Vegetation History and Archaeobotany

, Volume 24, Issue 2, pp 253–266 | Cite as

Reconstruction of past landscape openness using the Landscape Reconstruction Algorithm (LRA) applied on three local pollen sites in a southern Swedish biodiversity hotspot

  • Tove Hultberg
  • Marie-José Gaillard
  • Britt Grundmann
  • Matts Lindbladh
Original Article

Abstract

The Hornsö–Allgunnen area in south-eastern Sweden has been known as a biodiversity hotspot for insects for about a century and is considered to host the most species-rich insect fauna in northern Europe. Several hypotheses for the causes behind this biodiversity have been put forward, but never tested for more than small parts of the area. We analyse here the possible role of the area’s vegetation-cover history, in particular vegetation openness. We use pollen data from three sites in the Hornsö–Allgunnen area and apply the recently developed Landscape Reconstruction Algorithm (LRA) for quantitative reconstruction of past vegetation abundance at the local spatial scale. The study suggests that the area was dominated by diverse, relatively open forest during at least the last 3,000 years. Several tree taxa, such as Pinus, Betula and Quercus that were all suggested to be important for the present diversity, have a long continuity at the local spatial scale and were common until recently. Small proportions of anthropogenic pollen indicators were found, suggesting small-scale agriculture, which however did not considerably affect the area’s overall tree species composition. We propose that fire was the main cause for the open character of the area’s wooded landscape during the Holocene and, indirectly, an important agent behind the high insect diversity. However, the richness of insects was (and is) most likely also favoured by the long continuity of Quercus, and by the warm and dry local climate. The LRA provides a more realistic estimate of the taxa composition as compared to pollen percentages alone, both for arboreal and non-arboreal taxa. The differences between pollen percentages and LRA-estimates of plant abundance can be important to consider when causes behind high modern diversity are interpreted from fossil pollen records. Our results demonstrate the benefits of using the LRA along with traditional pollen percentages.

Keywords

Pinus sylvestris Quercus robur Saproxylic beetles Forest openness Southern Sweden 

Introduction

The Hornsö–Allgunnen area in south-eastern Sweden has been known as an insect hotspot for about a century. It is considered to host the most species-rich insect fauna in northern Europe; in particular many rare, wood-associated insects are found in the area (Ehnström and Axelsson 2002; Nilsson and Huggert 2001). Several hypotheses have been put forward to explain this high biodiversity, including warm local climate, frequent fires until the 20th century, large variation among habitats, and abundant old trees and dead wood (Anonymous 2008; Lindbladh et al. 2003). However, the causes for the large insect diversity are still not fully known and might not be possible to determine using only the contemporary information on the area. Long-term palaeoecological data could provide additional important insights. In the largest published survey of the area’s insect population, Nilsson and Huggert (2001) suggest that “for saproxylic insects the biotope availability and the history of the area are crucial”. Furthermore, there is a general acceptance today of the importance of an area’s long-term ecological history for the understanding of present biodiversity and for setting up adequate management strategies to maintain and/or favour it (Foster 2002; Lindbladh et al. 2013, Willis et al. 2007; Willis and Birks 2006).

The vegetation history of the Hornsö–Allgunnen area has been investigated earlier to some extent. Pollen and charcoal analysis in combination with dendrochronology at one of the sites, Skärsgölarna, was carried out by Lindbladh et al. (2003). Fire scars in dead and living trees, as well as fossil charcoal data, suggested that local fires had been common for millennia. The authors concluded that frequent and continuous fires kept the site relatively open and probably were a prerequisite not only for the long-term dominance of Pinus sylvestris (scots pine), Betula spp. (birch) and Calluna vulgaris (heather), but also for the almost total absence of all shade-tolerant tree species. However, that study included the analysis of only one site (a small forest hollow sensu e.g. Jacobson and Bradshaw 1981, and Overballe-Petersen and Bradshaw 2011) with a small pollen source area, which severely hampered interpretation regarding the vegetation history of the entire area. Furthermore, only pollen percentages were used for vegetation reconstruction, which does not reflect actual vegetation composition in terms of abundance of each taxon (Sugita 2007a, b).

In this study we use pollen data from three small sites in the Hornsö–Allgunnen area and two large sites in the surrounding region, and apply the Landscape Reconstruction Algorithm (LRA) (Sugita 2007a, b) to estimate the percentage cover of plant taxa at the local spatial scale, i.e. within the pollen source area (relevant source area of pollen, RSAP sensu Sugita 1994) of the three small sites. Several approaches or methods have been developed to “translate” pollen data into vegetation abundance, e.g. Davis (1963), Andersen (1972), Webb et al. (1981), Björse et al. (1996) and Sugita (2007a, b). LRA is a new tool for vegetation reconstruction that was found to perform better than previously tested models and correction factors (Hellman et al. 2008a; Nielsen and Odgaard 2010, Sugita et al. 2010). Using three sites enables us to reconstruct the vegetation of a larger part of the study area and to detect possible vegetation differences among sites. Our purpose is to look for possible links between present day insect biodiversity and the long-term vegetation history of the Hornsö–Allgunnen area in terms of plant abundance, vegetation/forest openness, and fire activity. We also analyse the hypothesis that open forest and fire-resistant and/or fire-prone vegetation has a long continuity in the Hornsö–Allgunnen area (Lindbladh et al. 2003).

Materials and methods

Study area

The Hornsö–Allgunnen area (57°01′N, 16°07′E) (Fig. 1) is located in the county of Kalmar, south-eastern Sweden. The ice sheet of the Weichselian glaciation retreated from the area ~12500bp and Hornsö–Allgunnen lies at ~85 m.a.s.l., just below the highest coast line (Anonymous 2006; Fredén 2002; Nilsson et al. 1995; www.sgu.se2010). The region is part of the trans-Scandinavian granite-porphyry belt, dominated by granite, but with elements of acid volcanic bedrock, as well as basic bedrock covered by moraine with abundant boulders (SGU 2009a, b). The Hornsö–Allgunnen bedrock has a pH of 5.2–5.8 (Fredén 2002). The average temperature is −2 °C in January and +18 °C in July (www.sna.se2010), and the average annual temperature (6.6 °C) is among the warmest in the region (Alexandersson and Eggertsson Karlström 2001). Annual precipitation is ~500 mm/year, which is lower than both the Swedish and the south Swedish inland averages (both ~700 mm/year) (www.smhi.se2014).
Fig. 1

Location of the surveyed sites (shaded = “Eco Park”-area) and location in Sweden. Map: Emma Holmström

Large parts of the Hornsö–Allgunnen area used to be a grazed common (Swedish: allmänning), but were purchased by the Swedish state in 1866. Today most of the area is owned by the state forest company Sveaskog that manages it as an “Eco Park” with multiple uses. At least 50 % of the area is set aside from wood production and managed primarily for preservation of biodiversity and/or recreation. The park is freely accessible to visitors and not formally protected, but includes several nature reserves. At present the park vegetation is characterized by Picea abies (spruce) (~40 %), Pinus sylvestris and Betula sp. (Anonymous 2008). Pinus, Quercus robur (oak), Betula and Fagus sylvatica (beech) have been suggested to be specifically important for the diversity of the wood associated beetles in the area, but Hornsö–Allgunnen also contains valuable biotopes of Picea and Populus tremula (aspen) (Anonymous 2008; Speight 1989). The villages of Hornsö and Allgunnen neighbour the park, which also includes several other smaller villages. The first known farms in the area are from the 16th century, but archaeological records indicate that the area was inhabited before the middle ages (Anonymous 2006; Brunius and Ferm 1990; Ferm et al. 1987; www.raa.se2011). Today’s population density in the area is low, about 1/4 of the national mean of 23 inhabitants/km2 (Anonymous 2011).

