LUNDQUA Thesis 30 Interaction of Holocene climate, water balance, vegetation, fire, and cultural land-use in the Swedish Borderland Heather Almquist .. Jacobson l\;f.Sc., Vrml Avhandling att I1lcd tillstand fdln Matcmatisk·Naturvctcnskapliga Fakultctcn vid Lunds univcrsitd t('lr avWggandc av filosolic doktorscxamcn oflCnlligcn t('lrsvaras i nalurgcogra~iska institutioncns ft1rd:isningssal. St))vcgatan U. Lund. fn.'dagen den 11 mars 1Q Q4 kl. 10 L ~ ~ ~ Lund 1994 Lund University. Department of Quaternary Geology In memory of my grandmother Svea Larson Almquist who loved Sweden and all the natural world - Contents 1 Introduction ............................................. . 2 Description of the Study Area ................................. " 2.1 The Borderland .................................. " 2.1.1 Geology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . " 2.1.2 Climate ..................................... " 2.1.3 Vegetation ..................... . . . . . . . . " 2.2 The HaIlefors Area ................. . . . . . . . . . . . . 7 2.2.1 Surficial Geology ............................ 7 2.2.2 Post-Glacial Geologic History . . . . . . . . . . . . . . . . . . . . . 7 2.2.3 Modem Vegetation and Soils ..................... 7 2.2.4 Cultural History ............................... 7 2.3 Investigated Field Sites .................................... 8 2.3.1 Lilla Gloppsjon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.3.2 LjustjarDen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3.3 Local Peat Deposits ................................ 10 .J Field and Analytical Methods ................................. 11 3.1 Reconnaissance of Field Sites ............................... . I1 3.2 Core Collection. Description, and Subsampling ............. . 11 3.2.1 Lilla Gloppsjon ................................. . 11 3.2.2 Ljustjamen .............................. . 11 3.2.3 Local Peat Deposits .......................... . 12 3.3 RadiocaJbon Dating . . . . . . . . . . . . . . . . . . . . . . ....... . 12 3.4 Sedimentologic Analyses ......................... . 12 3.4.1 Bulk Organic. Inorganic. and Carbonate Fractions '" 12 3.4.2 Biogenic Silica .............................. . 12 3.4.3 Magnetic Properties . . . . . . . . . . . . . . . . . . . . . . . . . 12 3.4.4 Coarse Organic and Inorganic Matter ....... . 13 3.5 Fossil Analyses ..................... . 13 3.5.1 Pollen ................ . 13 3.5.2 Macrofossils ......................... . 13 3.5.3 Charcoal ................... . 13 3.6 Statistical Analyses ................................. . 15 4 Chronology............................................... 16 4.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 4.2 Assumptions of the Age Models . . . . . . . . . . . . . . . . . . . . . . . . . .. 16 4.3 Age Models. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 4.3.1 LiIla Gloppsjon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17 4.3.2 Ljustjamen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1ft S Local Pollen Assemblage Zones ................................ 20 5.1 Lilla Gloppsjon ..................... ...... 20 5.2 Ljustjamen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 22 6 Lake-Level History ......................................... 26 6.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 6.1.1 Research Objectives . . . . . . . . . . 26 6.1.2 Importance of Paleohydrologic Records in Em'ironmcntal and Climate Reconstruction 26 6.1.3 Sensitivity of Lake-Level Records . . . . . . . . . . . .. ........ 26 6.2 Lake-Level History of the Hilllefors Area 29 6.2.1 Evidence from Lilla GloppsjOn . . . . . . . . . . . . 29 .- 622 Evidence from Ljustjarnen ............................ 37 62 \ Comparison of Sandplain & Upland Records. . . . . . . . . . . . . . .. 47 6.24 Relative Sensitivity of the Lake-Level Indicators .,. . . . . . . . . .. 48 7 Vegetation History. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . .. 50 7.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 50 7.1.1 Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 50 7.1.2 Factors Influencing Vegetation ......................... 50 7.1.3 Detennining Local Presence of Tree Taxa from Pollen Data ..... 50 7.2 Vegetation History of the Hiillefors Area ........................ 53 7.2.1 Pioneer Vegetation (9600t09IOOBP) ................... 53 7.2.2 Pinus Forest (9100 to 8200 BP) ........................ 54 7.2.3 Mixed Ulmus Forest (8200 to 6200 BP) .................. 54 7.2.4 Mixed Quercus Forest (6200 to 2200 BP) . . . . . . . . . . . . . . . .. 54 7.2.5 Mixed Picea Forest (2200 BP to present) ................. 56 7.2.6 Comparison of Sandplain and Upland Areas .... . . . . . . . . . . .. 57 8 Fire History .............................................. 58 8.1 Background. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 58 8.1.1 Research Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 58 8.1.2 The Importance of Fire in Northern Temperate Forests . . . . . . . .. 