**5. Hydrotopographic types distribution and area**

Each polygon was classified into Hydrotopographic wetland types, slightly modified after the wetland types used in the Swedish Wetland Inventory (Gunnarsson & Löfroth, 2009). Of the 18 wetland types (compared with 17 types in Swedish VMI), 14 were recorded in the 116 squares for northern Sweden. There were five types that were relatively common in terms of area and percentage of total wetlands: Flat (topogenous) fen with 44%, Sloping fen with 14%, Flat or weakly raised bog with 12%, String flark fen with 11%, and Non-mire wetland with 6% (Table 3, Fig. 5). Three other types achieved moderately low levels: Limnogenous fen 4%, String mixed mire 5%, and Mosaic mixed mire 3%. The remaining six types were quite uncommon. It is significant that Flat fens are by far the most abundant wetlands. It is also noteworthy that the Non-mire wetlands are among the more common of the wetlands. The wetland types missing from the inventory were Plateau bog, Domed bog, Blanket bog,

and Palsa mire. Plateau bogs and Domed bogs are more common in the south of Sweden where there is more heat and more precipitation, allowing these bogs to grow upwards into distinct raised bodies (similar to Finland, Seppä, 1996). Blanket bogs are missing in our

Main Ecosystem Characteristics and Distribution of Wetlands

2. **Northern boreal** – Flat fen, Sloping fen, String mixed mire.

common types are Flat fens and Flat or weakly raised bogs.

5. **Coastal boreal** – Limnogenous fen and Strongly influenced (disturbed) fen.

Northern boreal

The main wetlands characterizing the zones are:

Arctic/ Alpine

– northern, Non-mire wetland.

mire.

Hydrotopographic wetland type

Strongly influenced

Flat or weakly raised bog

Mosaic mixed

Unclassified mire- Weakly raised bog or fen

Strongly influenced bog

mire

fen

in Boreal and Alpine Landscapes in Northern Sweden Under Climate Change 205

1. **Arctic/Alpine** – Limnogenous fen, Hill fen, Unclassified mire, Weakly raised bog or fen

3. **Upper middle boreal** – String flark fen, Flat or weakly raised bog, Mosaic-mixed

4. **Lower middle boreal** – None of the Hydrotopographic types peak in this zone, but

Some of these distributions are as expected, e.g., the Sloping fens occur in terrains that have more relief and higher effective moisture regimes, and we expect the steeper Hill fens to be more common where there are steep hills and mountains with groundwater discharges and springs on lower and toe slopes. Strongly influenced fens peak in the Coastal boreal zone, but also occur in Lower and Upper middle boreal zones where the anthropogenic influences

Flat fen 4.49 (1.08) **11.92** (1.89) 11.36 (1.86) 6.98 (1.96) 7.46 (2.01) 8.99 (0.87) Limnogenous fen **1.58** (1.10) 0.28 (0.13) 0.14 (0.09) 0.20 (0.16) 1.5 (1.47) 0.63 (0.28) Sloping fen (3-7%) 2.85 (0.87) **5.56** (1.75) 1.93 (0.58) 1.64 (1.26) 0.48 (0.36) 2.90 (0.58) Hill fen (>7%) **0.72** (0.34) 0.39 (0.23) 0.11 (0.09) 0 0 0.29 (0.10) String flark fen 0.08 (0.08) 2.13 (0.83) **4.53** (1.82) 2.19 (1.84) 0.65 (0.32) 2.19 (0.62)

Sloping raised bog 0 0 **0.02** (0.02) 0 0 0.01 (0.01)

Net bog 0 0.14 (0.14) 0 0 0 0.04 (0.04)

String mixed mire 0 **2.28** (1.60) 0.26 (0.12) 0.07 (0.07) 0.60 (0.39) 0.76 (0.44)

Non-mire wetland **2.65** (1.32) 1.02 (0.52) 0.75 (0.29) 0.09 (0.04) 0.72 (0.41) 1.12 (0.33)

Table 3. Estimates of the percentage of terrestrial area for 14 Hydrotopographic types in five Elevation Zones and the total study area with standard errors in parentheses. The largest

areas for each Hydrotopographic type across zones are indicated in bold.