The three small wetlands were selected on the basis of location in order to reflect different parts of the area, in combination with sufficient sediment depth and preservation for palaeoecological analysis. They are located within ~8 km of each other (Fig. 1). Lillegölen (see also ESM) is a forested peat bog of ~250 × 100 m overgrown by sparse Pinus and Betula with a field layer dominated by Vaccinium spp. (bilberry) and Sphagnum spp. (peat moss). The peat core was collected close to the edge of the wetland. Skärsgölarna is a ~50 × 30 m wetland surrounded by forest on sandy and gravely moraine with conspicuous boulders. It is sparsely forested with mainly Betula, with an understory of Sphagnum spp., Dicranum spp. (fork moss) and Pleurozium schreberi (red-stemmed feather moss), as well as Vaccinium spp. and Rhododendron tomentosum (marsh labrador tea) (Lindbladh et al. 2003). Ekenäs (see also ESM) is a ~30 × 70 m wetland overgrown by Salix spp. (willows) and single Betula trees, with a field layer of Carex vesicaria (bladder sedge), Calliergon cordifolium (calliergon moss) and Sphagnum palustre (blunt-leaved bog moss), and surrounded by forested sandy moraine with boulders (Valdemardotter 2001).

The application of the LRA requires estimates of regional vegetation abundance, which is best obtained using pollen records from large sites, preferably lakes (Sugita 2007b). In this study, we use the same Holocene pollen records as Cui et al. (2013) as these are from the large lakes with pollen records that are closest to our study area, i.e. Lake Trummen (Digerfeldt 1972) and Lake Kansjön (Jacobsson unpublished, data extracted from the European Pollen Database). They both have a radius of ~250 m and are located ~77 and 118 km from the Hornsö–Allgunnen area respectively (Fig. 1).

Field and laboratory techniques

Field and laboratory techniques for Skärsgölarna are described in Lindbladh et al. (2003). At Ekenäs and Lillegölen sediment cores were collected using a Wardenaar sampler (Wardenaar 1987) in 2001 and 2009 respectively. The 75 cm peat monolith from Ekenäs was divided into 0.4 cm pieces, 25 of which were subsampled for pollen (0.5 ml/sample, using a volumetric sampler) and 29 for charcoal (1 ml/sample, measured by liquid replacement) (Valdemardotter 2001). For Lillegölen, the peat monolith (91.5 cm) was sliced into 1.25 cm pieces. 55 slices were subsampled for pollen and charcoal analysis (1 ml/sample and 1–2 ml/sample, respectively, measured by liquid replacement). Subsamples were stored cold in labelled plastic bags.

Pollen preparation was carried out according to standard procedure (Moore et al. 1991). Between 508 and 931 pollen grains from seed plants were counted in the samples from Lillegölen, except for the 87.5 cm level (203 grains). In the Ekenäs profile, a minimum of 400 grains/level was counted (Valdemardotter 2001). Grains that were not possible to identify because they were hidden, corroded, broken or folded were not included in the pollen sum. The identification was carried out using the identification keys of Moore et al. (1991), the pollen atlas of Reille (1992) and the reference collection at the Southern Swedish Forest Research Centre at SLU Alnarp. Subsamples for macrocharcoal analysis were dissolved in ~5 % sodium hydroxide (NaOH) solution for ≥24 h before they were washed through two sieves with 5 and 0.28 mm meshes. The dissolved material was sorted using a Nikon SMZ645 stereo microscope and stored cold in a solution of ~1 % hydrochloric acid (HCl). The samples for charcoal analysis from the profiles of Skärsgölarna were collected as continuous series, whereas the Lillegölen profile was sampled at 1.25–3.5 cm intervals and Ekenäs at ca. 2.5 cm intervals.

Radiocarbon dating and chronologies

14C datings were performed using plant macro remains or bulk sediment (Table 1). 14C dates from the three small sites and the two lakes were calibrated to calendar years using the software Clam 1.0.2 (Blaauw 2010) that applies the IntCal09 calibration curve. Age-depth models were constructed simultaneously using the same software and the smooth spline model (Table 1; Fig. 2). Confidence intervals were calculated at 2 SD (95 %). Henceforth, all ages are given in calibrated years before present (cal. yrs bp).
Table 1

Radiocarbon dates for the sites Lillegölen, Skärsgölarna and Ekenäs

Beta Beta Analytic, Miami. LuS radiocarbon dating laboratory, Lund University. Ua radiocarbon dating laboratories, Uppsala University

* Omitted from analysis

Fig. 2

Age/depth models for the three study sites established using the CLAM software. The scale is in calibrated years before present (bp = 2000)

The Landscape Reconstruction Algorithm (LRA)

The LRA consists of two models; the REVEALS model (Regional Estimates of VEgetation Abundance from Large Sites) estimates regional vegetation composition using pollen data from large sites or multiple small sites and the LOVE model (LOcal Vegetation Estimates) calculates relative vegetation abundance within the Relevant Source Area of Pollen (RSAP sensu Sugita 1994) using pollen data from small sites and REVEALS-based estimates of regional vegetation abundance (Sugita 2007a, b; Sugita et al. 2010). REVEALS was shown to model modern regional vegetation in southern Sweden reasonably well for an area of 100 × 100 km using modern pollen data from large lakes (Hellman et al. 2008a, b). The RSAP is defined as the area beyond which the pollen-vegetation relationship does not improve, i.e. it is the minimum area for which vegetation can be reconstructed in quantitative terms (e.g. percentage cover) (Sugita 1994). The major factor influencing the size of RSAP is the spatial distribution of plant taxa, where larger vegetation patch size can cause larger RSAP (Bunting et al. 2004; Broström et al. 2005; Nielsen and Sugita 2005; Gaillard et al. 2008; Hellman et al. 2009a, b). Neither wind speed and dispersal properties of pollen grains (Nielsen and Sugita 2005) nor atmospheric conditions (stable or unstable) (Gaillard et al. 2008) seem to affect the RSAP much. The RSAP of small sites (bogs or lakes <0.5 ha) in southern Sweden was estimated to 1–2.5 km radius using hypothetical past landscapes and a forward modelling approach (Hellman et al. 2009a, b), even though RSAPs of 200–400 m have been found for modern pollen assemblages in moss polsters (e.g. Broström et al. 2005; Mazier et al. 2008), which would be analogous to pollen deposition in very small bogs of ca. 0.5 m diameter. Henceforth, the RSAP is simply referred to as the surrounding of the site. The LRA (REVEALS and LOVE) is still experimental, but has been evaluated in several studies in North America (Sugita et al. 2010), Denmark (Nielsen and Odgaard 2010), and southern Sweden (Cui et al. 2013; Fredh 2012).