58 8.1.3 Factors Affecting Fire ............................... 58 8.1.4 Limitations of Charcoal Data in Reconstructing Fire History ..... 60 8.2 Fire Histol)' of the Hiillefors Area . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 61 8.2.1 Regional Overview ................................. 61 8.2.2 Local Fire Regimes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 61 8.2.3 Factors Controlling Fire in Sandplain and Upland Areas . . . . . . .. 61 9 Synthesis................................................. 63 9.1 Influences of Climate & Disturbance on Forest Development .......... 63 9.1.1 Local Interactions Among Water Balance. Fire. and Vegetation ... 63 9.12 Influence of Human Activities on Regional Forest Development .. 63 9.2 Paleoclimatic Implications of Geographic Patterns in Water Balance ...... 68 9.2.1 Factors Controlling the Modem Climate of Scandinavia ........ 68 9.2.2 Comparison of Water Balance Records. . . . . . . . . . . . . . . . . . .. 69 9.2.3 Implications for the Paleoclimate of Scandinavia .' . . . . . . . . .. 71 9.3 Final Comments ......................................... 71 Ackno~·ledgements ...................................... 73 References ............................................ 7 .. Plates 1-3 ii 1 Introduction The "Borderland" of south-central Sweden is a strongly contrasting geologic settings, the histOlY physiographically complex region that includes of fire and its impact on vegetation of the region hilly upland areas interrupted by valleys, many may have been particularly complex. The fire of which contain glacio-fluvial outwash plains. history of the Swedish Borderland has not previ It corresponds to an important modem transition ously been studied. zone for climate and vegetation, i.e. the southern To summarize, past interactions of vegetation, margin of the northern boreal forest, and is com water balance, fire, and human activities have parable to that transition elsewhere in the north implications for both climate reconstructions and ern hemisphere. Because this ecotone is accen for our general understanding of landscape de tuated in Sweden by a distinct gradient in topog velopment. Climatic reconstructions based pri raphy and surficial geology, the ecosystems and marily on fossil-pollen data are strengthened and vegetation of the region have likely always been tested by independent stratigraphic evidence for especially sensitive to past changes in climate. those other factors. In many ways water balance Vegetation is influenced by climate through provides a critical independent measure of cli complex interactions among many aspects of the mate. Comparison of the direction and timing of environment. These include air temperature, changes in regional moisture balance across sou water balance, and fire regime, each of which thern Scandinavia may help to reveal the mecha operates at a different spatial scale. In the nisms that control Holocene climate there. The broadest sense, temperature is the most important Borderland is ideally located for reconciliation of determinant for vegetation, so subcontinentaI the disparate climate records from southern scale reconstructions of paleotemperatures can be Scandinavia. Lake-level changes in the Border based on fossil-pollen data. However, even in land, for example, can be compared with lake north-temperate regions, water balance can influ level changes in southern Sweden and with gla ence vegetation, either directly through the phys cier fluctuations in western Norway (Fig. I) to iology of the species, or indirectly by influenc determine whether differences between the latter ing fire regime. Although many high-quality two records result from a latitudinal difference in paleohydrological studies have been carried out climate or simply from differential response of in southern Sweden (Digerfeldt 1971, 1972, glaciers and lakes to climate change. 1974, 1975, 1976, 1988; Fig.I), and some infor Lake-level reconstructions provide the most di mation regarding water balance can be inferred rect assessment of past water-balance. However, from studies of glacier fluctuations in western in north-temperate regions, where fluctuations in Norway (Nesje 1992), those records do not cor lake level are relatively small and the strati respond closely with one another, and they re graphic evidence of those changes is subtle, spe veal little about Holocene water balance in cen cial care is required to identify the best sites for tral Sweden. such studies and the most useful indicators of On a local scale, disturbances such as fire and lake stand. No such studies have been complet human activities (Bradshaw & Hannon 1992; ed previously for northern or central Scandina Foster 1993), along with geologic and edaphic via. factors (Brubaker 1975; Jacobson 1979), can