Upper middle boreal

Lower middle boreal

0 0.06 (0.06) 0.36 (0.33) 0.19 (0.11) **1.60** (1.15) 0.31 (0.15)

0.06 (0.05) 1.42 (0.93) **4.90** (1.36) 4.72 (3.35) 1.29 (0.66) 2.54 (0.69)

0 0 **0.56** (0.56) 0.30 (0.30) 0.35 (0.18) 0.23 (0.16)

0.12 (0.13) 0.76 (0.52) **1.69** (1.08) 0 0 0.68 (0.32)

**0.21** (0.21) 0.08 (0.07) 0.12 (0.11) 0 0.01 (0.01) 0.10 (0.06)

Coastal boreal

Total study area

sample because the precipitation is too low (ca. 600 mm). They are rare in Sweden, but are found in Jämtland County where precipitation is high enough to allow their development (Götbrink & Haglund, 2010). Palsa mires were not found in the sample, but they do occur in the subalpine zone in the study area. Palsas from near Hemavan in Västerbotten (Zuidhoff & Kolstrup, 2000), and in three locations northwards in Norrbotten (Zuidhof, 2003), have been studied from the points of view of climatic data, soil temperatures, and geomorphological properties. Palsas are frozen peat structures that occur as the raised mound features of a mixed mire. They are difficult to recognize from aerial photos owing to their similarity to the small bog mound features in mixed mires at lower elevations.

Sloping bog and Net bog were both recorded only once in the sample in northern Sweden. Most sloping peat bodies in northern Sweden are minerotrophic sloping fens, sloping mixed mires with strings and flarks, or flark fens, by virtue of receiving some amount of input from mineral soil. However, large weakly sloping peatlands can become quite acidic and poor in base cations in their centres, and may only have scattered minerotrophic indicators (e.g., Carex rostrata). Net (patterned) bogs may be the result of the conflux of two weak flow patterns coming together approximately at right angles, which cause ridges to form at right angles to each other, eventually merging into net-shaped ridges. The flarks in the netshaped ridges become isolated from mineral soil water influx, and eventually the whole net and flark complex becomes ombrogenous.

Fig. 5. Estimates of the total areas of the 14 Hydrotopographic mire types in the study area of Northern Sweden. Error bars indicate standard errors.

#### **5.1 Hydrotopographic wetland types across five zones**

In Table 3 we present the area percentage for the 14 types across the zones. We found that a majority of the types (eight) peaked in the Northern and Upper middle boreal zones. These two zones had also the highest proportion of wetland cover in total (Table 1). Most of the other six types had highest peaks in adjacent zones, and Hill fen which was most frequent in Arctic/Alpine had highest mean area in Upper middle boreal. Since there were only three Hill fen polygons sampled in the Upper middle boreal, at least one of these had a very large area.

The main wetlands characterizing the zones are:

204 Ecosystems Biodiversity

sample because the precipitation is too low (ca. 600 mm). They are rare in Sweden, but are found in Jämtland County where precipitation is high enough to allow their development (Götbrink & Haglund, 2010). Palsa mires were not found in the sample, but they do occur in the subalpine zone in the study area. Palsas from near Hemavan in Västerbotten (Zuidhoff & Kolstrup, 2000), and in three locations northwards in Norrbotten (Zuidhof, 2003), have been studied from the points of view of climatic data, soil temperatures, and geomorphological properties. Palsas are frozen peat structures that occur as the raised mound features of a mixed mire. They are difficult to recognize from aerial photos owing to

their similarity to the small bog mound features in mixed mires at lower elevations.

and flark complex becomes ombrogenous.

Sloping bog and Net bog were both recorded only once in the sample in northern Sweden. Most sloping peat bodies in northern Sweden are minerotrophic sloping fens, sloping mixed mires with strings and flarks, or flark fens, by virtue of receiving some amount of input from mineral soil. However, large weakly sloping peatlands can become quite acidic and poor in base cations in their centres, and may only have scattered minerotrophic indicators (e.g., Carex rostrata). Net (patterned) bogs may be the result of the conflux of two weak flow patterns coming together approximately at right angles, which cause ridges to form at right angles to each other, eventually merging into net-shaped ridges. The flarks in the netshaped ridges become isolated from mineral soil water influx, and eventually the whole net

Fig. 5. Estimates of the total areas of the 14 Hydrotopographic mire types in the study area

In Table 3 we present the area percentage for the 14 types across the zones. We found that a majority of the types (eight) peaked in the Northern and Upper middle boreal zones. These two zones had also the highest proportion of wetland cover in total (Table 1). Most of the other six types had highest peaks in adjacent zones, and Hill fen which was most frequent in Arctic/Alpine had highest mean area in Upper middle boreal. Since there were only three Hill fen polygons sampled in the Upper middle boreal, at least one of

of Northern Sweden. Error bars indicate standard errors.