In this study the regional vegetation cover (percentage cover of each taxon) was estimated using REVEALS and pollen data from the large lakes Trummen and Kansjön, while local vegetation cover within the surroundings of the wetlands Lillegölen, Skärsgölarna and Ekenäs was reconstructed using LOVE. We selected 20 pollen taxa (Table 2) (representing 95.2–100 % of the total pollen counts for the three small sites) for which pollen productivity estimates (PPEs) were available (Mazier et al. 2012). Calluna was not separated from Ericaceae in the pollen analysis of the Kansjön sediment. Hence, we assumed that all Ericaceae pollen from Kansjön were from Calluna. We applied the PPEs, standard errors and fall speed of pollen according to the latest synthesis of PPEs in north-western Europe, by Mazier et al. (2012). PPEs from southern Sweden were used, in accordance to Broström et al. (2004), except for Cerealia (cereals), Calluna (heather), Plantagolanceolata (ribwort plantain) and Rumexacetosa-type (sorrel), for which Danish PPEs (Nielsen 2004) were used following the recommendations of Hellman et al. (2008b).
Table 2

The PPEs (pollen productivity estimates) and fall speed of pollen used in the REVEALS and LOVE reconstruction at the study sites

Taxa

Fall speed (m/s)

PPE

Standard dev.

Alnus

0.021

4.2

0.14

Betula

0.024

8.87

0.13

Calluna

0.038

1.1

0.05

Carpinus

0.042

2.53

0.07

Cerealia

0.06

0.75

0.04

Corylus

0.025

1.4

0.04

Cyperaceae

0.035

1

0.16

Fagus

0.057

6.67

0.17

Filipendula

0.006

2.48

0.82

Fraxinus

0.022

0.67

0.03

Juniperus

0.016

2.07

0.04

Picea

0.056

1.76

0

Pinus

0.031

5.66

0

Plantago lanceolata

0.029

0.9

0.23

Poaceae

0.035

1

0

Quercus

0.035

7.53

0.08

Rumex acetosa-type

0.018

1.56

0.09

Salix

0.022

1.27

0.31

Tilia

0.032

0.8

0.03

Ulmus

0.032

1.27

0.05

In our study, the REVEALS and LOVE runs were performed using the computer programmes REVEALS.v4.2.2.forpublic.fast.exe (lake version) and LOVE.v3.4.3.fast.exe (bog version), respectively (Sugita 2011 unpublished). The maximum extent of the regional vegetation (Zmax) was set to 100 km because vegetation beyond this distance was assumed to contribute <10 % of the pollen to the studied lake basins (Hellman et al. 2008a, b; Sugita et al. 2007) the minimum and maximum RSAP of the small sites were set to 50 and 5,000 m respectively and the calculation increment to 25 m. The length of the time windows (Table 3) was chosen to achieve the best temporal resolution for each site, and hence vary over time with coarser resolution in the older parts of the chronologies. A minimum value of 1 % of estimated vegetation cover was used to define the time of establishment/local presence of a taxon (note that Figs. 3, 4, 5, 6, 7 also present lower vegetation shares than 1 %). When interpreting the LOVE-estimated vegetation cover, the values with error estimates larger than the mean LOVE estimates are considered as not different from zero (following Cui et al. 2013).
Table 3

Time windows (cal. years bp) used for the REVEALS and LOVE reconstructions at the study sites, and the summed pollen counts for each site and time window

Time windows LOVE

Pollen counts Lillegölen

Pollen counts Ekenäs

Pollen counts Skärsgölarna

Time windows REVEALS

Pollen counts Trummen

Pollen counts Kansjön

0–100

1,860

429

1,686

0–200

2,626

976

100–200

2,283

455

1,193

   

200–300

2,384

442

801

200–400

2,538

1,099

300–400

2,782

469

831

   

400–800

3,654

1,655

2,497

400–800

2,583

989

800–1,200

2,042

820

1,706

800–1,200

4,820

1,151

1,200–1,500

1,167

414

1,623

1,200–1,500

2,540

1,063

1,500–2,000

1,445

879

2,404

1,500–2,000

2,330

 

2,000–2,500

1,632

426

2,282

2,000–2,500

9,377

3,246

2,500–3,000

1,624

2,059

403

2,500–3,000

9,542

1,147

3,000–3,500

771

1,274

 

3,000–3,500

10,162

1,198

3,500–4,000

654

885

 

3,500–4,000

9,688

1,114

4,000–4,500

1,294

  

4,000–4,500

7,107

1,086

4,500–5,000

616

  

4,500–5,000

9,998

1,098

5,000–6,000

1,209

  

5,000–6,000

16,998

3,237

6,000–7,000

5,376

  

6,000–7,000

16,214

2,102

7,000–8,000

2,326

  

7,000–8,000

19,598

3,101

Fig. 3

Pollen records from the large lakes Lake Trummen and Lake Kansjön: pollen percentages of the 12 most common pollen taxa in the LRA application. Note the different scales on the x axis

Fig. 4

Mean REVEALS estimates of regional plant percentage cover at Lake Trummen and Lake Kansjön; note the different scales on the x axis

Fig. 5

Pollen percentages and estimated vegetation cover at Skärsgölarna; note the different scales on the x axis

Fig. 6

Pollen percentages and estimated vegetation cover at Lillegölen; note the different scales on the x-axis. b. Relevant source area of pollen (RSAP) calculated on the basis of the pollen records from all three sites, i.e. three sites 1–2500 bp, two sites 2500–3500 bp and one site 3500–6000 bp

Fig. 7

Pollen percentages and estimated vegetation cover at Ekenäs; note the different scales on the x axis

Results

Chronologies

All three age-depth models imply that the accumulation rates are relatively constant from mid Holocene until ~1000 bp when they increase (Fig. 2). The latter is a common feature of peat-bogs accumulations in southern Sweden (Olsson et al. 2010).

Pollen percentages versus REVEALS and LOVE model-based vegetation cover

As a result of the correction of biases due to differences in pollen productivity and dispersal properties, the REVEALS and LOVE-based plant percentage-cover (henceforth called estimated cover or REVEALS estimate/LOVE estimate) is generally higher than the pollen percentages for Fagus, Tilia, Picea, Poaceae, Cerealia, Plantago, Rumex and Calluna. In contrast, it is generally lower than the pollen percentages for Betula, Pinus and Quercus. These results are in agreement with the test of the REVEALS model using modern pollen and vegetation data in southern Sweden (Hellman et al. 2008a, b) and the LOVE model using modern and historical pollen and vegetation data in central Småland (Fredh 2012; Cui 2013). However, the percentage calculation may sometimes result in either higher or lower REVEALS/LOVE estimates than the pollen percentage for the same taxon but in different levels of the stratigraphy, depending on the estimated abundance of the other taxa at the level in question. This phenomenon is seen for instance in the LOVE reconstructions of Pinus at the three Hornsö–Allgunnen sites.

