ACTA PHYTOGEOGRAPHICA SUECICA 80 EDIDIT SVENSKA V AXTGEOGRAFISKA SALLSKAPET
Martin Diekmann
Deciduous forest vegetation in Boreo-nemoral Scandinavia
UPPSALA
1994ACTA PHYTOGEOGRAPHICA SUECICA 80 EDIDIT SVENSKA VAXTGEOGRAFISKA SALLSKAPET
Martin Diekmann
Deciduous forest vegetation in Boreo-nemoral Scandinavia
OPULUS PRESS AB UPPSALA 1994
ISBN 9 1-721 0-480-5 (cloth) ISSN 0084-59 14
Editor: Erik Sjogren
Editorial Board:
A. W .H. Darnman, Storrs, CT F.J.A. Daniels, MUnster L. Ericson, Umea
D. Glenn-Lewin, Ames, lA 0. Hamann, Copenhagen H. Sjors, Uppsala H. Trass, Tartu
Technical Editor: Marijke van der Maarel-Versluys
© Martin Diekmann 1 994 Cover illustration: Reno Lottmann
Edidit:
Svenska Vaxtgeografiska Sallskapet Villavagen 14, S-752 36 Uppsala DTP: OPULUS PRESS AB
Printed in Sweden, 1 994 by Eklundshof Grafiska AB, U ppsala
Deciduous forest vegetation in Boreo-nemoral Scandinavia 3
Abstract. Martin Diekrnann. 1 994. Deciduous forest vegetation in Boreo-nemoral Scandinavia - Acta Phytogeogr. Suec. 80, Uppsala. 1 1 2 pp. ISBN 9 1 -721 0-080-X. (9 1 -72 10-480-5).
This study aimed at an investigation of the vegetation ecology of deciduous hardwood forests in the Boreo-nemoral zone of Scan
dinavia. Different community types were described with regard to their species composition and structure, differentiation, geo
graphic distribution, ecological conditions and dynamics, as well as their affinities to community types in other vegetation zones in northern and central Europe.
Field work was conducted in S Sweden and SE Norway. In total 367 releves were made. In each stand, structural character
istics as well as cover-abundance values of all vascular plants and bryophytes were recorded. The releves were analyzed by means of cluster analysis (program TWINSP AN) and ordination (program Correspondence Analysis from the program package CANOCO). Vegetation-environment relationships were analyzed by correlating the explanatory variables with releve scores on the ordination axes.
In 69 stands of two areas (Oland and the mainland of eastern Sweden), tree size was analyzed in order to reveal the population structure of the most important tree species in the main forest types. Successional trends were inferred from the relative densi
ties of species in different DBH (diameter at breast height) classes.
Environmental studies were carried out in 17 stands of mesotrophic and eutrophic forests in the vicinity of Uppsala, including inclination, heat index, maximum temperature, mini
mum temperature, temperature range, evaporation, wind expo
sure, canopy cover, light, dry bulk density, organic matter, moisture, pH(H20), pH(KCl) and nitrogen. Canonical Corre
spondence Analysis (CANOCO program package) and correla
tion analysis were used to reveal the relationships between vegetation and environment.
Four forest types were described, with in total nine commu
nities: (1) oligotrophic oak forests: Quercus petraea-Frangula alnus community, Quercus robur-Betula pendula community;
(2) mesotrophic mixed deciduous forests: Quercus robur-Tilia cordata community, Quercus robur-Euonymus europaeus com
munity, Quercus robur-Fraxinus excelsior community; (3) eutrophic elm-ash forests: Ulmus glabra-Fraxinus excelsior community, Ulmus minor-Fraxinus excelsior community; (4) eutrophic alder-ash forests: Fraxinus excelsior-Prunus padus community, Fraxinus excelsior-Alnus glutinosa community.
Mesotrophic forests represent the most widespread and charac
teristic forest type of the Boreo-nemoral zone, without floristically similar counterparts in the Nemoral zone.
The underlying environmental factors for the vegetational differentiation on the forest type level were a complex-gradient in nutrient status, connected with variation in light conditions, and a moisture gradient. On the community level, mainly geo
graphic-climatic factors were operating, in particular an East
West and a humidity gradient. Anthropogenic factors were less important.
Ellenberg' s central-European indicator values for light, moisture, nitrogen and reaction proved to be useful in character
izing the Scandinavian forests, although some species seem to occur under environmental conditions which differ from those in central Europe. The Boreo-nemoral forests are as species-rich as the Nemoral forests; the absence of several southerly distrib
uted taxa is 'compensated' for by comparatively higher frequen
cies of other species. Community types on soils with an interme
diate to moderately high fertility have the highest species rich
ness.
Most stands have been influenced strongly by human activ
ity and are compositionally unstable. In eutrophic elm-ash for
ests, Quercus spp. and Fraxinus excelsior were better repre
sented with larger size classes, whereas Acer platanoides and Ulmus spp. had higher frequencies in smaller size classes. It is suggested that this pattern can be interpreted in terms of a future replacement series: in a succession following termination of human impact, Quercus and Fraxinus will be partly replaced by Ulmus and Acer. The population structure in mesotrophic for
ests revealed that Quercus will decrease, whereas Ulmus and Fraxinus and, in particular, Acer and Tilia, will increase. In oligotrophic forests, however, Quercus will maintain a domi
nant position, while Picea may increase in importance. The tree species dynamics will also lead to increasing canopy closure and, subsequently, compositional changes in the other vegeta
tion layers.
The environmental studies in the selected forest stands re
vealed that pH (together with nutrients) and light were the primary factors related to vegetational variation in the field and bottom layers. Other properties of the soil and the climate of the forest as well as physiographic factors, were of minor signifi
cance. Mesotrophic forests had significantly lower soil reaction, higher light intensities and higher maximum temperatures than eutrophic forests. They also tended to have a lower soil mois
ture, a wider temperature range, lower minimum temperatures and a higher evaporation.
As a whole, deciduous hardwood forests in Boreo-nemoral Scandinavia have become rare and threatened, particularly the eutrophic forests. Existing forests, especially those with a long, uninterrupted history, should therefore be maintained and pro
tected.
Martin Diekmann, Department of Ecological Botany, Uppsala University, Villaviigen 14, S-752 36 Uppsala, Sweden.