**5.1 Hydrotopographic wetland types across five zones** 

these had a very large area.


Some of these distributions are as expected, e.g., the Sloping fens occur in terrains that have more relief and higher effective moisture regimes, and we expect the steeper Hill fens to be more common where there are steep hills and mountains with groundwater discharges and springs on lower and toe slopes. Strongly influenced fens peak in the Coastal boreal zone, but also occur in Lower and Upper middle boreal zones where the anthropogenic influences


Table 3. Estimates of the percentage of terrestrial area for 14 Hydrotopographic types in five Elevation Zones and the total study area with standard errors in parentheses. The largest areas for each Hydrotopographic type across zones are indicated in bold.

Main Ecosystem Characteristics and Distribution of Wetlands

(e.g., Gunnarsson & Löfroth, 2009).

Field and bottom layer type Occurrence

in Boreal and Alpine Landscapes in Northern Sweden Under Climate Change 207

summarize the occurrence of the Field and bottom cover types in the 3229 polygons. The most common types are Sphagnum (33.5%), Low sedge (21.1%), Dwarf shrub (17.6%), Graminoid/dwarf shrub (11.8%), Graminoid- and/or herb (6.8%), and Tall sedges/graminoids (3.3%). Five percent of the polygons were missing a field/bottom layer, probably owing to coverage by shadows, water, peat extraction, or dense cover of shrub or tree vegetation (the

In Table 4 we give a short, selected list of some dominant species that characterize the different Field and bottom layer types. Since the Field and bottom types were developed for air photo interpretation, they are necessarily quite broad and reflect main appearance and life form of the vegetation. Finer community types of course can be identified with the more detailed ground survey work that is done in each of the permanent squares. Many other species lists and descriptions of the wetland vegetation in northern Sweden can be found

**Field/bottom layer missing** 5.0 Peat extraction or water, and polygons with

**Graminoid and/or herb** 6.8 Graminoids: *Carex rostrata*, *Carex lasiocarpa*,

**Graminoid-dwarf shrub** 11.8 See Graminoids above and Dwarf Shrubs

**Dwarf shrub** 17.6 *Andromeda polifolia*, *Betula nana*, *Calluna* 

**Reeds** 0.1 *Phragmites australis*, *Typha latifolia*, *Iris* 

**Tall sedges/graminoids** 3.3 *Equisetum fluviatile*, *Scirpus lacustris*, *Carex* 

**Low sedge** 21.1 *Carex chordorrhiza*, *C. livida*, *Eriophorum* 

**Sphagnum mosses** 33.5 *Sphagnum balticum*, *S. capillifolium*, *S.* 

**Logging residues** 0.0 Branches, twigs and leaves.

(3229 total polygons), and typical abundant species for each type.

**Other mosses** 0.6 *Scorpidium scorpioides*, *Campylium stellatum*,

**Layer cannot be interpreted** 0.1 Shadows from structures on the ground or

Table 4. Field and bottom cover Types, and percent frequency of the types in the data set

*nitens*.

**Lichen-dwarf shrub** 0.1 See Lichen and Dwarf Shrub. **Lichen** 0.2 *Cetraria islandica*, *Cladonia mitis*, *C.* 

Typical species

> 50 % tree cover.

*Potentilla palustris*.

*Vaccinium uliginosum*.

*alpinum*, *T. cespitosum*.

*squarrosum*, *S. tenellum*.

below.

*pseudacorus*.

*Calamagrostis purpurea*, *Eriophorum angustifolium*, *Molinia coerula*; Herbs: *Dactylorrhiza* spp., *Menyanthes trifoliata*,

*vulgaris*, *Empetrum nigrum*, *Ledum palustre*,

*rangiferina*, *Racomitrium lanuginosum*.