The first goal of this research is to develop a strongly influence vegetation and landscape de holistic, paleoenvironmentaI reconstruction of the velopment. For thousands of years humans have HaIlefors area, which lies within the Borderland played an important role in the environments of region in south-central Sweden. Changes in wa Europe. Their impacts on the landscapes of sou ter balance, "egetation, fire regime, and human thern Sweden, the British Isles, and central Eu disturbance are assessed for two contrasting sites rope are well documented. In contrast, little is within the complex landscape -- a comparison of known about the timing and extent of cultural sandplain with upland areas -- to help assess the land-use in central Sweden, or whether human importance of variations in topography and eda activities there could have been of local or re phic conditions on forest development. Specifi gional importance in forest development. Fire, cally. this study assesses (l) the Holocene histo either natural or human-induced, can act as a ry of water balance, (2) the local interactions significant agent of disturbance that influences among water balance, fire, and vegetation, and vegetation through complex feedback mecha (3) the possible local and regional significance nisms, mediated by local edaphic and topograph of human activities on forest development. ic conditions. Because the Borderland includes A broader aim of this research is to compare 1 - • 12' 20' Barents Sea m a.s.!. 2500 1000 500 68' 200 68' 0 ; Norwegian Sea 60' 56' Baltic North Sea Sea 300 km 28' Fig. 1. Locations of (1) the study area, south-central Sweden, (2) previous studies of Holocene lake-levels, southern Sweden, and (3) studies of glacier fluctuations, Jostedalsbreen, western Norway. 2 the timing and direction of lake-level fluctua of the research is to develop further the tech tions in the Swedish Borderland with those in niques for reconstructing lake-levels in unpro southern Sweden and with glacier fluctuations in ductive lakes of north-temperate regions, and to westcrn Norway. Such comparisons may reveal determine the relative hydrologic sensitivity of past shifts in airmass boundaries and the position drainage vs. seepage lakes, as well as the rela or strength of cyclonic storm tracks in Scandina tive usefulness of the various sedimentologic via. Both may be detccted from temporal and indicators for determining past water-levels in spatial patterns in water balance. The final aim lakes of each type. 3 - 2 Description of the Study Area 2.1 The Borderland frost leaves the ground after May 10. (Those events have occurred over a month earlier in this decade, however.) The Swedish "Borderland" (sensu Fransson During the period 1901 through 1930, mean 1965) extends from the Hardanger Upland of annual precipitation in the Borderland was 550 Norway, which lies 100 km east of Bergen, to to 800 mmlyr, with approximately twice as much the province of Gastrikland on the Swedish east falling during summer and fall than during win coast. It is an important geologic, climatic, and ter and spring. Mean July temperatures ranged floristic transition zone, particularly in the west around 15°C, and mean January temperatures ern and central portions (Nordiska Ministerradet were approximately _5°C (Angstrom 1974). The 1984; Fig.2). (patterns are complicated by mari "vegetation period" (the numbers of days in time effects toward the Baltic coast.) Distinct which the mean temperature is greater than 3°C) latitudinal changes in altitude, geomorphology, was c. 190 days, mean annual precipitation and and Quaternary deposits intensify the climatic annual potential water-balance (P-PET), ranged gradient. Consequently, many plant taxa reach between 700 and 900 mm, and between 0 and 20 their distributional limits in this area. mm, respectively (Angstrom 1974; Fig.2). Alti tudinal gradients within the hilly landscape also affect climate. Precipitation increases c. 100 2.1.1 Geology mm for every 100 m elevation, for example (Sjors 1948). The central portion of the Borderland corre sponds roughly to "Bergslagen", the old mining districts of southern Dalarna, northern Viistma 2.1.3 Vegetation nland, and eastern Vannland. Bergslagen is an area of "chequer plateau land and broad valleys" Vegetation and biotic complexes of Sweden have (De Geer 1910) marking the topographic bound been described most recently by Lindquist & ary between the northern upland hills, which rise Lundqvist (1962) and Sjors (1963, 1965b) (Table to 300 or 400 m a.s.l., and the southern lowland 1). The Borderland corresponds to a narrow plains, which lie below 100 m a.s.l. (Fransson phytogeographical transition zone through which 1965). the Boreo-Nemoral region in the south is re The highest former Late Glacial coastline lies placed by the Southern Boreal region in the around 180 to 190 m a.s.l. (Granlund 