Acta Phytogeogr. Suec. 80
1 . 1 Aims
1.2 Nature conservation aspects
1.3 Deciduous hardwood forests in northern Europe: a brief literature review 1.4 Vegetation and forest history
2 Study area
2. 1 The Boreo-nemoral zone and its delimitation 2.2 Geology and Soils
2.3 Climate
3 Analysis of vegetation and vegetation-environment relations 3 .1 General
3.2 Sampling procedure 3.3 Data treatment 3.4 Tables
3 .5 Nomenclature of forest communities
3.6 Nomenclature of species and comments on difficult taxa 4 Forest communities
4. 1 Introduction: Survey of clusters 4.2 Oligotrophic oak forests
4.3 Mesotrophic mixed deciduous forests 4.4 Eutrophic elm-ash forests
4.5 Eutrophic alder-ash forests 4.6 Syntaxonomy
5 The population structure of trees and forest dynamics 5 . 1 Introduction
5.2 Methods 5.3 Results 5.4 Discussion
6 Environmental studies 6. 1 Introduction
6.2 Soil analysis and climatic measurements 6.3 Results
6.4 Discussion
7 General discussion and conclusions 7. 1 Species
7.2 Communities and environment
7.3 Geographic distribution of communities and the Boreo-nemoral zone 7.4 Structure and dynamics
7.5 Implications for nature conservation 8 Acknowledgements
9 References
5 5 6 8
12 12 1 3 1 5
1 8 1 8 1 8 1 8 20 20 20
22 22 23 32 48 55 64
69 69 70 7 1 74
79 79 80 82 89 95 95 98 10 1 102 103 104
105
1 Introduction
1.1 Aims
This study aims at a floristic and ecological description of deciduous forest communities in the Boreo-nemoral zone of Scandinavia. Field investigations have mainly been carried out in Sweden, partly for practical reasons but also because the deciduous forests in Sweden are less well known than their Norwegian counterparts. The study deals with forests of a type called 'adellovskog' in Swedish (broad-leaved hardwood forest), composed of A cer platanoides, Fraxinus excelsior, Quercus spp., Tilia cordata and Ulmus spp. The term 'adel' ('noble') refers to the species' relatively high demands for nutrients and temperature, as well as their greater economic importance and higher status, compared with 'trivial' species such as Betula spp. and Populus tremula. Also Fagus sylvatica and Carpinus betulus are often treated as 'noble' tree species. However, forests composed of these two species have their main distribution in the Nemoral zone and form only minor stands within the study area. Beech forests will therefore not be dealt with, and particularly since they have been described in detail by Lindquist ( 1 93 1, 1 932), Lindgren (e.g. 1 970, 1 975) and recently by T. 0kland ( 1 988). Neither will young successional forests with e.g.
Betula spp. be considered, or wet swamp forests with Alnus glutinosa, which have a different species composi
tion. The study thus comprises forests which phytosocio
logically can be arranged in the class Querco-Fagetea.
Since the study only deals with fairly natural forests with a more or less closed canopy, the term 'forest' will be used throughout instead of 'woodland', which is a term with a wider meaning, including both forests and parks, wooded meadows, wooded pastures, etc.
More specifically, the study foremost aims at a classi
fication of the deciduous forest communities in Scandina
via. They will be described with respect to their species composition, structure, geographic distribution, differen
tiation and ecological conditions. Ordination in connec
tion with correlation analysis will be used to reveal the importance of physiographic and macroclimatic param
eters for community differentiation. Species indicator values will be applied to evaluate the importance of edaphic and climatic factors for species responses. Emphasis is also placed on the successional trends of the main com
munities, derived from the population structure of tree species. Measurements of soil and climatic parameters, carried out in a selection of forest stands belonging to different forest communities, will be related to species
composition in order to reveal the primary causes for community differentiation. The Boreo-nemoral commu
nities will also be discussed with respect to their central and western European counterparts.
1.2 Nature conservation aspects
Climatic deterioration as well as millennia of cultivation and exploitation have diminished the area of deciduous forest in Scandinavia to a small fraction of its prehistoric size. The majority of sites suitable for deciduous forest are now covered by arable land and spruce plantations. The remaining stands have been influenced and changed by land use practises such as logging, pollarding and grazing, earlier also thinning for haymaking in wooded meadows.
Deciduous forests seem to be common in many parts of southern and central Sweden. However, this impression is delusive, since they usually are confined to open areas or fringes around cultivated land, lake-shores and hillsides, and in the vicinity of old roads and villages. Only occa
sionally, stands of deciduous forest are surrounded by an extensive area of coniferous forest.
The changes in agriculture and forestry (planting of conifers and decreasing importance of grazing in forests and former wooded meadows) since the end of the last century have resulted in a slight increase of the deciduous forest area. Many of the stands originating from older times have become mature and economically attractive for selective logging or complete deforestation, the more so as the value of timber from broad-leaved tree species has increased. This has only rarely, as yet, led to replant
ing with broad-leaved trees. During the course of the study, the author has witnessed the destruction of many stands. In the Mittlandsskogen area on Gland, a large part of the mature forest has been completely or partly cut down during only five years.
Motives for the preservation of deciduous forests are manifold, e.g. their high floristic and faunistic diversity, as compared with conifer forests (especially plantations), the occurrence of rare plant and animal species bound to deciduous forests, their importance for recreation and, in some cases, as monuments of former land use and cultural history.
Acta Phytogeogr. Suec. 80
1.3 Deciduous hardwood forests in north
em Europe: a brief literature review
Among the first to call attention to the deciduous forest vegetation in Sweden was Linnaeus (1745) who, in his book on the journey to bland and Gotland in 17 41, wrote as follows:
" ... The road passed through the most beautiful groves one could ever see, which for beauty surpassed all other places in Sweden and competed with all in Europe; they consisted of linden, hazel and oak, with the ground smooth and green, without rocks or moss . .. " ( transl. by Asberg &
Steam 1973).
General description
Until the beginning of the 20th century, botanical interest was mainly focused on plant geography and systematics.
The increasing importance of industrial forestry led to a growing interest in forest research, which, however, mainly aimed at a high productivity of certain favoured tree species, i.e. Picea abies and Pinus sylvestris. One of the first detailed ecological descriptions of deciduous vegeta
tion in the Boreo-nemoral zone was the study of wooded meadows on Aland by Palmgren (1915-1917). A mono
graph on the flora and vegetation in V pp land by Almquist (1929) contains a classification of deciduous forests that is mainly based on structural characteristics and domi
nances of species. During the following decades, several books and papers on deciduous forests were published.
However, a few attempts were made to give a comprehen
sive account of these forests, e.g. by Selander (1955). At least for Boreo-nemoral Sweden, deciduous forests were rarely dealt with in detail, in contrast to other vegetation types of a much wider distribution, such as mires.
Deciduous forests are of a limited extension, but often very striking in the coniferous or cultivated landscape. In many publications, single forest stands have been de
scribed, particularly from some provinces in southeastern Sweden, for example bland (e.g. Du Rietz 1917; Sterner 1926; 0. Johansson 1982 and K. Larsson 1982) and Sodermanland (e.g. Halden 1950; Ryberg 1956, 1971). In Norway, Skogen (1971) made an ecological and plant geographic study of the world's northernmost oak forest.
In many areas, inventories of valuable forest stands in connection with nature conservation projects were carried out, e.g. by Sjogren et al. (1974) and Ekstam (1979) on bland, and Korsmo (e.g. 1974) in Norway. These publi
cations often contain forest classification systems for mapping purposes. With regard to nature conservation aspects, the flora and vegetation of deciduous forests have been treated by e.g. Bergendorff et al. (1979), IngelOg (1981), Anon. (1982), Almgren et al. (1984) and IngelOg et al. (1984).
Several studies cover fairly large areas, treating the (forest) vegetation of a parish or larger district. In Swe
den, Ivarsson (1962) and Wallin (1973) described the deciduous vegetation in parts of BohusHi.n and Vastergot
land, respectively. Both give information on community differentiation, forest history and successional aspects.