*acuta*, *C. aquatilis*, *C. rostrata*, *C. lasiocarpa*.

*vaginatum*, *Rhynchospora alba*, *Trichophorum* 

*cuspidatum*, *S. fuscum*, *S. papillosum*, *S.* 

*Drepanocladus revolvens*, *Tomenthypnum* 

from clouds obstruct the view.

field/bottom layer is not registered when the trees have > 50 % crown cover).

(% of polygons)

have been greatest. The greatest frequencies of Limnogenous fen and Non-mire wetlands are in the Arctic/Alpine, which may relate to higher areas of floodplains in the upper reaches of rivers and along lakes there than in interior regions.

The Hydrotopographic types are defined on the basis of slope of the topography, flow patterns of the water, surface patterns of vegetation and Microtopographic elements, and the mineralogical characteristics of the water. A fundamental division is **minerogenous (fen)**, influenced to any degree by water derived from mineral soil, versus **ombrogenous (bog)**  nourished only by rain water. Minerogenous is further subdivided into flat (**topogenous)** versus sloping **(soligenous)**. Surface movement of water and other complex interrelationships with the surface vegetation and near surface peat causes the formation of string or net patterns. One must take into account the influence of adjacent open water systems and occasional flooding, and this is done in the categories of **Limnogenous fens and Marsh.**  Obviously, it is essential to develop diagnostic features for these conditions, and much of the NILS training consists of interpreting these features from features on the remote imagery. First and foremost is separating ombrogenous from minerogenous. This means using peatland form and vegetation to recognize bog as rounded or elongated peat bodies that are isolated from mineral soil water influence, and are flat or slightly raised in their centres.

We recognize several lines of variation or gradients of Hydrotopographic types within the classification:


It must be noted that the NILS inventory thus far does not distinguish the wetlandspeatlands-mires that are well-covered with trees, known as swamp forest, treed peatlands, and treed wetlands. These conditions have been assessed by Hånell (1989) and should be added to the NILS air photo interpretation scheme for a more complete representation of the total wetland-mire-peatland area. Many of the forested peatlands have been influenced or disturbed by drainage owing to long history of forest drainage to improve growth of trees. It has yet to be determined how the open wetlands-mires-peatlands included in NILS will merge with the treed wetlands. Undoubtedly the area of wetlands-peatlands will be considerably expanded when the treed wetland areas are added.

Some north-south gradients in distribution of types across the study area are discernible, indicating that there are higher frequencies of Sloping fen, Hill fen, Strongly influenced fen, Strongly influenced bog, and Mosaic mixed mire in the south. Higher percentages for the Strongly influenced fens and bog types in the south can most likely be explained by more frequent farming and forestry activities. Higher percentages of Limnogenous fens and Nonmire wetlands in the north may be related to a higher frequency of flooded lake and river margins and floodplains. There are also higher frequencies of String flark fens and String mixed mire types in the north, suggesting a more common occurrence of soligenous flow in sloping drainage ways and higher flooding stages during spring melt, which is associated with more frequent string and flark patterns.

#### **6. Field and bottom cover**

General classes of vegetation physiognomy are recognized under Field and bottom cover. In the photo interpretation, the most dominant class was assigned to each polygon. In Table 4 we

have been greatest. The greatest frequencies of Limnogenous fen and Non-mire wetlands are in the Arctic/Alpine, which may relate to higher areas of floodplains in the upper

The Hydrotopographic types are defined on the basis of slope of the topography, flow patterns of the water, surface patterns of vegetation and Microtopographic elements, and the mineralogical characteristics of the water. A fundamental division is **minerogenous (fen)**, influenced to any degree by water derived from mineral soil, versus **ombrogenous (bog)**  nourished only by rain water. Minerogenous is further subdivided into flat (**topogenous)** versus sloping **(soligenous)**. Surface movement of water and other complex interrelationships with the surface vegetation and near surface peat causes the formation of string or net patterns. One must take into account the influence of adjacent open water systems and occasional flooding, and this is done in the categories of **Limnogenous fens and Marsh.**  Obviously, it is essential to develop diagnostic features for these conditions, and much of the NILS training consists of interpreting these features from features on the remote imagery. First and foremost is separating ombrogenous from minerogenous. This means using peatland form and vegetation to recognize bog as rounded or elongated peat bodies that are isolated from