Regional spatial scale—pollen percentage and REVEALS estimates of large lakes

The pollen percentages of Trummen and Kansjön are comparable (Fig. 3). However, there are generally higher percentages of Pinus and Fagus, and lower percentages of Picea, Quercus and Tilia at Trummen than at Kansjön. According to the REVEALS mean estimated vegetation cover (combining the two regional sites), there are <30 % Picea and 10–30 % Pinus (Fig. 4). For Tilia, the mean estimated cover is 5–16 % until ~2500 bp, when it decreases to a few per cent. Poaceae is increasing from ~4500 bp and reach mean estimates between 15 and 20 %. Poaceae are much more abundant than Calluna at both sites. Pollen percentages of Cerealia are low at both sites, whereas the mean estimated cover reaches 10–12 % in the topmost time windows. Also both Rumex and Plantagolanceolata have higher mean estimated cover than pollen percentages, but even so, the estimated cover never exceeds a few per cent and are mostly not different from zero when the error estimate is taken into account.

Local spatial scale—pollen percentages and LOVE estimates of small wetlands

The LOVE estimates of plant abundance for a given taxon represent the percentage cover of that taxon for an area equal to the RSAP of the small study sites. The RSAP of the three Hornsö–Allgunnen sites together varies between 650 and 4,025 m radius over time, with an average of ~2,000 m (Fig. 6b). Because the RSAP estimates are based on data from a small number of sites (one to three depending on the time window) it is not possible to interpret the variation of RSAP between time windows in terms of differences in the vegetation composition and structure. These values of RSAP imply that the LOVE-based reconstruction of vegetation cover is valid for an area of minimum 650–4,000 m.

At all three Hornsö–Allgunnen sites, the LOVE estimates (Figs. 5, 6, 7) for Pinus fluctuate around 50 % throughout the period, and the difference between pollen percentages and LOVE estimates is small. In comparison, LOVE estimates of Picea are much higher than pollen percentages. However, the error estimates are large for most of the time windows, and local presence of Picea cannot be determined until ca. 350 bp at Ekenäs (up to 32 % cover), and in modern time at Lillegölen (19 %) and Skärsgölarna (up to 11 %). The LOVE estimates of Quercus are lower than pollen percentages and show values large enough to interpret Quercus as locally present from 800 bp at Lillegölen, from 600 to 200 bp at Skärsgölarna and from 600 bp at Ekenäs. For Fagus, the LOVE estimates are somewhat lower than the pollen percentages. At all three Hornsö–Allgunnen sites, Fagus occurs in low amounts and the LOVE estimates have error estimates too large to confirm its presence. There is a major difference between pollen percentages and LOVE estimates of Tilia. Whereas pollen percentages rarely exceed a few per cent, the estimated cover reaches values up to 17 % at Lillegölen and around 10 % at the two other sites. At Lillegölen, Tilia is most abundant 3000 bp, and at the two other sites at the beginning of their respective chronologies. At both Skärsgölarna and Ekenäs, Tilia has large error estimates and cannot be considered locally present after 2000 and 1500 bp respectively, whereas it is present in Lillegölen until ca. 600 bp. Betula is largely overrepresented using pollen percentages. It is however relatively abundant locally, particularly at Lillegölen (13–37 % estimated cover 6000 bp to present) and Ekenäs (12–29 % during most of the period 3500–100 bp and somewhat lower for the last century). At Skärsgölarna the estimated cover is lower, <10 % throughout most of the studied period, except for the last century when it exceeded 10 %, however most values are not different from zero when taking the error estimates into account.

Most of the non-arboreal taxa are underrepresented using pollen percentages. Calluna occurs in low amounts both as pollen percentages and estimated cover, except at Skärsgölarna where it is characterized by LOVE estimates of 30–40 % during most of the studied period except for the last two centuries during which it constituted 50–60 % of the estimated vegetation. At Lillegölen and Ekenäs, Calluna constituted <9 and <4 % of the estimated vegetation respectively. LOVE estimates of Poaceae are notably higher than indicated by pollen percentages and it is generally not abundant in the estimated vegetation of the three sampled sites. It occurs at Lillegölen (~10 % until 4500 bp, 22 % at 2500 bp, and 8 % at 400 bp) and Ekenäs (30–45 % 1200–400 bp), while it is not present at Skärsgölarna.

The pollen percentages of the anthropogenic indicators (in this study Cerealia, Plantagolanceolata and Rumex) are low and the LOVE estimates are not significantly different from zero due to large error estimates, which could be due to low pollen counts especially for short time windows. However, the LOVE estimates values (irrespective of error estimates) are consistently higher than the pollen percentages, indicating that the latter underestimate the cover of these taxa.

Macrocharcoal

The amount of charcoal fragments differs among the Hornsö–Allgunnen sites, from a few (<2.5) fragments/ml in Lillegölen to ~1000 fragments/ml in Skärsgölarna (Fig. 8). The profile of Skärsgölarna exhibits two well-delimited periods of high macrocharcoal values, ca. 1800–1200 bp and 500–250 bp. Charcoal fragments occur from 2500 bp, i.e. from the oldest part of the chronology onwards. In the profile of Ekenäs, macrocharcoal are mainly found for the period 1700–400 bp, followed by low amounts until 250 bp. The record from Lillegölen includes only a few episodes with charcoal, at 6000, 3250 and 300–250 bp.
Fig. 8

Number of charcoal fragments/ml sediment; note the different scales on the x axis

Discussion

The study of three local sites provides a picture of past vegetation for the Hornsö–Allgunnen area at higher spatial and temporal resolutions than was available earlier (Valdemardotter 2001; Lindbladh et al. 2003). The between-site differences in vegetation are relatively large, indicating differences in habitat between the sites. However, there are many important similarities during the last 3,000 years, such as a large estimated cover of Pinus and other light-demanding taxa (e.g. Calluna, Betula and Poaceae). The comparably large share of Pinus and the late establishment of Fagus and Picea at the three sites indicate that the major characteristics of the forests in the study area also differ from the general tree species composition in the region as indicated by the REVEALS estimates.

Forest openness

Regarding past openness in the Hornsö–Allgunnen study area, large estimated proportions of Calluna in Skärsgölarna and Poaceae in Ekenäs probably characterized the field layer during several millennia, indicating a rather open landscape. However, tree taxa can also provide valuable information on openness. The shade-intolerant Pinus represents 40–60 % of the vegetation cover both at Skärsgölarna and Ekenäs—substantially more than in the region as a whole. The similarly light-dependent Betula was common in the study area (although less abundant in the surroundings of Skärsgölarna). At Lillegölen the estimated cover of Betula fluctuates around 30 % and at Ekenäs around 20 %, indicating light conditions due to the relatively little shade cast by light-demanding trees such as Betula (Messier and Bellefleur 1988). The relatively light-demanding Quercus (Diekmann 1996) seems to have a different history in Hornsö–Allgunnen than in the region as a whole, as well as in southern Scandinavia in general (Lindbladh and Foster 2010). In Lillegölen and Ekenäs, the presence of Quercus was sustained during the 19th and 20th centuries while it decreased significantly in the rest of southern Scandinavia, both according to historical documents (Eliasson and Nilsson 2002) and pollen studies (Lindbladh and Foster 2010).