Sjogren (1961, 1964) described the different forest types and bryophyte communities of deciduous forests on bland.
Olsson (1974, 1975) studied the south Swedish sand vegetation, including oak forests on acid soils. Here, the classification was accompanied by extensive soil analyses.
In Finland, literature on deciduous forest in the Boreo
nemoral zone is almost lacking, probably due to the restricted extension of the zone and the rareness of this vegetation type. An exception is the monograph on eutrophic deciduous woods in the SW archipelago by Hinneri (1972). However, many descriptions of wooded meadows exist, e.g. by Palmgren ( 1915-1917), Cedercreutz (1927) and Hregstrom (1983). Several publications from Finland deal with forest communities of more northern but adjacent areas belonging to the southern Boreal zone, including mixed thermophilous forests with Tilia cordata andAlnus-rich forests, e.g. Tapio (1953), Koponen (1967) and Makirinta (1968). Compared with Sweden, the de
ciduous forest vegetation in Norway is much better known.
Especially since the beginning of the seventies, many studies from different parts of Norway have been pub
lished, often using the Braun-Blanquet approach, e.g. A.
Bj�rnstad (1971), Aune (1973), Fremstad (1979) and 0vstedal (1985). Kielland-Lund (1981) made a compre
hensive study of the forest communities in SE Norway.
For each community, information is given on the syntaxo
nomy, general species composition, differentiation, eco
logical conditions, dynamics and chorological aspects.
Other publications treat the widely distributed, Alnus incana-rich forests on river banks, e.g. Fremstad &
0vstedal (1978) and Klokk (1980, 1982), having counter
parts in Sweden and Finland mainly in the Boreal zone, but to some extent also in the western parts of the Boreo
nemoral zone.
Classification systems and nomenclature of forest com
munities
Vegetation science in Fennoscandia has not followed one particular approach, either with respect to methodology, or in syntaxonomical classification (Trass & Malmer 1973;
Malmer 1974). Swedish phytosociology, and the 'Uppsala School' in particular, is more a tradition than a fixed School. There is no generally accepted system of hierar
chically structured vegetation units, as exists in central Europe where the Braun-Blanquet approach is used by practically all vegetation scientists. Deciduous forest com
munities in Sweden have been described using different methods and nomenclatural systems, e.g. by Julin (1948),
Deciduous forest vegetation in Boreo-nemoral Scandinavia 7
Ivarsson (1962), Sjogren (1964), Wallin (1973) andKlotzli (1975a). Several Norwegian authors have followed or adapted the Braun-Blanquet approach, e.g. Bj�rnstad (1971) and Kielland-Lund (1981).
Only a few attempts have been made to work out a classification for the whole of Scandinavia or the Boreo
nemoral zone, respectively. Kielland-Lund (1971, 1973) presented the first schemes in connection with the Inter
national Biological Programme. Each community is char
acterized by a short general description and a species list, including dominant and differential species. Vevle (1983) published a preliminary survey of higher syntaxa for Norway, and a synopsis of Norwegian forest communities was recently given by Kielland-Lund (1994). Classifica
tions were also presented by Bergendorff et al. (1979) and Anon. (1984). As the most important criterion for the distinction of forest types, the dominance of tree species has been used. Both authors give valuable lists of syno
nyms of communities from the North European literature.
None of the systems mentioned is based on a comprehen
sive tabular comparison of releves from different regions.
Klotzli (1975a) proposed a classification derived from such a comparison, but his system is based on a rather small number of releves. Comparisons of central Euro
pean and Scandinavian forest communities have mainly dealt with conifer forests (Matuszkiewicz 1962; Kielland
Lund 1967 and Neuhausl 1969). However, North Euro
pean wet alder forests have been included in a monograph on Alnus incana-rich communities in Europe by Schwabe (1985).
Multivariate analysis
Modem multivariate techniques, such as cluster analysis and ordination, have only exceptionally been applied to Fennoscandian deciduous forest vegetation. One of the first studies was performed by Ammar (1978) on the vegetation (and its underlying environmental factors) of shore ridges on bland. Rlihling & Tyler (1986) and Brunet (e.g. 1991) use cluster analysis for the description and comparison of deciduous forest communities in southern
most parts of Sweden. In Finland, Makirinta (1990) and Tonteri et al. (1990) have used both classification and ordination techniques for their analyses of forests in the southern part of the country. Their studies include a comparison of modem techniques with the traditional Finnish approach, i.e. Cajander's forest site type classifi
cation. Cluster analysis and ordination were also applied by Heikkinen (1991) in order to analyse the esker vegeta
tion in an area in S Finland. His study shows the impor
tance of physiographic and microclimatic factors for the differentiation of forest communities. In Norway, fern
rich vegetation has been analyzed with the aid of multi
variate methods by Odland et al. ( 1990) and Odland (1991, 1992). Thus, most studies have concentrated on
either particular forest types or limited areas. A compre
hensive treatment of deciduous forests in Sweden, using multivariate methods, has not been carried out so far.
Productivity
Apart from classification and description of communities and their relationship with environmental factors, the productivity of deciduous forest species has been ad
dressed. In the framework of the International Biological Programme, Hyttebom (1975) and H. Persson (1975) investigated the above-ground woody production and field
layer and below-ground production, respectively, of a deciduous woodland in the province of Uppland. The latter study also presents many phenological data on the seasonal development of herbs and grasses.
Population biology and forest dynamics
The population ecology of deciduous forest species has not gained much attention up to now. An exception is formed by the thorough demographic study of the peren
nial herbs Hepatica nobilis and Sanicula europaea by Inghe & Tamm ( 1985), where the authors demonstrate the importance of weather conditions, particularly summer drought, for flowering and mortality of the investigated species. Among other studied herbs are Primula veris (Tamm 1972), Fragaria moschata (Ryberg 1986) and Lathyrus vemus (Ehrlen 1992). Brunet (1994) studied the demography of the woodland grasses F estuca altissima, Hordelymus europaeus, Bromus ramosus and B. benekenii.
With respect to deciduous tree species, Fraxinus excelsior was treated by Hulden (1941) and Tapper (1992), and Quercus robur by C. Andersson (1994). Much more is known about the population biology, as well as popula
tion genetics, of conifers such as Picea abies and Pinus sylvestris.
With respect to forest dynamics (both natural and anthropogenic), attention has foremost been given to three different subjects: firstly, the acidification of forest soils and its impact on species composition; secondly, the succession of deciduous forests, particularly concerning woody species; thirdly, the impacts of climatic change on forest ecosystems.
The negative consequences of soil acidification due to acid precipitation have first been described for central Europe, but have recently become a major topic in Swe
den as well. Research has concentrated on S and SW Sweden where the acidification is most pronounced due to a comparatively high amount of rainfall. The first results of these long-term changes of soil acidity and its impact on soil nutrient contents and species composition have been presented by Hallbacken & Tamm (1986), Falkengren-Grerup (e.g. 1986, 1987), Tamm & Hallbacken (1988) and O.N. Bj�rnstad (1991). Norden (1992)
Acta Phytogeogr. Suec. 80
compared the impact of different deciduous forest species on soil acidification and element fluxes in Skane.
