelago, including at Krossfjorden, and on the island of Nordaustlandet.

Along the west coast of Spitsbergen there is a belt of metamorphic rocks, which are the remains of the formation of a chain of mountains. In this area there are occurrences of harder rocks, which may result in acidic substrates, and with larger local variation than in sedimentary areas. In the northwest, from just south of Magdalenefjorden, and northwards and eastwards to the west side of Raudfjorden, there is an area of granite, which is a source of extremely acidic substrates. Corresponding conditions are found in places on the north side of Nordaustlandet. These are the only places on Svalbard where the ecosystem type Boulder field T27 can develop because the hard granite weathers extremely slowly into fine material. East of Wijdefjorden (Ny-Friesland), and towards the north part of Nordaustlandet, there are older rocks with a mixture of sedimentary, intrusive and metamorphic rocks, and thus greater variation in the substrates.

Assessed ecosystem types

All major types at the natural system level in NiN, with occurrences in the Arctic, are essentially considered to be assessment entities. A thorough review was undertaken of all minor types, combinations of minor types, as well as "fractions" of minor types created on the basis of the classification system in NiN. This was done to assess the extent to which combinations/fractions of minor types could be subject to a greater level of threat than the major type of which they are a part, and thereby qualify as separate assessment entities. In total, 29 entities that are relevant in the High Arctic were assessed, 26 of which are at the major type level. Three entities are nevertheless combinations or fractions of minor types.

A combination of Bird-cliff meadows (Bird-cliff meadow T8-1, T8-2, T8-3) was created to enable Red List assessments of the ecosystem types associated with larger bird cliffs, without including areas heavily fertilized with avian excrement in flat terrain and on islands; the latter being sub-types that are on the rise due to the increase in geese populations.

It was also necessary to single out and create fractions of permafrost-dependent wetlands as assessment entities. However, the peat-forming and salt-affected areas of wetlands dependent on permafrost are not captured by the current descriptive classifications in NiN. The descriptions of wetland types are gathered under the major type Arctic permafrost wetland V7. This type is described as an environment characterised by swamps without significant peat formation, dominated by vascular plant helophytes and mosses that float in still water that lies level with, or just above, the surface of the ground. Therefore wetland types that are peat-producing and dependent on permafrost, which have a far greater distibution on Svalbard than areas with swamp characteristics, fall outside the definition of Arctic permafrost wetland V7. Neither can such areas be incorporated into Open fen V1. Ordinary mire systems are not dependent on permafrost and meltwater.

In order to integrate peat-forming, permafrost dependent wetlands into the NiN system, as assessment entities, we have introduced the terms "open fen" and "fen margin". These ecosystem types have a water table a few centimetres below the surface of the ground, and produce peat. There is often a lush and dense vegetation of grass and sedge species (not halophytes). The ecosystem types within Arctic permafrost wetland V7 have been renamed "Permafrost freshwater swamp" to distinguish them from peat-producing wetland systems. We have also included "Permafrost brackish swamp" which represents the brackish water version of the corresponding system. The latter represents a variation that is not currently described in NiN. The dividing line between open fen and permafrost swamp largely corresponds with the results of Vanderpuye (2002) who described wetland areas in Sassendalen. Strongly intermediate to extremely calcareous (lime-rich) permafrost fen edge and permafrost brackish water swamp were eventually assessed separately for the Red List as fractions of Arctic permafrost wetland.

We also emphasise that we have selected a stricter definition of the term moss tundra than that which is described in NiN, in order to avoid an overlap with the wetland ecosystem types mentioned above. Our assessments of Moss tundra encompass only moss-dominated, peat-producing areas in sloping terrain without any still water, as the meltwater immediately drains away, and the peat gradually becomes integrated with the permafrost. Such systems have a shallow active layer and are particularly well-developed beneath bird cliffs. Nonetheless we do not regard avian excrement as a prerequisite for the development of this ecosystem type.

Ecosystem types on the Red List

Of the 29 types which are assessed, 7 ecosystem types are on the Red List.

Critically Endangered ecosystem types

Thermal springs, Arctic steppe and Lime-rich fen margin are assessed as Critically Endangered CR on Svalbard. The reason they are critically endangered is a very small distribution area, combined with ongoing environmental degradation or decline (criterion B).