We recognize several lines of variation or gradients of Hydrotopographic types within the

It must be noted that the NILS inventory thus far does not distinguish the wetlandspeatlands-mires that are well-covered with trees, known as swamp forest, treed peatlands, and treed wetlands. These conditions have been assessed by Hånell (1989) and should be added to the NILS air photo interpretation scheme for a more complete representation of the total wetland-mire-peatland area. Many of the forested peatlands have been influenced or disturbed by drainage owing to long history of forest drainage to improve growth of trees. It has yet to be determined how the open wetlands-mires-peatlands included in NILS will merge with the treed wetlands. Undoubtedly the area of wetlands-peatlands will be

Some north-south gradients in distribution of types across the study area are discernible, indicating that there are higher frequencies of Sloping fen, Hill fen, Strongly influenced fen, Strongly influenced bog, and Mosaic mixed mire in the south. Higher percentages for the Strongly influenced fens and bog types in the south can most likely be explained by more frequent farming and forestry activities. Higher percentages of Limnogenous fens and Nonmire wetlands in the north may be related to a higher frequency of flooded lake and river margins and floodplains. There are also higher frequencies of String flark fens and String mixed mire types in the north, suggesting a more common occurrence of soligenous flow in sloping drainage ways and higher flooding stages during spring melt, which is associated

General classes of vegetation physiognomy are recognized under Field and bottom cover. In the photo interpretation, the most dominant class was assigned to each polygon. In Table 4 we

reaches of rivers and along lakes there than in interior regions.

mineral soil water influence, and are flat or slightly raised in their centres.

4. Sloping fens →String flark fens→ String mixed mires→Net bog 5. String mixed mires→ Mosaic mixed mires→ Palsa mixed mires

considerably expanded when the treed wetland areas are added.

classification:

1. Flat fens → Sloping fens→ Hill fens 2. Flat fens→Flat or weakly raised bogs

with more frequent string and flark patterns.

**6. Field and bottom cover** 

3. Sloping fens→Sloping bogs

summarize the occurrence of the Field and bottom cover types in the 3229 polygons. The most common types are Sphagnum (33.5%), Low sedge (21.1%), Dwarf shrub (17.6%), Graminoid/dwarf shrub (11.8%), Graminoid- and/or herb (6.8%), and Tall sedges/graminoids (3.3%). Five percent of the polygons were missing a field/bottom layer, probably owing to coverage by shadows, water, peat extraction, or dense cover of shrub or tree vegetation (the field/bottom layer is not registered when the trees have > 50 % crown cover).

In Table 4 we give a short, selected list of some dominant species that characterize the different Field and bottom layer types. Since the Field and bottom types were developed for air photo interpretation, they are necessarily quite broad and reflect main appearance and life form of the vegetation. Finer community types of course can be identified with the more detailed ground survey work that is done in each of the permanent squares. Many other species lists and descriptions of the wetland vegetation in northern Sweden can be found (e.g., Gunnarsson & Löfroth, 2009).


Table 4. Field and bottom cover Types, and percent frequency of the types in the data set (3229 total polygons), and typical abundant species for each type.

Main Ecosystem Characteristics and Distribution of Wetlands

classification one must rely on the dominant life form.

these two types, with a Pearson r value of -0.635, P-Value < 0.0001.

bottom. Flark pool, Bog pool, and Marsh were very infrequent.

**7. Microtopographic series** 

Microtopographic

element

raised bog.

in Boreal and Alpine Landscapes in Northern Sweden Under Climate Change 209

most common Hydrotopographic types. By summarizing the results from a Chi-square analysis, the relation of Field/bottom types to different Hydrotopographic wetland types can be seen (Table 5). The relationship is such that if the observed is higher than the expected by random, there is a positive association, while if it is lower there is a negative association. For example, Graminoid herb was rather strongly and positively associated with Limnogenous fen and Non-mire wetland, while Sphagnum type was strongly positively associated with Flat Fen, and strongly negatively associated with Flat or weakly