As most deciduous forests in Scandinavia are in a highly dynamic state, successional trends and the poten
tial 'climax' vegetation have been addressed by many authors (e.g. Lindquist 1938 and Ivarsson 1962). How
ever, the discussions are seldom based on field observa
tions. Ryberg (1971) studied the diameter and age distri
bution of tree species with regard to forest history and succession. Through repeated analysis of permanent plots established between 1925 and 1935, long-term changes in the composition of the tree layer could be demonstrated for Dalby Soderskog in Skane, situated in the Nemoral zone (Lindquist 1938; Malmer et al. 1978; S. Persson 1980). This forest stand has been heavily disturbed during the past and is in a successional state towards a more natural species composition. In a study of Vardsatra Na
ture Reserve near Uppsala, Hytteborn (1986) compared densities and basal areas of tree species between 1912 (measured by R. Sernander) and 1985. The importance of forest history and former land use for the structure and species composition of a deciduous woodland on bland were emphasized by Ekstam & Sjogren (1973). Julin ( 1948) treated the ecology of wooded meadows after abandonment. The roles of particular species in forest succession were discussed for Alnus glutinosa by Fremstad ( 1983) and for both Fraxinus excelsior andAlnus glutinosa by Tapper (1992). Andersson (1994) studied the patterns of seedling emergence and early survival of Acer plata
noides, Quercus robur and Tilia cordata at a site in a transitional state between open grassland and deciduous forest. However, general trends of deciduous forest suc
cession in the Boreo-nemoral zone, as well as trends of particular species or forest communities, are still insuffi
ciently known. Due to the great but variable longevity of most tree species, predicting the future from present trends is an uncertain procedure.
Successional trends as described above may be super
imposed by long-term vegetation dynamics due to cli
matic changes, induced by an increasing carbon dioxide level and subsequent global warming. Modelling of the responses of Nordic forests to these climatic changes indicates that many tree species eventually may expand (Fagus sylvatica) or reduce (Picea abies) their distribu
tion areas, and that their competitive· relations may change (Sykes & Prentice 1993). As a consequence, the Boreo
nemoral zone in Sweden would 'move' northwards over a considerable distance.
1.4 Vegetation and forest history
All forests and forest sites have in some way been influ
enced and changed by the activity of man. In order to understand and interpret the structure and species compo-
sition of these forests, it is necessary to elaborate the forest history of the study area up to the present. Much of the information given in this chapter is based on the comprehensive studies of the vegetation and forest his
tory by Fries (1965) and Huntley (1988), and of the cultural history and former land use by Sjobeck (1931, 1933), Selander (1955) and Ekstam et al. (1984).
Vegetation history
After the last glaciation, trees started to re-invade north
ern Europe from the south during the early Post-glacial time (since about 8000 B.C.). Among the first woody species to appear were Betula spp., Populus tremula, Pinus sylvestris and Corylus avellana. Other thermophilous species than Corylus, such as Alnus glutinosa, Ulmus glabra, Fraxinus excelsior, Quercus robur and Tilia cordata, first immigrated at the beginning of the Late Boreal (ea. 6800 B.C.) due to a climatic amelioration.
During the Atlantic time (Older Stone Age, culmination about 4000 B.C.), a period of climatic optimum, rich deciduous forests covered extensive areas of Sweden.
These so-called mixed oak forests (Quercetum mixtum) occurred further to the north at that time, compared with presentday distribution. Evidence of this is found in pol
len spectra, fossil nuts of Corylus avellana, and isolated present occurrences of Corylus, Tilia and Ulmus glabra ssp. montana in Norrland. In southern Sweden, less de
manding trees such as Betula and Pinus were probably restricted to either dry and shallow or very wet soils. The early Sub-boreal time (since about 3000 B.C.) is charac
terized by a deterioration of the climate and an increasing human impact on the vegetation. As a result, the area of deciduous forest decreased again in the whole of Sweden.
At the beginning of the Sub-atlantic time (transition from Bronze to Iron Age, ea. 500 B.C.), the climate deterio
rated once more towards cooler and moister conditions and caused a further decrease of the Quercetum mixtum.
The present northern distribution of, e.g. Tilia cordata, reflects a relict status from the mid-Holocene. Besides, new forest tree species immigrated to northern Europe, Pie ea abies from the northeast and F agus sylvatica anJi Carpinus betulus from the south. Pinus and Picea gradu
ally became the most important constituents of North European forests, apart from Fagus and Quercus in the southwest.
Cultural history
During pre-agricultural time, human impact on the North European forest landscape was probably negligible. Set
tlements of the Younger Stone Age and Bronze Age were concentrated to areas in S Sweden that were naturally rich in deciduous forests. Farming played a less important role during these times than cattle-breeding, and the people
Deciduous forest vegetation in Boreo-nemoral Scandinavia 9
Fig. 1 . Wooded meadow , spring aspect. Scattered trees and shrubs alternate with open areas. Conspicuous are a large individual of Quercus robur and several Corylus avellana bushes. In the foreground, a traditional fire place can be seen, where the litter and small branches of the previous year have been burnt. Gotland, May
1989. Photo M. Diekmann.
burnt the woodland extensively in order to create clear
ings for pastures and later hay-meadows. Losses of nutri
ents and minerals due to constant grazing and burning caused a significant deterioration of the soil, except on the most fertile morainic tills and in wetlands which were naturally manured by periodical flooding. During the Middle Ages and later periods, the forest area decreased continuously, caused by an increasing population and its high demands for cultivated land, timber and fire wood. In parts of northern Europe, forests had vanished completely already before Modem times, for example in the west Norwegian coastland (Behre 1988). When the climate turned cooler and wetter at the beginning of the Iron Age, supply of winter fodder for cattle became increasingly difficult. This period gave rise to the partitioning of farm
land into different structural and functional units, and the creationo,of hay-meadows.
The former agrarian landscape was divided into 'inago
mark', the infield, infenced land, and 'utagomark' , the outfield, unfenced land. The latter comprised nutrient
poor heath and woodland and was used as common pas
ture. The infield land was made up of the rich soils with arable land and hay-meadows. The meadows provided winter fodder for the cattle, which in turn provided ma
nure for the fields. The harvest of grasses and herbs was difficult because tools for cutting, such as sickles and scythes, first became available during the course of the Iron Age. Instead, the farmers cut and broke off branches of trees of deciduous species, in order to dry the leaves.
Thus, probably about 1500 years ago, the first wooded meadows were created. They represented a mosaic of different vegetation types: small groves or groups of trees
alternated with open meadow areas without trees (Fig. 1).
The maintenance of wooded meadows was very time- and energy-demanding. Work consisted in removing and burn
ing litter, cutting of branches and twigs, collecting and drying their leaves, mowing of the grass and herbs and removing invading trees and shrubs. Following the mow
ing, cattle were usually allowed to graze during late summer and autumn. All this implied a constant loss of plant material which was not compensated for by fertiliz
ing. In the beginning, however, wooded meadows did not suffer so much from an impoverishment of the soil, be
cause trees with their extensive root systems permanently transferred nutrients from deeper soil layers to the ground.