The ecosystem type Thermal springs exists in two isolated areas on Bockfjorden. The actual distribution area is estimated to be less than 1 km2, and historic degradation has been registered from damage to the microstructures on the sinter terraces, caused by walking across them, and bathing in the largest springs. Arctic steppe also has a very small distribution area, limited to the clearly continental section primarily in Inner Wijdefjorden, as well as to fragments in Sassendalen, and perhaps an area in Dicksondalen that has not yet been investigated. The ecosystem type depends on very dry conditions and is expected to be extremely vulnerable during the transformation to a more humid climate. Lime-rich fen edge is very rare and exists in small scattered patches in calcareous areas in inner Spitsbergen. The ecosystem type has some very rare species and the land has already been converted to other uses in the Longyearbyen area and is probably over-grazed in the Ossian-Sars area in Kongsfjorden. Consequently, there has been a significant decline in area and probably ongoing degradation.

Near Threatened ecosystem types

Permafrost freshwater swamp, Polar desert, Bird cliff meadow and Tidal meadow are assessed as Near Threatened NT. Permafrost freshwater swamp is assessed as being extremely vulnerable to melting permafrost, with an increase in the thickness of the active layer. The consequence is significant degradation, in the form of a transformation in the direction of an open fen. The ecosystem type is presumed to expand slowly into new areas as bioclimatic limits gradually migrate, but not sufficiently quickly that a decrease in net area is avoided.

Polar desert, like Arctic steppe, is extremely sensitive to climate change. This is the coldest part of the Arctic and it is therefore not possible for the ecosystem type to expand into colder zones. The areas with Polar desert will evolve in the direction of Grass tundra, and we presume that this development with loss of area and degradation will occur so quickly that red-listing in the Near Threatened NT category will be required.

Bird cliff meadows associated with larger bird cliffs depend on populations of sea birds. Our assessment is that a large proportion of the seabird populations associated with bird cliffs can be impacted by changes in ocean currents and temperatures, migration of fish stocks, and the distribution of fishing quotas. The altered living conditions for sea birds will have an impact on the ecosystem type Bird cliff meadow (all levels of fertilization but perhaps the over-fertilized in particular). The changes in ocean temperatures, and the distribution of fish stocks, occur so quickly that we can expect a decline in area within 50 years, not only on the individual bird cliff, but also in terms of where birds nest, in order to have easy access to food resources. This is the basis on which we consider Bird cliff meadow to be Near Threatened NT in accordance with criterion D.

Tidal meadows are expected to be negatively impacted as fjords that previously had solid ice during winter are now characterised by open water and ice-scouring. The fragile Tidal meadows only develop where they are protected from breaking waves, and ice-scouring, by stable fjord ice. It has therefore been observed that areas that previously had Tidal meadows have become sterile mud flats. As a very large proportion of Tidal meadows on Svalbard exists in the actual fjord zone, the ecosystem type is red-listed in the category Near Threatened NT.

Ecosystem types of Least Concern LC

There are 22 assessment entities that are assessed as being of Least Concern LC on Svalbard. In common for these is that they are barely influenced by climate change, or that ongoing climate change either encourages the development of these particular ecosystem types, or makes it possible for them to expand into new areas almost as quickly as they collapse in other areas. One of these ecosystem types is Moss tundra which our assessment indicates will decline in the course of the assessment period of 50 years, but which nonetheless does not reach the threshold values for Near Threatened NT, according to the applicable definitions for criteria A and C. Permafrost melting will probably lead to the active layer becoming deeper in areas with Moss tundra, vascular plants will appear, and there will be a transformation to Arctic-alpine heath and lee side T3. The gradual formation of new Moss tundra, as areas in the north and at higher elevations develop suitable conditions, is presumed to be a much slower process, and there is therefore a net decline in this exclusively High Arctic ecosystem type.

Impact factors

The difficulty of accessing Svalbard, and strict conservation of large areas, means that the majority of impact factors associated with land use are hardly relevant. A small exception is tourism which causes wear and tear on landing areas which are subject to high use. This is a relevant impact factor for some few ecosystem types which have a very small distribution, such as the thermal springs in the Bockfjorden area.