1928). north. The Boreo-Nemoral forest is dominated Above that altitude, hilly areas are covered by by Picea abies (L.) Karst. and Pinus silvestris ablation till rich in sand and boulders, and broad 1., with significant amounts of Quercus robur valleys contain sandy deltas and outwash plains 1., Fraxinus excelsior L., Ulmus glabra Huds., (SjOrs 1948). Clayey marine sediments, wave Acer platanoides L., Tilia cordata Mill., Sorbus washed till, and bare bedrock outcrop are com intermedia (Ehrh.) Pers., Corylus avel/ana 1., mon below the marine limit. Thorough des and Alnus glutinosa (L.) Gaertn. In contrast, criptions of comparable surficial deposits from Alnus incana (L.) Moench joins A. glutinosa in neighboring Vannland are given by Lundqvist the Southern Boreal forest, and the temperate (1958). Bedrock of the region includes granite, deciduous component is restricted to scattered gneiss, and leptite, which are relatively hard and occurrences of Acer, Tilia, Ulmus, and Corylus insoluble (SjOrs 1965a, 1948). Industrial miner on favorable sites. Quercus is notably absent. als include silver, lead, and iron. The land is The northern distributional limit of Quercus currently rising at a rate of c. 0.5 cm/yr (Fromm (Fig.3) has itself long been considered an impor 1971). tant dividing line between southern and northern vegetation (WahJenberg 1826; Myrin 1832; E. Fries 1856, 1864). 2.1.2 Climate This ecotone, which comprises gradients in both physical and biological factors, including Winters are longer and colder in the Border secondary effects such as soil formation and land than on the southern plains. The snowpack modern land-use practices, initially was referred usually persists more than 120 days (Fig.2), to by Rutger Semander as "limes norrlandicus" springs are late and fast, and autumns come ear (Fries 1948) and by Du Rietz (1935) as "the ly. Ice usually leaves the lakes after May 1, and natural borderline between North and South 4 SNOW MEAN COVER JANUARY (days) TEMP. (OC) VEGETATION P-PET PERIOD (mm) (days) o 200 '---'---' km Fig. 2. Geographic distribution of vegetation zones and mire complexes in Sweden, in relation to trends in climatic parameters during the period 1901 to 1930, including duration of snow cover, mean January temperture, duration of the vegetation period, annual precipitation, annual potential waler balance (P-PET), and mean July temperature (from Angstrom 1974). The Swedish Borderland is indicated by the shaded pattern. Sweden". The floristic boundary is actually the boundary, while many other species includ rather diffuse and amounts to changes in species ing Betula nana L., Alnus incana, Salix population density rather than presence/absence lapponum L., SeJagineJla seJaginoides L. Link, (Fransson 1965). Some northern species such as Carex brunnescens (pers.) Poir., and C. vaginata Cicerbita alpina CL.) Wallr. (= Lactuca alpina Tausch, are simply reduced in frequency there. CL.) Gray), Tofieldia pusi/la (Michx.) Pers., Some southern species penetrate into the north Phleum alpinum L., Epilobium hornemannii ern valleys, but overall species diversity decreas Rchb., E. lactijlorum Hausskn., and Sparganium es northward (Hulten 1971). hyperboreum Beurl. are virtually absent south of 5 - Table 1. Approximately corresponding phytogeographical c/assificatio~ sche",.es Jor Sweden inc!.uding "vegetation regions" oJ Lindquist & Lundqvist (1962). "forest regions oJ ~ors (1965b). and bIOtIC regions" oJ Sj6rs (1963). The heavy line corresponds to the Borderland transItIOn zone. Vegetation Regions Forest Regions Biotic Regions Mountain Alpine Alpine Mountain-Birch Forest Birchwood Sub-Alpine Birch Woodland ---------M-o--u-n-ta-i-n- -T-a-i-g-a- ------- ----------B--o-r-e-o--M--o-n-t-a-n-e- ---------- Northern ----------C-e-n-t-ra-l- -T-a-i-g-a- -------- Coniferous Main Boreal South Taiga Southern Boreal Oak & S ruce Forest Southern Coniferous Boreo-Nemoral Oceanic Oak Forest Southern Northern Deciduous Nemoral Beech Forest (" (b) , , \ \ I I I r' Fig. 3. Distributional limits oJ selected tree taxa (after llulU!n 1971). including (a) approximate northern boundaries oJ (--) Quercus rohur. (-- -) Acer platanoides. (_. _) Tilia cordata. (_ .. _) Fraxinus excelsior. and ( .... ) Ulmus glabra. and (b) sourthern and western boundaries oJ (_) Picea abies. and (-... ) Pinus sylvestris. The Borderland also coincides with a shift from of Bergslagen also has a distributional gradient predominantly raised bogs in the south to from southeast to northwest (SjMs 1948), and eccentrically sloping bogs in the north (Fig.2). At decreasing humidity may be the primary factor higher altitudes, soligenous fens become more limiting northern species. common than ombrogenous bogs. According to Fransson (1965), those changes result more from gradients in precipitation and geologic setting than from differences in temperature. Mire flora 6 ,