They were bound to old settled areas and therefore con
centrated to S Sweden and the provinces along the Baltic Sea, from northern Skane and Blekinge up to Uppland.
However, wooded meadows could also be found in cen
tral and northern Sweden, for example in Dalarna, where trivial broad-leaved trees (e.g. Betula spp., Populus tremula, Sorbus aucuparia, Salix spp. and Alnus incana) replaced the more demanding, thermophilous species (Sjors 1954 ). The extension of wooded meadows culminated during the 18th century. From this time on, but particu
larly during the 19th century, changes in agricultural economy and implements entailed a major change in the agrarian and forest landscape.
In general, the relocation into larger blocks of farm
land, resulting from three land reform events ('storskifte', 'enskifte' and 'lagaskifte'), led to a more effective agri
culture. The former outfield land was now private-owned and subjected to exploitation, especially from the copper
and ironworks which required great quantities of char-
Acta Phytogeogr. Suec. 80
coal. Due to milder forest laws and an increasing demand for timber, the total forest area decreased rapidly in many provinces during the 19th century, e.g. on bland (Daniels
son 1918). Parts of BohusHin were almost completely deforested in 1850, as a result of the high demand for timber, smallwood and fuel for agriculture, and for mak
ing herring oil (Fries 1958). Also in Halland, the forest area reached its minimum in the middle of the 19th century (Malmstrom 1939). At the same time, grazing increased considerably, and many closed forests were turned into wooded pastures. In this stage, more or less natural, primary forests had become quite rare and were restricted to special sites. Forests on steep slopes, e.g. at edges of the plateau hills, were less accessible to grazing animals and unsuitable for cultivation, and probably only subjected to selective logging. Woodland belonging to big estates or the Crown was sometimes preserved for hunting reasons.
After the introduction of more effective ploughs and artificial fertilizers, arable lands, including leys and ferti
lized meadows, extended their areas at the expense of wooded meadows, which suffered from deterioration of soil and were either cut down and turned into arable land and pastures, or allowed to regrow into forest. Many of the present deciduous forests originate from former wooded meadows, often through an intervening stage of wooded pastures. The rapid decrease of the wooded meadow area can be illustrated by some figures from Gotland, where about 32 000 ha in the year 1900 had become reduced to only 284 ha in 1983.
At the turn of the century, grazing in forests and wooded pastures decreased due to a more intensified agriculture (Anon. 1982). The need for timber decreased too, as other materials and energy sources became more important. Many pastures, wooded pastures and meadows turned into forest again, and the deciduous forest ex
tended its area considerably. This extension has been most pronounced in central bland which was almost devoid of trees in the middle of the 18th century (Daniels
son 1918), but now carries the largest continuous area of deciduous forest in S Sweden, Mittlandsskogen. How
ever, large areas suitable for deciduous forest were planted with conifers, mainly spruce. This planting activity started already in the 19th century (e.g. Malmstrom 1939; Fries 1958).
Tree species composition of wooded meadows As many deciduous forests have arisen from wooded meadows, the management of the latter must be given some consideration. Certain trees and shrubs of high 'quality' and usefulness were selected and favoured, oth
ers repressed or eliminated. As a consequence, the natural frequencies of important tree species of the Boreo-nemoral zone were altered.
Most important for the supply of winter fodder in the wooded meadow was Fraxinus excelsior. Its ability to resprout after repeated cutting of branches and twigs made it very popular, not only in northern Europe, but also in central Europe (Pettersson 1958; Ellenberg 1986). Pol
larded ash trees can still be seen as remnants in forests and meadows in many parts of the continent (Behre 1988).
Hence, the species is nowadays often found on the former infield land in the neighbourhood of old villages and farms, especially in Blekinge and Smclland. On the other hand, Fraxinus was greatly reduced on the former outfield land due to its sensitivity to grazing.
Ulmus glabra was not particularly favoured as a wooded meadow tree. However, elm was often planted close to large estates and castles. According to Lindquist (1932), many of the planted trees belonged to introduced provenances of southern origin. Ulmus minor could be found in wooded meadows on bland and Gotland, and its ability to resprout as root suckers made it a fast colonizer on abandoned pastures (Pettersson 1958).
Another selected tree species was Tilia cordata. It provided leaf fodder, bast fibers for ropes, and valuable timber which was easy to process (Anon. 1982). Linden played an important role in areas of bee-keeping and bast production. However, on Gotland it was repressed through human influence (Pettersson 1958). Besides, it generally suffered from grazing in the former outfield land, but often survived among large boulders.
Quercus robur and Q. petraea were the tree species that were considered the most desirable, both in northern and central Europe (Ellenberg 1986; Behre 1988). They were usually the only species to be allowed to grow tall.
Oak hardly provided any leaf fodder, but raising of pigs was very dependent on a good supply of acorns. For this purpose, also beech (Fagus sylvatica) was allowed to form large stands on the estates in the south. Besides, both oak and beech gave valuable timber, and oak bark could be used for tannery. Land owners kept oak forests for the purpose of hunting, and the Crown preserved and even planted oak to ensure a supply of timber for warships.
Already in medieval times (14th century), a law was enacted against the cutting of oak and hazel. A decree from 1569 prohibited the cutting of oak, hazel and apple on bland (Danielsson 1918). Due to its low demand for nutrients, Quercus probably also had a good chance of survival in pastures of the outfield land with its deterio
rated soils. However, grazing must have prevented regen
eration of oak.
Even other 'fruit-bearing' species such as Malus sylvestris and Corylus avellana (in the south locally also Fagus sylvatica) were selected. Hazel gave valuable wood and nutrient-rich leaves for winter fodder.
Acer platanoides was neither selected for in wooded meadows, nor needed as timber or fire wood. On Gotland, human activity eliminated it from major parts of the island
Deciduous forest vegetation in Boreo-nemoral Scandinavia 11
(Pettersson 1958). However, it was popular as an orna
mental tree close to farms and villages.
Conifers, especially Picea, were usually considered undesirable in the wooded meadows and were removed (Hregstrom 1983).
Summarizing these effects of management, Quercus robur, Fraxinus excelsior and Corylus avellana (locally also Tilia cordata) were the most favoured species. In contrast, Betula spp., Ulmus spp. and Acer platanoides were not favoured, and conifer trees were usually disfavoured. This selection of species has had a great impact on the composition of deciduous forests until recent times (Ekstam & Sjogren 1973).
Effects of grazing
Forests and wooded pastures of the former outfield land have been grazed by cattle, pigs, horses, goats and sheep for many centuries. After the land reforms in the first half of the 19th century, also many parts of the infield land suffered from a hard grazing pressure. The effects of this grazing have been numerous, depending on its intensity, duration and character.