Shade-tolerant tree species were on the other hand not a dominant feature in the area. Even if Tilia was probably relatively common both regionally and locally during some time windows, it was rare in Skärsgölarna during most of the surveyed time period, and in Ekenäs from ~1,500 bp onwards. The highly shade-tolerant Picea and Fagus both established late at the sampled sites. Despite the suggested importance of the latter for today’s insect fauna in the area, Fagus has not been abundant in Hornsö–Allgunnen, but could have occurred in small, scattered stands as has been suggested for other hotspot areas close to the edge of the Fagus distribution (Hannon et al. 2010).

In summary the pollen data and LRA reconstructions suggest that the Hornsö–Allgunnen area has been covered by relatively open forest during several millenia. Agriculture is known to have occurred in southern Scandinavia for 6,000 years (Berglund 1991), also in marginal areas from at least 2700 bp (Lagerås et al. 1995), which could have been a contributing factor behind the openness. Our results suggest, however, that cultivation of cereals was scarce in the studied area, but, based on the relatively large proportion of Poaceae and Calluna, grazing by domestic animals might have contributed to maintaining the openess of the forest. This is confirmed for historical times as Skärsgölarna and Lillegölen are located in a part of Hornsö–Allgunnen that belonged to a grazed outland (a common) up to the 19th century (Anonymous 2006).

Continuous fire

Even if grazing did contribute to some extent, the data from this and other studies (e.g. Niklasson 2011; Niklasson and Drakenberg 2001; Niklasson et al. 2002, 2010a) raise the question of fire as the possibly largest factor contributing to long-term openness of the forest in the Hornsö–Allgunnen area in the past. The charcoal records differ between the three small sites, with large concentrations at Skärsgölarna and smaller ones at both Ekenäs and Lillegölen, indicating local fires rather than large fires over the entire Hornsö–Allgunnen, which is consistent with tree-ring studies from the area (Niklasson et al. 2010a; Niklasson et al. unpublished). Fire intervals of 30–40 years or even shorter were documented in the area over the last 500 years by dated fire scars, which suggest human-induced fires at least for half a century (Niklasson 2011; Niklasson and Drakenberg 2001; Niklasson et al. 2002, 2010a). This is also an indication of the important role of forest fires.

Furthermore, the region is characterised by having among the lowest annual precipitations in the country (www.smhi.se2014) and the highest number of annual lightning ignitions (0.23 ignitions/10,000 ha) (Granström 1993). The recurring fires up to recent time in combination with dry soils are also possible causes of the absence of Picea in the area until the 19th century. Picea is sensitive to fire and, hence, most likely did not establish and/or expand until the fires ceased (Niklasson and Drakenberg 2001; Niklasson et al. 2010b; Ohlson et al. 2011; Tryterud 2003; Wallenius et al. 2007).

Links between forest history and present day biodiversity

Our results show that the Hornsö–Allgunnen area was dominated by taxa that are shade-intolerant and more or less fire-resistant and/or fire-prone, such as Pinus, Betula, Calluna, Quercus and Poaceae (Nilsson et al. 2006; Zackrisson 1977) for millennia. Many shade-intolerant tree species cast less shade than shade-tolerant species, both among deciduous and coniferous species (Canham and Burbank 1994; Messier and Bellefleur 1988). Hence, there are strong indications that the Pinus dominance created a different environment with regard to light and temperature in the Hornsö–Allgunnen area as compared with sites dominated by shade-tolerant tree taxa, for instance Picea (Barbier et al. 2008; Sonohat et al. 2004; Wirth et al. 1999). Thus the late establishment of Picea, both in relation to other small hollow studies in southern Sweden (Bradshaw and Lindbladh 2005) and the regional vegetation as estimated by REVEALS (Fig. 4), probably contributed to maintain the open environment until recently.

These open conditions result in high solar insolation and warm microclimate, which have been suggested reasons for the insect species richness biodiversity in the area (Anonymous 2008), but are also favourable for the species diversity of wood dependent beetles in general (Lindhe et al. 2005). Inventories in Hornsö–Allgunnen have shown that considerably more beetle species are associated with sun-exposed sites than with shaded (Dahlberg and Stokland 2004; Naturcentrum 2009; Nilsson and Huggert 2001; Andersson, personal communication, 2009). However, there are also direct connections between fire and present insect biodiversity. Among the beetle species found in the area today, 31 % (245 taxa) are indicative of fire (Dahlberg and Stokland 2004; Naturcentrum 2009; Nilsson and Huggert 2001; Andersson, personal communication, 2009). Hornsö–Allgunnen has probably offered favourable conditions for fire-dependent species due to the late cessation of forest fires compared to many other areas in southern Sweden, where fires became uncommon earlier (Niklasson and Drakenberg 2001). Many of the late fires are likely to be due to ignitions in connection with slash-and-burn cultivation which is known to have occurred in the area until the mid-19th century (Eriksson and Franzén 1969), or to management fires to improve fodder for woodland grazing (e.g. Wäglind 2004).

A dominance of Pinus in this part of south-eastern Sweden has been shown in several other studies (Björse and Bradshaw 1998; Lindbladh and Foster 2010) and a vegetation history similar to that of Hornsö–Allgunnen, including recurring fires, was described at Storasjö ~50 km west of Hornsö–Allgunnen) (Cui et al. 2013). However, Storasjö is not known to host a particularly species-rich insect fauna today, indicating that frequent fires in the past and dominance of Pinus are not the only important factors for the high entomological biodiversity in the Hornsö–Allgunnen area. Another plausible reason for the difference in insect diversity between the two areas could be the higher proportion of Quercus in Hornsö–Allgunnen compared to Storasjö. Quercus hosts by far the most insect taxa in Sweden, in particular red-listed taxa (Jonsell et al. 1998). Also, the large contemporary insect fauna associated with Fagus in Hornsö–Allgunnen may be a result of the continuous presence of Quercus over a long period; the insect communities considered as characteristic of Quercus and Fagus are overlapping (Dahlberg and Stokland 2004; Jonsell et al. 1998). Further, the climatic conditions of Hornsö–Allgunnen are slightly warmer and drier than those of Storasjö (Alexandersson and Eggertsson Karlström 2001), which might provide a more favourable microclimate for other warmth-demanding insect species than at Storasjö.