Climate change

Undoubtedly the most significant impact factor is ongoing climate change. We have used the climate projections that are outlined in Førland et al. (2010). In brief, there is mention of moderate temperature increases during summer of around 2 – 3 degrees over the entire archipelago, but somewhat higher in the southeast around the island of Edgeøya. At other times of the year, the temperature increase is however more dramatic, up around 6-8 degrees during winter in the northeast. The change is associated with the absence of stable pack ice in this area during winter. In terms of precipitation, it is estimated to be relatively stable during summer, but the winter precipitation (snow) is expected to increase, especially in eastern parts of the archipelago from Edgeøya to Nordaustlandet (an increase of 30 – 40 % in the years leading up to 2040). It must however be mentioned that, in general, Svalbard currently has modest amounts of precipitation, and some areas are very dry. A doubling of winter precipitation is therefore not as dramatic as it sounds but can be enough to place marginal occurrences of areas in the inner fjord regions, which are clearly classified as having a continental climate, under great stress.

The effects of climate change will have both positive and negative impacts on the ecosystem types of Svalbard. Typically a number of ecosystem types that are associated with the mid-Arctic tundra zone, such as Arctic-alpine heath and lee side T3 and Snowbed T7, will increase their distribution due to the increase in distribution of the mid-Arctic tundra zone. Likewise the polar desert zone will decrease in distribution and the ecosytem type Polar desert T28 is probably in decline. Several other ecosystem types will decline in existing areas of occurrence but there will be a corresponding increase in newly available areas. This applies to Open fen V1 and Arctic alpine dry-grass heath T22.

Melting permafrost, due to climate change, will probably lead to changes in several ecosystem types. An increase in the depth of the active layer will lead to wetlands dependent on a shallow active layer developing more characteristics of terrestrial ground, and transitioning to other wetland types or terrestrial systems. Open active landslide T17 will clearly benefit from ongoing melting permafrost which leads to more landslides.

In general, it is the exclusively High Arctic ecosystem types that decline due to climate change. However, it is only a selection of these that appear on the Red List. The reason is that the changes occur so slowly that the thresholds for a decline in area, or degradation, are not reached within the assessment period which is a mere 50 years into the future.

Other impact factors

For the major type T8 Bird-cliff meadow the fertilizer effect is significant. The effect depends on variations in the seabird populations.

Grazing by geese and reindeer also influence the environment on Svalbard. In recent decades the introduction of reindeer to Brøggerhalvøya peninsula, and restricted hunting, as well as initiatives to increase geese populations, mean that these species have expanded throughout the archipelago and altered the vegetation in this period of time. We have, however, based our assessment on a more historical perspective, and subsequently consider that the extinction of reindeer on Brøggerhalvøya peninsula was temporary (less than 100 years), as well as the fact that geese populations have been decimated by hunting in other places where the species lives. Both reindeer and geese have coexisted with the ecosystem types that have developed on Svalbard over thousands of years. The consequences of the relatively recent human interventions in the populations, interventions which have been intended to return them to their previous level, we perceive therefore as a restoration of something natural. The exception is where there are some unique effects in local areas, such as on the Lovénøyane islands in Kongsfjorden where the increase in the population of geese threatens a plant that is a national responsibility species (NRS).

Existing knowledge

The very isolated location of Jan Mayen means that it is infrequently visited and poorly mapped. The characteristic lava field substrates, of which the entire island consists, are not included in the current version of Nature in Norway NiN. The base data is assessed for this reason as being so poor that it is not possible to make a sound classification decision regarding assessment entities or to carry out a Red List assessment. The strongly oceanic climate and the succession conditions on the lava fields are presumed nevertheless to be the dominant factors that affect the terrestrial environment of the small island (577 km2), and these factors are presumed to be stable.

The land areas of Svalbard are also difficult to access, and even in 2018 only a handful of areas had been field-surveyed in such detail that surfaces and vegetation units had been localised on maps. Some field-surveys in local areas were carried out several decades ago, such as in Gipsdalen valley (Elven et al. 1990) and on the Brøggerhalvøya peninsula (Brattbakk 1981). Some comprehensive vegetation maps of Svalbard have also been made, such as those by Brattbakk (1983), Elvebakk (2005) and Johansen et al. (2008). The first two are hand-drawn maps where the authors have attempted, to the best of their abilities, to subjectively extrapolate data from their experience of surveying in small areas, to apply to the whole of Svalbard. In the latter case, the extrapolation is carried out by the classification of satellite photos and is therefore more general. In common for all three, however, is that the base data for extrapolation is extremely restricted and the uncertainty is large. None of the three works have classifications that allow them to be transferred in any particular degree to the classification system in NiN.