Foraging of acorns by pigs had a largely positive impact on oak and beech forest. Trampling might have caused some damage, but greatly facilitated regeneration of trees due to uprooting of the soil. Grazing, e.g. by cattle, had a strong direct impact on the vegetation. Even a moderate grazing intensity can prevent regeneration of trees (Steen 1958), as shown for, e.g. Quercus robur (Brenner 1921) and Fraxinus excelsior (Hulden 1941). As a consequence, in the long run, and with some logging, the tree layer becomes more open, and the closed forest turns into a wooded pasture. Eventually, trees may totally dis
appear, as on the island Stora Karlso close to Gotland after centuries of intensive sheep grazing (Froman 1946). Shrub
species do not suffer in the same way, as several species are thorny or unpalatable, e.g. Rosa spp., Crataegus spp., Prunus spinosa and Juniperus communis. A high fre
quency of these species in deciduous forests may indicate former grazing.
A moderate grazing intensity often results in a higher species diversity due to a more open, lighter forest struc
ture and the constant removal of large, competitive species.
Releves in a grazed, open oak-ash forest on bland had species numbers up to 84 species/200 m2, including mosses (Diekmann 1988). On the other hand, intensive grazing pressure causes a floristic impoverishment. In general, the number of herbs decreases, whereas grasses and certain unpalatable herbs increase. With respect to structure, the field layer becomes lower and denser. As grazing pressure increases, animals behave less selectively.
In the long run, continuous grazing necessarily results in a deterioration of the soil, even if droppings and urine, in limited spots, may cause a much higher fertility. It may also change the morphology and density of the soil (Steen 1958).
Many centuries of intensive land use in the form of logging, mowing and grazing have altered forest struc
ture, species composition and soil conditions. Apart from the effects described above, the geographical isolation and historical discontinuity of present deciduous forests are important factors in terms of species richness. Studies in several parts of Europe have shown that continuous, ancient forests have a higher species richness than discon
tinuous, recent forests, and that isolation causes a de
crease in species (e.g. Peterken & Game 1984; Dzwonko
& Loster 1990; Brunet 1993). Thus, compared with natu
ral woodland of pre-historical time, many deciduous for
ests in northern Europe may have suffered from a floristic impoverishment of native forest species.
Acta Phytogeogr. Suec. 80
2. 1 The Boreo-nemoral zone and its de
limitation
In the same way as any extensive area of the earth, northern Europe can be divided into different vegetation zones. These are primarily defined according to prevail
ing types of vegetation. The criteria for the division and delimitation of such zones fall into three categories, bioclimatic, edaphic-topographic and botanical (Ahti et al. 1968). Macroclimate is of crucial importance as an underlying factor. The climatic approach normally uses thermal values, e.g. mean temperature, length of growing season, and temperature sums (Tuhkanen 1984). Edaphic
topographic divisions are based on soil types and/or the general topography of an area. A botanical division makes use of distribution patterns of single species or the occur
rence and dominance of certain plant communities. Apart from the criteria mentioned above, a division of zones can also be based on ecological criteria such as phenological data or productivity. Besides, even cultural and agricul
tural features are usually specific for certain vegetation zones. Attempts using different criteria often come to similar results. A vegetation zone can be further divided into vegetation sections based on characteristic features caused by the oceanity/continentality of the climate (Ahti et al. 1968). Vegetation belts designate altitudinally sepa
rated regions.
Several systems of vegetation zones in northern Eu
rope have been proposed by, e.g. Zoller (1956), Hustich (1960), Sjors (1963) and Ahti et al. (1968). A detailed map of the vegetation regions in Norway was published by Moen (1987). The vegetation zones of the Nordic countries according to Sjors are shown in Fig. 2. Rich deciduous forests with tree species of the genera Quercus, Acer, Tilia, Ulmus and Fraxinus are confined to the south
em- and westernmost parts of northern Europe. The Nemoral zone (Temperate zone according to Ahti et al.
1968) embraces Skane, the western parts of Halland and BohusHin, SE Blekinge and S bland as well as parts of the south coast of Norway. The forests of this zone are almost exclusively deciduous. Beech (Fagus sylvatica) is the dominating tree species on mesic soils throughout the zone except on bland and in the coastal parts of northern BohusHi.n as well as most of the Norwegian part. How
ever, on very dry or wet soils other broad-leaved species play a more prominent role. Norway spruce (Picea abies) originally was absent from the Nemoral zone, but can nowadays be found almost everywhere due to planting.
A much larger area of northern Europe, compared with the Nemoral zone, belongs to the Boreo-nemoral zone (Hemiboreal zone according to Ahti et al. 1968). The boundary between these two zones is usually defined as the natural southwestern distribution limit of Pie ea abies.
Because of the intensive planting of this species, as men
tioned previously, this limit cannot be accurately recon
structed any longer (Sjors 1965). Due to the scarcity of Picea on southernmost bland this part of the island is usually included in the Nemoral zone. The higher el
evated plateau in the province of Smaland ( 'sydsvenska hoglandet') is treated as a southern outlier of the Boreal
Fig. 2. Vegetation zones in northern Europe. 1 . Arctic zone; 2.
Alpine belt; 3-6. Boreal zone (3. Sub-alpine birch woodland belt; 4. Sub-arctic and Boreo-montane sub-zone; 5. Main Boreal zone; 6. Southern Boreal sub-zone); 7. Boreo-nemoral zone; 8.
Nemoral zone; 9. Western broad-leaved and pine forest region (North Atlantic pine-birch woodland and heath region). Accord
ing to Sjors (1963), from Anon. (1977), revised.
Deciduous forest vegetation in Boreo-nemoral Scandinavia 13
zone by Ahti et al. (1968).
The Boreo-nemoral zone is dominated by conifer for
ests consisting of Picea abies and Pinus sylvestris. De
ciduous forests depend on favourable climatic and/or edaphic conditions and have a more restricted distribu
tion. Nowadays, deciduous forests are more or less un
common in most parts of the Boreo-nemoral zone because man has turned them into agricultural land. Most Nemoral tree species are still represented, namely Quercus robur, Ulmus glabra, Fraxinus excelsior, Acer platanoides and Tilia cordata. Only Acer campestre and Tilia platyphyllos are absent, and Carpinus betulus is restricted to a narrow fringe close to the Nemoral zone. Fagus sylvatica forms stands in the southern and southwestern parts, with scat
tered occurrences further to the north. Quercus petraea has a western distribution, whereas Ulmus minor only occurs on the Baltic islands of Oland and Gotland. Ulmus laevis is confined to Oland and southern Finland.
The northern boundary of the Boreo-nemoral zone is usually defined as the distribution limit of Que reus robur.
It is well known as the 'limes norrlandicus'. In Sweden, this border line can be drawn from just north of the lower course of the river DaHilven, through central Vastmanland and Narke and north of Lake Vanern to the Norwegian border. The fairly abrupt change in the floristic and vegetational composition corresponds to a similar abrupt change of both topographic, edaphic and climatic condi
tions and, subsequently, of land use (Fries 1948; Fransson 1965). In the Boreal zone, many Nemoral tree species are totally absent or restricted to a narrow zone in the south.
The reason for this has often been considered to be the length of the growing season (or temperature sum) which decreases with increasing latitude. Generally, this favours the 'evergreen-strategy' of conifers (Schroder 1983).
Nemoral outposts in the form of small stands of ash, linden, maple and hazel are often bound to south-facing slopes with a favourable local climate and soil chemistry.
The western race of elm (Ulmus glabra ssp. montana) extends almost to the polar circle in Sweden and even farther in Norway.