The use of LRA in biodiversity research

The main obstacles in translating pollen data to actual vegetation cover are related to interspecific differences in pollen production and dispersal characteristics among taxa. Therefore, the LRA estimates provide a more realistic picture of the true taxa composition than do pollen percentages. In this study, important differences between pollen percentages and LRA-estimated plant cover were found for non-arboreal taxa such as Calluna and Poaceae. This implies that large-scale structures such as heathlands or forested bogs that often harbour these taxa could have been overlooked using pollen percentages. Moreover, the forest might have been perceived as denser and darker than it really was due to the underestimation of important field-layer taxa such as Calluna and Poaceae. The LRA reconstruction also reveals a different history for several arboreal taxa compared to pollen percentages. Knowledge of these differences could be crucial for the understanding and management of present diversity, and shows the benefit of using LRA estimates of plant cover in addition to pollen percentages for the reconstruction of the vegetation and landscape.

In addition to Pinus, Betula and Quercus that all have had a long continuity in our study area, Populus tremula (aspen) has been pointed out as an important tree species for the present insect biodiversity, but has never been common according to our study. However, poor pollen preservation in combination with a very faint surface ornamentation of the pollen wall makes it difficult to recognize damaged Populus grains (Campbell 1999), which could suggest that it is underestimated (Fægri and Iversen 1989), and it could hence not be ruled out that it might have been a characteristic tree species in Hornsö–Allgunnen.

Conclusions and implications for conservation strategies

Several tree species, such as Pinus, Betula and Quercus were all suggested to be important for the present insect diversity of the study area. Our study demonstrates that these taxa have a long continuity in the study area and were common until recently. This continuity in tree species composition and the persistence of Quercus until recently, in combination with frequent fires and particularly mild and dry climatic conditions, created open conditions with abundant sun-exposed surfaces with a dry and warm microclimate that was, and still is, essential for many of today’s rare insect species. We conclude that the area was dominated by open forest during at least 3,000 years.

However, the area was far from unused by humans. For instance at Ekenäs, small-scale agriculture and grazing by domestic animals could have been important factors for the maintenance of open forests, and the possible existence of grazed Calluna heaths cannot be ruled out in the surroundings of Skärsgölarna. The regular findings of charcoal point to fire as a main reason behind the openness, and we stress the importance of continued prescribed burnings in the area to maintain both the open character of the forest and the substrates caused by fire as this will favour both light-demanding and fire-dependant insect species.

Our study demonstrates that the quantitative vegetation/landscape history of biodiversity hotspots can provide important insights on the factors behind present biodiversity, and makes it possible to propose better founded recommendations for its long-term maintenance.

Notes

Acknowledgments

The authors thank Formas (diary number 215-2007-645) for funding, Gina Hannon for help with plant macro-fossil analysis, Shinya Sugita for supplying us with the computer programmes to implement the REVEALS and LOVE models, and Qiao-Yu Cui for software support. We also thank Thomas Giesecke and one anonymous reviewer for valuable comments that helped us to substantially improve earlier versions of the manuscript.

Supplementary material

334_2014_469_MOESM1_ESM.pdf (2.6 mb)
Supplementary material 1 (PDF 2697 kb)