From this analysis it is possible to indicate, in a general way, the main kinds of Field and bottom layers associated with each Hydrotopographic mire type. However, it is concluded that one cannot very reliably use Field/bottom layer type to indicate Hydrotopographic type, or vice versa. Instead one must use such defining attributes as, flow patterns, peatland shape, and physiographic location to classify polygons. And for Field / bottom layer

The definition of these types was based upon the microtopographic series - hummock, lawn, carpet, mud-bottom, pool series – introduced by Sjörs (1948) and now used internationally (e.g., Rydin and Jeglum 2006). Seven elements are recognized in the NILS system: Dwarf shrub dominated hummock, lawn, carpet, mud-bottom, flark pool, bog pool, and marsh. (Marsh as described in Rydin and Jeglum, 2006, is very similar to Sumpkarr as described in Allard, 2005). In conducting the photo interpretation, these types together should add up to 100% cover. In the data explored here hummock or lawn were the most common types, and often together totalled 100%. There was a strong inverse correlation between the values for

The area data for Microtopographic series in the polygons are given in Table 6, showing that the most common elements are, in decreasing order: Lawn, Hummock, Carpet, and Mud-

> Hummock asterisk 8423 (1196) 30.51 (2.64) Lawn 11272 (1149) 40.83 (2.47) Carpet 5614 (816) 20.33 (1.72) Mud-bottom 1886 (557) 6.83 (1.61) Flark pool 329 (180) 1.19 (0.56) Bog pool 82 (30) 0.30 (0.10) Marsh 62 (32) 0.23 (0.10)

Proportion of total mire area with standard

errors

Total area (km2) with standard errors

\*Small raised mounds usually built-up by Sphagnum and characterized by dwarf shrubs. Table 6. Estimates of total areas of mires of seven Microtopographic elements, and proportions of the total mire area studied, with standard errors (in parentheses).

The field and bottom cover types can be broadly related to the moisture and water table gradient. Dwarf shrubs, Lichen-dwarf shrubs, and Lichen types are well developed on hummock and mound levels of the microtopographic gradient. Graminoid and/or herb, and Graminoid-dwarf shrub are common types of lawns in fens or bogs. Low sedges are in the carpet phases of fens and bogs. Sphagnum occurs widely across the whole wetland gradient, except for frequently flooded sites and strongly shaded sites (trees, dense shrubs). Reeds and tall sedges/graminoids are on shores and frequently flooded locations, often with higher pH and base-richness. Often there is a Sphagnum bottom layer beneath the dwarf shrub, graminoid, and low sedge types.

There are also general relationships of these Field and bottom cover types with the pH-baserichness gradient. Dwarf shrub and dwarf shrub-lichen tend to have quite low pH-base status and are in bogs or in the bog phase of mixed mires. The sequence tall graminoids, graminoids, and low sedges tend to follow a gradient of high to low pH-base levels. *Sphagnum*-rich sites have a range of species that differentiate both on a moisture-water table gradient and a pHbase richness gradient from rich to poor fen and into ombrogenous bog. Tall reeds and emergents, floating plants, and submerged plants are in water at edges of rivers and lakes, and vary from circumneutral to weakly acid and moderate to low base levels.


**6.1 Relationships of Field and bottom cover to Hydrotopographic types** 

To explore relationships of the field and ground layer vegetation with hydrotopographic types we reduced the data set to the six most common Field and bottom types, and the eight

Table 5. The relations between the Hydrological mire type and associated Field/bottom layers in the dataset, from analyses with Chi-square test.

most common Hydrotopographic types. By summarizing the results from a Chi-square analysis, the relation of Field/bottom types to different Hydrotopographic wetland types can be seen (Table 5). The relationship is such that if the observed is higher than the expected by random, there is a positive association, while if it is lower there is a negative association. For example, Graminoid herb was rather strongly and positively associated with Limnogenous fen and Non-mire wetland, while Sphagnum type was strongly positively associated with Flat Fen, and strongly negatively associated with Flat or weakly raised bog.

From this analysis it is possible to indicate, in a general way, the main kinds of Field and bottom layers associated with each Hydrotopographic mire type. However, it is concluded that one cannot very reliably use Field/bottom layer type to indicate Hydrotopographic type, or vice versa. Instead one must use such defining attributes as, flow patterns, peatland shape, and physiographic location to classify polygons. And for Field / bottom layer classification one must rely on the dominant life form.