In connection with the development of the first edition of Nature in Norway (1.0), descriptions of High Arctic major and minor types were developed, which incorporated descriptions of presumed Arctic parallels to alpine types. However, during the revision of NiN, which resulted in versions 2.0 and 2.1, the High Arctic ecosystem types were not re-assessed. We are aware of only one case where there has been an attempt to use NiN for classification of the environment on Svalbard (Eidesen et al. 2018). This report concluded that there is considerable variation in the natural landscape on Svalbard, of basic character, that has not been clarified, and that is not described in NiN. This impression has been confirmed during the development of the Red List assessment for High Arctic ecosystem types. There is a considerable need for further development of the NiN system so that its application on Svalbard can be improved, and preferably without the underlying assumption that the Arctic variation reflects the alpine environment in Scandinavia. In the following section we mention some of the most significant problems we have had to consider during work on the Red List.

Perhaps the most fundamental issue is that the importance of permafrost as a drainage inhibitor, and an active layer, is not used in a further classification of wetlands, and to separate tundra from alpine heath in T3 and T22. Furthermore, we believe there is a need to work out the boundaries between Arctic steppe T10 and Exposed ridge T14, as well as defining several minor types within Arctic steppe T10. Variation in marine salinity is relevant for wetland systems in the Arctic but is not included as part of the definition of the only described wetland type (Arctic permafrost wetland V7). The term moss tundra is used in the literature for several types of areas with thick moss mats on Svalbard. In NiN, Moss tundra T9 is associated with avian excrement, or secondarily with reindeer excrement. The significance of manure to develop the ecosystem type appears nevertheless to be poorly documented, and in our opinion the ecosystem type has been problematic to handle. Finally, we would like to mention that the nomenclature of Polar desert T28 is unfortunate as the ecosystem type can be confused with the bioclimatic zone of the same name. There is a need for a better description and classification of this major type.

Expert Committee

The members of the expert committee for Svalbard – High Arctic terrestrial areas were Geir Arnesen (chair), Reidar Elven and Kristian Hassel. 

References

Brattbakk I (1981). Brøggerhalvøya, Svalbard, vegetasjonskart 1:10,000. Oslo: Norsk Polarinstitutt. [In Norwegian.]

Brattbakk I (1983). Svalbards vegetasjon. In: Baadsvik K, Klokk T, Rønning OI. Fagmøte i vegetasjonsøkologi på; Kongsvoll 16.-18.3. 1980. Trondheim, Det Kongelige Norske Videnskabers Selskab, Museet. University of Trondheim. [In Norwegian.]

Eidesen PB, Arnesen G, Elven R, Søli G (2018). Kartlegging av Ringhorndalen, Wijdefjorden: En uutforsket arktisk oase. Rapport til Svalbard miljøvernfond. [In Norwegian.]

Elvebakk A (2005). A vegetation map of Svalbard on the scale 1:3.5 mill. Phytocoenologia 2005; Vol. 35.

Elvebakk A, Nilsen L (2002). Indre Wijdefjorden med sidefjordar: eit botanisk unikt steppeområde. Rapport til Sysselmannen på Svalbard utgitt av Universitetet i Tromsø 66 p. [In Norwegian.]

Elven R, Eriksen MB, Elvebakk A, Johansen BE, Engelskjøn T (1990). Gipsdalen, central Svalbard; Flora, vegetation, and botanical values. - In: Brekke B, Hansson R (eds.) Environmental atlas Gipsdalen, Svalbard. 61. Norwegian Polar Institute, Oslo. 27-66.

Førland EJ, Benestad RE, Flatøy F, Hanssen-Bauer I, Haugen JE, Isaksen K, Sorteberg A, Ådlandsvik B (2010). Klimautvikling i Nord-Norge og på Svalbard i perioden 1900-2100. Klimaendringer i norsk Arktis. Norsk Polarinstitutt rapportserie 135: 52. [In Norwegian.]

Johansen B, Tømmervik H, Karlsen SR (2009). Vegetasjonskart over Svalbard basert på satellittdata. Dokumentasjon av metoder og vegetasjonsbeskrivelser. NINA Report 456. 54 p. [In Norwegian.]

Joly D, Arnesen G, Malnes E, Nilsen L (2016). Building an indicator to characterize the thermal conditions for plant growth on an Arctic archipelago, Svalbard. Ecological Indicators 66:623-631

Vanderpuye AW, Elvebakk A, Nilsen L (2002). Plant communities along environmental gradients of high-arctic mires in Sassendalen, Svalbard. Journal of Vegetation Science 13: 875-884.