Apart from Sweden, the coastal region of southemmost Finland, including Aland, belongs to the Boreo-nemoral zone. Even here, the distribution limit of Quercus robur forms the border line to the adjacent zone, again due to a fairly abrupt change of topographic and climatic condi
tions (Jalas 1957). On the other side of the Baltic sea, the Boreo-nemoral zone includes the Baltic states and adja
cent regions mainly to the east.
The Boreo-nemoral zone comprises also SE Norway, namely the Oslofiord region as far north as Lake Mjfbsa.
According to Moen (1987), also some small areas along the western fiords as far north as the Trondheim region belong to this zone. The phytogeographical position of the hyperatlantic southern and southwestern coast of Norway up to the Trondheim region has been much discussed.
According to Ahti et al. ( 1968), the southern part of this area belongs to the Nemoral zone, and the northern part to the Boreo-nemoral zone. Moen (1987) distinguished a Coastal section as a major phytogeographical unit of its own, characterized by a general scarcity of forest due to logging and subsequent burning, grazing and hay-mak
ing. Sjors (1963), however, defines it as a separate North Atlantic pine-birch woodland and heath region where both Picea abies and Fagus sylvatica are nearly absent.
Here Pinus sylvestris and some Nemoral broad-leaved tree species play an important role. Forests of this region will therefore be included in the comparison of forest vegetation from the Boreo-nemoral zone.
2.2 Geology and Soils
Topography and Geomorphology
In S Sweden, altitude does not vary much. Most areas, particularly in the east, lie below 100 m a.s.l. The interior parts of SmiHand, the so-called south Swedish Highland, only reach about 300 m, but they function as a barrier for the predominant westerly winds, resulting in a rain-shadow situation in SE Sweden. Some areas in Vastergotland and bstergotland have altitudes exceeding 250 m, the most conspicuous being Omberg (263 m) at the eastern side of Lake Vattern and the plateau-like stratified hills in Vastergotland, rising from the surrounding lowland, e.g.
Billingen (299 m), Mosseberg (327 m) and Kinnekulle (306 m). Whereas SW Finland is rather level throughout the study area, S Norway is mountainous, and only the coastal areas and deep inland valleys lie below 500 m.
As to the relative relief, large areas of the Boreo
nemoral zone in S Sweden can be described as plains or joint-valley landscape (Rudberg 1987). The central parts form an undulating hilly landscape of varying altitude.
Pre-Quaternary bedrock
S Sweden, like northern Europe as a whole east of the Scandes, is characterized by predominantly very ancient bedrock, part of the extensive Baltic Shield. Precambrian crystalline rocks, mainly granites and gneisses, cover a major part of the area (Rudberg 1987). These are rather acid and resistant to weathering. Basic (alkaline) crystal
line rocks appear at only a few places, e.g. close to Almunge in the province of Uppland.
The Precambrian stratum forms the basement for lo
calized Palaeozoic sediments of varying thickness, mainly Cambrian sandstone and alum shale, Ordovician lime
stone and Silurian shales. These soft, often calcareous, rocks are preserved only in a few regions, with their largest extension in the province of Jamtland (Fig. 3). Due to their physical and chemical properties, weathering
Acta Phytogeogr. Suec. 80
�
Rocks rich in lime�
Soils rich in lime 1 1 1 Transport direction of I I disintegrated rocl<s rich in lime. . · · �- .
P:lt. ·- -
..� :
\
lOO l·, 200 km
Fig. 3. Distribution of calcareous rocks and soils in Sweden.
Arrows show the dispersal of disintegrated limestone by glacial drift. From Magnusson et al. ( 1963), revised.
results in favourable soils suitable for growth of decidu
ous forest. Within the study area, Carnbro-Silurian rocks can be found in:
- bstergotland (between Lake Vattem and Lake Roxen);
- Vastergotland (Billingen, hills of Falbygden, Halleberg, Hunneberg, Kinnekulle and Lugm1s hill; sedimen
tary rocks often covered by Permian dolerite sills);
- Narke (area west of Lake Hjalmaren);
- bland and Gotland, where the layers form the visible portion of a continuous sequence in the Baltic region;
- some parts of Skane.
Jotnian and Carnbrian sandstones occur in some areas, e.g. in the coastal region of Smaland.
Mesozoic rocks are restricted to Skane, which physio
graphically (and phytogeographically) forms a part of west-central Europe (Sjors 1965a). In SW Finland and on Aland, the bedrock consists exclusively of Precambrian rocks.
In contrast, the bedrock in S Norway is more variable, including Palaeozoic strata. This is particularly true for the Oslofiord region, where Permian sediments cover large areas.
Quaternary deposits
Because Scandinavia has been a land area from the late Silurian up to the present, it has a long history of erosion and denudation (Rudberg 1987). Of major importance are the repeated glaciations during the Pleistocene which resulted in a total (or almost total) coverage by ice. Glacial and glaciofluvial erosion had a great geomorphological impact on the form of the landscape.
After the last glaciation, most of the present land surface of Sweden was covered by till, mostly rich in boulders, glacial or marine clay and coarse-grained glaciofluvial sediments. These deposits vary as to type and distribution, depending on the form of the ice sheet and its movement, type of deglaciation, etc. A large part of S Sweden was under water during the retreat of the ice sheet. When the land emerged from the sea, the glacial material was sorted and partly transported due to wave action (Sjors 1965a). As a consequence, the till cover has been removed from hills and exposed sites and re-depos
ited in lower, sheltered areas. It fills depressions and softens the irregularities of the basal bedrock. Among the apparent surface features are radial and terminal moraines as well as drumlins (G. Lundqvist 1959). Originating from the areas with Palaeozoic sediments, calcareous material has been dispersed southwards by glacial drift into large areas of Vastergotland, Smaland, bstergotland and Narke (Fig. 3, Magnusson et al. 1963). Calcareous soils were also transported to eastern Uppland and Sodermanland, originating from the Ordovician bedrock on the bottom of the Bothnian Sea (Sjors 1965a). How
ever, large parts of the area are no longer covered by any Quaternary deposits, but show exposed bedrock. This is particularly apparent in Bohuslan on the Swedish west coast, in places on the east coast and also on the Baltic islands of bland and Gotland (G. Lundqvist 1959).
Apart from glacial drift, we find glaciofluvial deposits transported by melt water streams which are usually built up of coarse-grained material. The eskers, e.g. the Uppsala esker, are quite striking features on the flat, cultivated
Deciduous forest vegetation in Boreo-nemoral Scandinavia 1 5
plains. Particularly i n the lower elevated parts of the study area, fine-grained sediments have been deposited due to the influence of both standing (fresh or brackish) and running water. Aeolian sands, both early Post-glacial and recent, are only locally frequent and, within southern Sweden, concentrated to the province of Vastergotland and the Baltic islands of bland and Gotland (G. Lundqvist 1959). Peatlands in S Sweden only locally cover large areas, e.g. in Vastergotland and Sm:lland.