References

  1. Alexandersson H, Eggertsson Karlström C (2001) Temperaturen och nederbörden i Sverige 1961–1990. Referensnormaler—utgåva 2. SMHI, NorrköpingGoogle Scholar
  2. Andersen ST (1972) The differential pollen productivity of trees and its significance for the interpretation of a pollen diagram from a forested region. In: Birks HJ, West RG (eds) Quaternary plant ecology. Blackwell Scientific, OxfordGoogle Scholar
  3. Anonymous (2006) Beslut och skötselplan för naturreservatet Allgunnen, Högsby och Nybro kommun, Kalmar län. KalmarGoogle Scholar
  4. Anonymous (2008) Ekoparksplan Hornsö. SveaskogGoogle Scholar
  5. Anonymous (2011) Statistisk årsbok 2011Google Scholar
  6. Barbier S, Gosselin F, Balandier P (2008) Influence of tree species on understory vegetation diversity and mechanisms involved: a critical review for temperate and boreal forests. For Ecol Man 254:1–15CrossRefGoogle Scholar
  7. Berglund BE (1991) The cultural landscape during 6000 years in southern Sweden: the Ystad project. Ecological bulletin, vol 41. Munksgaard, CopenhagenGoogle Scholar
  8. Björse G, Bradshaw R (1998) 2000 years of forest dynamics in southern Sweden: suggestions for forest management. For Ecol Man 104:15–26CrossRefGoogle Scholar
  9. Björse G, Bradshaw RHW, Michelson DB (1996) Calibration of regional pollen data to construct maps of former forest types in southern Sweden. J Paleolimnol 16:67–78CrossRefGoogle Scholar
  10. Blaauw M (2010) Methods and code for ‘classical’ age-modelling of radiocarbon sequences. Quat Geochron 5:512–518CrossRefGoogle Scholar
  11. Bradshaw RHW, Lindbladh M (2005) Regional spread and stand-scale establishment of Fagus sylvatica and Picea abies in Scandinavia. Ecol 86(7):1,679–1,686CrossRefGoogle Scholar
  12. Broström A, Sugita S, Gaillard M-J (2004) Pollen productivity estimates for the reconstruction of past vegetation cover in the cultural landscape of southern Sweden. Holocene 14:368–381CrossRefGoogle Scholar
  13. Broström A, Sugita S, Gaillard M-J, Pilesjö P (2005) Estimating the spatial scale of pollen dispersal in the cultural landscape of southern Sweden. Holocene 15:252–262CrossRefGoogle Scholar
  14. Brunius J, Ferm O (1990) Handbörd och Stranda. Riksantikvarieämbetet, StockholmGoogle Scholar
  15. Bunting MJ, Gaillard M-J, Sugita S, Middleton R, Broström A (2004) Vegetation structure and pollen source area. Holocene 15:651–660CrossRefGoogle Scholar
  16. Campbell ID (1999) Quaternary pollen taphonomy: examples of differential redeposition and differential preservation. Palaeogeogr Palaeoclimatol Palaeoecol 149:245–256CrossRefGoogle Scholar
  17. Canham CD, Burbank DH (1994) Causes and consequences of resource heterogeneity in forests: interspecific variations in light transmission by canopy trees. Can J For Res 24:337–349CrossRefGoogle Scholar
  18. Cui Q (2013) Fire history in the hemiboreal and southern boreal zones of southern Sweden during 11,000 years. Relationships with past vegetation composition and human activities and implications for biodiversity issues. Linnaeus University Dissertations No 155/2013, Linnaeus University PressGoogle Scholar
  19. Cui Q, Gaillard M-J, Lemdahl G, Sugita S, Greisman A, Jacobson GL, Olsson F (2013) The role of tree composition in Holocene fire history of the hemiboreal and southern boreal zones of southern Sweden, as revealed by the application of the Landscape Reconstruction Algorithm: implications for biodiversity and climate-change issues. Holocene 23:1,745–1,761CrossRefGoogle Scholar
  20. Dahlberg A, Stokland JN (2004) Vedlevande arters krav på substrat. Skogsstyrelsen, JönköpingGoogle Scholar
  21. Davis MB (1963) On the theory of pollen analysis. Am J Sci 261:897–912CrossRefGoogle Scholar
  22. Diekmann M (1996) Ecological behaviour of deciduous hardwood trees in boreo-nemoral Sweden in relation to light and soil conditions. For Ecol Man 86:1–14CrossRefGoogle Scholar
  23. Digerfeldt G (1972) The postglacial development of Lake Trummen, Sweden. Regional vegetation history, water level changes and paleo limnology. Folia Limnol Scand 1972:1–104Google Scholar
  24. Ehnström B, Axelsson R (2002) Insektsgnag i bark och ved. Artdatabanken, SLU, UppsalaGoogle Scholar
  25. Eliasson P, Nilsson SG (2002) You should hate young oaks and young noblemen: the environmental history of oaks in eighteenth- and ninetheenth-century Sweden. Environ Hist 7:657–675Google Scholar
  26. Eriksson H, Franzén O (1969) Högsbyboken. Högsby kommun, HögsbyGoogle Scholar
  27. Fægri K, Iversen J (1989) Textbook of pollen analysis, 4th edn. The Blackburn Press, CaldwellGoogle Scholar
  28. Ferm O, Rahmqvist S, Thor L (1987) Möre. Norra och Södra Möre. Riksantikvarieämbetet, StockholmGoogle Scholar
  29. Foster DR (2002) Insights from historical geography to ecology and conservation: lessons from the New England landscape. J Biogeogr 29:1,269–1,275CrossRefGoogle Scholar
  30. Fredén C (2002) Berg och jord. Sveriges nationalatlas. Lantmäteriverket, GävleGoogle Scholar
  31. Fredh D (2012) The impact of past land-use change on high-resolution pollen data. Lund University, LundGoogle Scholar
  32. Gaillard M-J, Sugita S, Bunting J et al (2008) The use of modelling and simulation approach in reconstructing past landscapes from fossil pollen data: a review and results from the POLLANDCAL network. Veget Hist Archaeobot 17:419–443CrossRefGoogle Scholar
  33. Granström A (1993) Spatial and temporal variation in lightning ignitions in Sweden. J Veget Sci 4:737–744CrossRefGoogle Scholar
  34. Hannon G, Niklasson M, Brunet J, Eliasson P, Lindbladh M (2010) How long has the ‘hotspot’ been ‘hot’? Past stand-scale structures at Siggaboda nature reserve in southern Sweden. Biodivers Conserv 19:2,167–2,187CrossRefGoogle Scholar
  35. Hellman SEV, Gaillard M-J, Broström A, Sugita S (2008a) The REVEALS model, a new tool to estimate past regional plant abundance from pollen data in large lakes: validation in southern Sweden. J Quat Sci 23:21–42CrossRefGoogle Scholar
  36. Hellman SEV, Gaillard M-J, Broström A, Sugita S (2008b) Effects of the sampling design and selection of parameter values on pollen-based quantitative reconstructions of regional vegetation: a case study in southern Sweden using the REVEALS model. Veg Hist Archaeobot 17:445–459CrossRefGoogle Scholar
  37. Hellman SEV, Bunting MJ, Gaillard M-J (2009a) Relevant source area of pollen in patchy cultural landscapes and signals of anthropogenic landscape disturbance in the pollen record: a simulation approach. Rev Palaeobot Palynol 153:245–258CrossRefGoogle Scholar
  38. Hellman SEV, Gaillard M-J, Bunting MJ, Mazier F (2009b) Estimating the relevant source area of pollen in the past cultural landscapes of southern Sweden: a forward modelling approach. Rev Palaeobot Palynol 153:259–271CrossRefGoogle Scholar
  39. Jacobson GL, Bradshaw RHW (1981) The selection of sites for paleovegetational studies. Quat Res 16:80–96CrossRefGoogle Scholar
  40. Jonsell M, Weslien J, Ehnström B (1998) Substrate requirements of red-listed saproxylic invertebrates in Sweden. Biodivers Conserv 7:749–764CrossRefGoogle Scholar
  41. Lagerås P, Jansson K, Vestbö A (1995) Land-use history of the Axlarp area in the Småland uplands, southern Sweden: palaeoecological and archaeological investigations. Veg Hist Archaeobot 4:223–234CrossRefGoogle Scholar
  42. Lindbladh M, Foster DR (2010) Dynamics of long-lived foundation species: the history of Quercus in southern Scandinavia. J Ecol 98:1,330–1,345CrossRefGoogle Scholar
  43. Lindbladh M, Niklasson M, Nilsson SG (2003) Long-time record of fire and open canopy in a high biodiversity forest in southeast Sweden. Biol Conserv 114:231–243CrossRefGoogle Scholar
  44. Lindbladh M, Fraver S, Edvardsson J, Felton A (2013) Past forest composition, structures and processes: how paleoecology can contribute to forest conservation. Biol Conserv 168:116–127CrossRefGoogle Scholar
  45. Lindhe A, Lindelöw Å, Åsenbland N (2005) Saproxylic beetles in standing dead wood density in relation to substrate sun-exposure and diameter. Biodivers Conserv 14:3,033–3,053CrossRefGoogle Scholar