Soils
In S Sweden, brown earths play an equally important role as podsols (Troedsson & Nykvist 1973). They are more or less restricted to areas with alkaline bedrock and/or cal
careous Quaternary deposits. The vegetation formerly connected with brown earth was deciduous forest, but the brown forest soils have been transformed into agricultural land to a great extent. As to the humus form, podsols usually are combined with mor and brown earths with mull, respectively.
2.3 Climate
Scandinavia enjoys a higher annual temperature than any other area at a comparable latitude. With its location at the western edge of the big Eurasian landmass, it is influ
enced by warm westerly and south-westerly winds and the Gulf Stream. Due to the proximity of the Atlantic Ocean and Baltic Sea, the climate is fairly maritime, and the winters are relatively mild.
Temperature
Within the study area, the annual mean temperature is mainly a function of latitude. Besides, increasing altitude as well as increasing distance to the sea or extensive inland waters result in decreasing annual mean tempera
tures (Eriksson 1982). The values vary between 5.0 °C and 7.0 °C, compared with 7.0 oc to 8.0 oc in southemmost Sweden. In the surroundings of the big lakes (Lake Vanern, Lake Vattern and Lake Malaren), the annual mean tem
perature is comparatively high. In contrast, the interior parts of Smaland show comparatively low values (down to 4.8 °C).
Despite the great latitudinal extension of the area, the mean temperatures of the warmest month (July) are fairly equal, ranging from about 17 oc to 15 °C, with the lower values at higher altitudes. The value for Uppsala (16.5 °C) is about the same as for most climate stations in Skane, situated 600 km to the south (Eriksson 1982).
The mean temperature of the coldest month (usually February) shows a much higher variability. In general, it decreases with latitude and distance to the sea. Whereas
Growing season (days)
400 km
Fig. 4. Length of the growing season (threshold +5 °C) in Sweden, Finland and northern Norway. From Tuhkanen ( 1980).
the values for Blekinge and bland vary between - 1.0 oc and - 2.0 °C, they drop off to - 4.0 oc to - 5.0 oc in the inner parts of Sm:lland and the northern provinces of the Boreo-nemoral zone, e.g. Uppland. The mild winters of the coastal areas of S Sweden (and W Norway) are of great importance for the vegetation and allow frost-sensi
tive species to extend their distribution far to the north (Sjors 1965a; Hulten 1971).
The risk of frost during the growth period of plants is also of botanical importance. It is particularly low in coastal areas and in the surroundings of the big lakes, both in spring and autumn (Angstrom 1974).
Growing season
The growing season is defined as the time period with a mean diurnal temperature above a certain threshold (usu
ally 5 °C). Fig. 4 shows the length of the growing season in Sweden and Finland according to Tuhkanen (1980).
The isoclines correspond well with the borders of the main vegetation zones. The southemmost parts of Swe
den, belonging to the Nemoral zone, enjoy the longest
Acta Phytogeogr. Suec. 80
l:r
2500Annual precipitation
Fig. 5. Annual precipitation in mm in 1 93 1 - 1 960 in northern Europe. From Nordseth ( 1 987), revised.
growing season, i.e. Skane and the southern parts of the Swedish west and east coast. The Baltic islands of bland and Gotland show comparatively high values. The 180- days isocline is well in accordance with the distribution limit of Que reus robur both in Sweden and Finland, and, hence, with the borderline of the Boreo-nemoral zone.
The interior elevated parts of Smaland have a growing season shorter than 180 days which supports the decision of Ahti et al. (1968) to include this region in the Boreal zone. The use of temperature sum (with a 5 oc threshold) instead of growing season reveals a very similar isocline pattern (Tuhkanen 1980).
Precipitation
Due to the prevailing wind directions and orientation of the Scandes, northern Europe shows a strong West-East gradient in precipitation. Whereas the annual amount of rain in western Norway exceeds 2000 mm, it may fall below 400 mm in areas situated in the rain-shadow of mountains, e.g. in valleys of south-central Norway and in Finnmark (Fig. 5, Nordseth 1987). In accordance with this gradient, the highest values within the study area are shown by the western provinces of Sweden. The average annual precipitation amounts to 900-1100 mm in parts of Halland, BohusHin, Smciland and Vastergotland, but only
de Martonne's i ndex
70 56
0 1 00 200 km
Fig. 6. de Martonne's index in Sweden, applied to the growing season only (based on a 3 oc threshold). From Angstrom ( 1 974 ), revised.
to 400-500 (600) mm at the Swedish east coast and the Baltic islands of bland and Gotland. There is an appreci
able increase with altitude. On the other hand, areas in the vicinity of the big lakes show comparatively low values.
With respect to the seasonal distribution of rain, maxi
mum values almost always occur in late summer (July and August). A second maximum is shown in autumn by some stations at the west coast, due to the increasing frequency of south-westerly winds during this season. The lowest values are generally found in late winter/early spring (Angstrom 1974).
Deciduous forest vegetation in Boreo-nemoral Scandinavia 1 7
Snow keeps the temperature at ground close to the freezing point, even when the air temperature is much below 0 °C. This is of great importance for plant growth.
The average duration of snow cover varies between ea. 40 days in the southernmost parts of Sweden and ea. 1 20 days at the northern border of the Boreo-nemoral zone.
However, the actual duration fluctuates greatly from year to year.
Climatic indices Moisture conditions
In order to characterize the relative climatic humidity or aridity of an area, one usually relates precipitation and temperature (or evaporation) to each other in a formula (Tuhkanen 1 980). One of the most widespread indices is de Martonne' s 'indice d'aridite' (better called humidity index because values increase with humidity) which has the form: H = P/(T + 1 0), where P is the annual precipita
tion and T the annual mean temperature in °C (de Martonne 1 926). In a slightly different form, the index can also be calculated for shorter time periods. Fig. 6 (Angstrom 1 97 4) gives de Martonne' s index based on the growing season for Sweden. It is in quite good accordance with the distribution of rain as given above, with the highest values in SW Sweden and extreme low values in a narrow band along the Swedish east coast and on bland and Gotland.
The surroundings of the big lakes show comparatively low humidity, and the increase with altitude is even clearer than for precipitation alone. The moisture conditions can be graphically displayed in the climatic diagrammes of Waiter (Waiter & Lieth 1 960- 1 967).
Continentality/maritimity
Continentality indices can be based on both thermal and hygric properties of the climate. Often we make use of the annual temperature range, corrected for the effect of lati
tude (Tuhkanen 1 980). In Conrad' s continentality index [C = 1 ,7 X A/sin (% + 1 0 °C)], A denotes the annual temperature range and % the latitude (Conrad 1 946). The higher the value of C, the more continental is the climate.
Within Sweden, C-values are lowest in the coastal areas
Continentality index
400 km
Fig. 7. Conrad's continentality index in Sweden, Finland and northern Norway. From Tuhkanen ( 1980).
of southern Sweden and increase towards the interior and towards the north (Fig. 7, Tuhkanen 1 980). A similar pattern of relative continentality and maritimity results from a survey of temperature anomalies, a measure of the deviation of A and annual mean temperature from the mean for the latitude in question (Angstrom 1 974). W Norway has an extreme oceanic climate with high pre
cipitation and a small annual temperature range.
Acta Phytogeogr. Suec. 80