  46. Mazier F, Broström A, Gaillard M-J, Sugita S, Vittoz P, Buttler A (2008) Pollen productivity estimates and relevant source area of pollen for selected plant taxa in the pasture woodland landscape of the Jura Mountains (Switzerland). Veget Hist Archaeobot 17:479–495CrossRefGoogle Scholar
  47. Mazier F, Gaillard M-J, Kuneš P, Sugita S, Trondman A-K, Broström A (2012) Testing the effect of site selection and parameter setting on REVEALS-model estimates of plant abundance using the Czech quaternary palynological database. Rev Palaeobot Palynol 187:38–49CrossRefGoogle Scholar
  48. Messier C, Bellefleur P (1988) Light quantity and quality on the forest floor of pioneer and climax stages in a birch–beech–sugar maple stand. Can J For Res 18:615–622CrossRefGoogle Scholar
  49. Moore PD, Webb JA, Collinson ME (1991) Pollen analysis, 2nd edn. Blackwell, OxfordGoogle Scholar
  50. Naturcentrum (2009) Vedskalbaggar på brandfält i Hornsö. StenungsundGoogle Scholar
  51. Nielsen AB (2004) Modelling pollen sedimentation in Danish lakes around AD 1800—an attempt to validate the POLLSCAPE model. J Biogeogr 31:1,693–1,709CrossRefGoogle Scholar
  52. Nielsen AB, Odgaard BV (2010) Quantitative landscape dynamics in Denmark through the last three millenia based on the Landscape Reconstruction Algorithm. Veg Hist Archaeobot 19:375–387CrossRefGoogle Scholar
  53. Nielsen AB, Sugita S (2005) Estimating relevant source area of pollen for small Danish lakes around AD 1800. Holocene 15:1,006–1,020CrossRefGoogle Scholar
  54. Niklasson M (2011) Brandhistorik i sydöstra Sverige. Länsstyrelsen Kalmar län, KalmarGoogle Scholar
  55. Niklasson M, Drakenberg B (2001) A 600-year tree-ring fire history from Norra Kvills National Park, southern Sweden: implications for conservation strategies in the hemiboreal zone. Biol Conserv 101:63–71CrossRefGoogle Scholar
  56. Niklasson M, Lindbladh M, Björkman L (2002) A long-term record of Quercus decline, logging and fires in a southern Swedish Fagus-Picea forest. J Veg Sci 13:765–774Google Scholar
  57. Niklasson M, Drobyshev I, Zielonka T (2010a) A 400-year history of fires on islands in south-east Sweden. Int J Wildland Fire 19:1,050–1,058CrossRefGoogle Scholar
  58. Niklasson M, Zin E, Zielonka T, Feijen M, Korczyk AF, Churski M, Samojlik T, Jedrzejewska B, Gutowski JM, Brzeziecki B (2010b) A 350-year tree-ring fire record from Bialowieza Primeval Forest, Poland: implications for Central European lowland fire history. J Ecol 98:1,319–1,329CrossRefGoogle Scholar
  59. Nilsson SG, Huggert L (2001) Vedinsektsfaunan i Hornsö - Allgunnenområdet i östra Småland. Länsstyrelsen Kalmar län, KalmarGoogle Scholar
  60. Nilsson L, Isendahl P, Nilsson Eriksson B (1995) Odlingslandskapet i Kalmar län - bevarandeprogram. Högsby kommun. Länsstyrelsen i Kalmar län, KalmarGoogle Scholar
  61. Nilsson SG, Niklasson M, Hedin J, Eliasson P, Ljungberg H (2006) Biodiversity and sustainable forestry in changing landscapes-principles and southern Sweden as an example. J Sustain For 21:11–43CrossRefGoogle Scholar
  62. Ohlson M, Brown KJ, Birks HJB et al (2011) Invasion of Norway spruce diversifies the fire regime in boreal European forests. J Ecol 99:395–403Google Scholar
  63. Olsson F, Gaillard M-J, Lemdahl G et al (2010) A continuous record of fire covering the last 10,500 calendar years from southern Sweden: the role of climate and human activities. Palaeogeogr Palaeoclimatol Palaeoecol 291:128–141CrossRefGoogle Scholar
  64. Overballe-Petersen MV, Bradshaw RHW (2011) The selection of small forest hollows for pollen analysis in boreal and temperate forest regions. Palynology 35:146–153CrossRefGoogle Scholar
  65. Reille M (1992) Pollen et spores d’Europe et d’Afrique du Nord. Laboratoire de Botanique Historique et Palynologie, MarseilleGoogle Scholar
  66. SGU (2009a) Cartographic material, map of bedrock. SGU (Geological Survey of Sweden), UppsalaGoogle Scholar
  67. SGU (2009b) Cartographic material, map of soil types. SGU (Geological Survey of Sweden), UppsalaGoogle Scholar
  68. Sonohat G, Balandier P, Ruchaud F (2004) Predicting solar radiation transmittance in the understory of even-aged coniferous stands in temperate forests. Ann For Sci 61:629–641CrossRefGoogle Scholar
  69. Speight MCD (1989) Saproxylic invertebrates and their conservation. Council of Europe, StrasbourgGoogle Scholar
  70. Sugita S (1994) Pollen representation of vegetation in quaternary sediments: theory and method in patchy vegetation. J Ecol 82:881–897CrossRefGoogle Scholar
  71. Sugita S (2007a) Theory of quantitative reconstruction of vegetation I: pollen from large sites REVEALS regional vegetation composition. Holocene 17:229CrossRefGoogle Scholar
  72. Sugita S (2007b) Theory of quantitative reconstruction of vegetation II: all you need is LOVE. Holocene 17:243–257CrossRefGoogle Scholar
  73. Sugita S, Galliard MJ, Hellman S, Broström A (2007) Model-based reconstruction of vegetation and landscape using fossil pollen. Proceedings of the 35th international conference on computer applications and quantitative methods in archaeology (CAA), BerlinGoogle Scholar
  74. Sugita S, Parshall T, Calcote R, Walker K (2010) Testing the Landscape Reconstruction Algorithm for spatially explicit reconstruction of vegetation in northern Michigan and Wisconsin. Quat Res 74:289–300CrossRefGoogle Scholar
  75. Tryterud E (2003) Forest fire history in Norway: from fire-disturbed pine forests to fire-free spruce forests. Ecography 26:161–170CrossRefGoogle Scholar
  76. Valdemardotter Å (2001) Vegetation development and fire history in a long term perspective at Ekenäs in the Hornsö area. Southern Swedish Forest Research Centre, Swedish Agricultural University, AlnarpGoogle Scholar
  77. Wäglind J (2004) En översiktlig brandhistorisk analys av Storasjöområdets naturreservat, Kronobergs län. MSc Thesis M16, University of Kalmar, KalmarGoogle Scholar
  78. Wallenius TH, Lilja S, Kuuluvainen T (2007) Fire history and tree species composition in managed Picea abies stands in southern Finland: implications for restoration. For Ecol Man 250:89–95CrossRefGoogle Scholar
  79. Wardenaar ECP (1987) A new hand tool for cutting peat profiles. Can J Bot 65:1,772–1,773CrossRefGoogle Scholar
  80. Webb T, Howe SE, Bradshaw RHW, Heide KM (1981) Estimating plant abundances from pollen percentages: the use of regression analysis. Rev Palaeobot Palynol 34:269–300CrossRefGoogle Scholar
  81. Willis KJ, Birks HJB (2006) What is natural? The need for a long-term perspective in biodiversity conservation. Science 314:1,261–1,265CrossRefGoogle Scholar
  82. Willis KJ, Araujo MB, Bennett KD, Figueroa-Rangel B, Froyd CA, Myers N (2007) How can a knowledge of the past help to conserve the future? Biodiversity conservation and the relevance of long-term ecological studies. Phil Trans R Soc B 362:175–186CrossRefGoogle Scholar
  83. Wirth C, Schulze E-D, Schulze W et al (1999) Above-ground biomass and structure of pristine siberian scots pine forests as controlled by competition and fire. Oecologia 121:66–80CrossRefGoogle Scholar
  84. www.raa.se (2011) Swedish National Heritage Board web page. Accessed 14 April 2011
  85. www.sgu.se (2010) Geological Survey of Sweden web page. Accessed 5 Oct 2010
  86. www.smhi.se (2014) Swedish Meteorological and Hydrological Institute web page. Accessed 12 Jan 2014
  87. www.sna.se (2010) National Atlas of Sweden web page. Accessed 5 Oct 2010
  88. Zackrisson O (1977) Influence of forest fires on north Swedish boreal forest. Oikos 29:22–32CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • Tove Hultberg
    • 1
  • Marie-José Gaillard
    • 2
  • Britt Grundmann
    • 3
  • Matts Lindbladh
    • 1
  1. 1.Swedish University of Agricultural Sciences, Southern Swedish Forest Research CentreAlnarpSweden
  2. 2.Department of Biology and Environmental ScienceLinnaeus UniversityKalmarSweden
  3. 3.Department of Biology and Environmental ScienceTechnische Universität DresdenTharandtGermany

Personalised recommendations