SEARCH

SEARCH BY CITATION

Keywords:

  • AMS 14C dating;
  • deforestation;
  • fire intervals;
  • palaeoecology;
  • wood charcoal

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

1 If, within a vegetation type, fire regimes are climate dependent, then fire patterns should be synchronous at regional scales. If they are not synchronous, then fires may be dependent on local processes such as human-induced disturbances.

2 Two fire chronologies were developed using 34-radiocarbon dating measured by accelerator mass spectrometry (AMS) from wood charcoal buried in soil. These charcoal fragments were sampled in two study sites 10 km apart in the Maurienne valley (Southern Vanoise massif, Savoy, France).

3 Asynchronous temporal fire patterns were seen at Aussois and Saint-Michel-de-Maurienne. This demonstrates the dependence of fires on local- or stand-scale environmental forcing; any direct relationship with climate is therefore rejected.

4 Slash-and-burn practices are probably the main source of Holocene fires in the Maurienne valley. However, deforestation did not occur throughout a site in any period, nor simultaneously at the same elevation in two different sites 10 km apart. The cultural landscape was shaped as early as the Neolithic and the Bronze Age, between 6000 and 3000 bp.

5 Deforestation at both study sites probably occurred in many stages in many small areas. The fire intervals were c. 500–1000 years. Deforested areas increased in extent over 2000–4000 years, until the present-day cultural landscape was established. This process stopped c. 2500 bp at Saint-Michel-de-Maurienne but is still active at Aussois.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

In the forests of the European Alps, human activities are the main disturbances influencing present-day vegetation patterns (Stern 1983). Since they may mask the impact of global climate changes in the Late Holocene palaeoecological records, they are a potential source of confusion for research attempting to forecast future climates from the reconstruction of past changes (Pons 1995).

In some areas of North America, Australia and the Mediterranean basin, the current natural fire regime appears to be climate dependent (Clark 1989; Tarrega & Luis-Calabuig 1990; Johnson 1992; Whelan 1995). If such an effect can be observed globally during the Holocene, then fire patterns would be synchronous at regional scales, but if not then fire regimes might be more dependent on local processes such as human-caused disturbances (Clark & Royall 1995, 1996).

In the Alps, fire appears to have been the major process causing deforestation during the Holocene (Burga 1976; Thinon 1992; Vorren et al. 1993; Wick 1994; Tinner et al. 1996) although little is known about its importance and the consequences on vegetation patterns. Palaeoenvironmental studies in both northern and southern Europe show that slash-and-burn agriculture has been an important cultivation method since the Neolithic (Iversen 1949; Vuorela 1970, 1986; Tolonen 1978; Huttunen 1980; Pons & Thinon 1987; Behre 1988; Birks et al. 1988; Clark et al. 1989; Berglund 1991; Litt 1992; Carcaillet et al. 1997). Fire frequency may therefore have a key role in explaining the spatial pattern of plant communities and long-term vegetation dynamics at the landscape level (Zackrisson 1977; Delarze et al. 1992; Bradshaw 1993; Tatoni et al. 1994; Björkman & Bradshaw 1996; Zackrisson et al. 1996).

Charcoal fragments are not time-stratified in alpine and subalpine soils in the Alps above 1700 m a.s.l., instead they show a stochastic distribution with soil depth (Carcaillet 1996). Independent stratification methods are thus necessary for the palaeoenvironmental analysis of such sediment. The statistical weight method (Hillaire-Marcel & Occhietti 1977) can be used to establish Holocene fire chronologies and to pinpoint, in space and time, events such as sea-level oscillations (Occhietti & Hillaire-Marcel 1977), sand dune development (Filion 1984; Desponts & Payette 1993), peat development (Mathieu et al. 1987; Bussière et al. 1996), fire chronologies (Payette & Gagnon 1985; Mathieu et al. 1987; Millet & Payette 1987; Payette & Morneau 1993), tree-line movement and distribution (Gagnon & Payette 1981; Payette & Gagnon 1985; Lavoie & Payette 1996) and vegetation change (Filion & Quinty 1993; Bhiry & Filion 1996).

Charcoal fragments, found buried in soils, are a result of fires in woody vegetation that took place during the Holocene (Jacquiot et al. 1973a; Thinon 1978) or the Pleistocene (van Vliet-Lanoë 1988; Hopkins et al. 1993). If charcoal fragments are not immediately buried after fire, they are exposed to strong erosion-fragmentation processes through transportation, and thus accumulate in lowland areas, depressions, humid zones and torrential deposits (Clark 1988; Meyer et al. 1992; Thinon 1992). Since 14C-dating requires charcoal fragments of at least 1 mm in size their aerial transport more than a few hundred metres from the source area is unlikely (Wein et al. 1987; Clark 1988; Thinon 1992; Earle et al. 1996; Whitlock & Millspaugh 1996). Even if large particles c. 1 mm size are transported away from a stand they will be exposed to intense fragmentation by frost–thaw and biological activities (roots, earthworms), particularly if buried for a long time (Carcaillet & Talon 1996), and for this reason they will not be dated by 14C. If soil sampling follows the methods suggested by Carcaillet & Thinon (1996), charcoal fragments from one soil profile give a reliable pattern of an area from a few hundred to a few thousand m2 (Thinon 1992; Carcaillet 1996).

In this paper, long-term records of fire occurrence were compared for two study areas situated higher than 1700 m a.s.l. in the Western Alps, using 34-radiocarbon dating by accelerator mass spectrometry (AMS) of wood charcoal fragments extracted from soil. The relative importance of environmental forcing, climate and human impact as factors controlling fire occurrence and frequency is discussed.

Study area

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

The two study sites, Aussois and Saint-Michel-de-Maurienne, are located in the upper Maurienne valley, 10 km apart. This valley delimits the southern portion of the Vanoise and Grand Arc massifs, in the Northern French Alps (Fig. 1). Climatic data were obtained from Saint-Michel-de-Maurienne (1360 m a.s.l.) and Aussois (1490 m a.s.l.) weather stations, both with southern exposure (Fournier 1985): the mean annual temperatures were, respectively, 7.0 and 6.2 °C; the mean temperature for the coldest month (January) was −0.7 and −3.2 °C, and that of the warmest month (July) 15.0 and 15.2 °C. The mean annual snow cover was 3–4 months per year at 1400 m a.s.l. on southern slopes. Mean monthly precipitation was 79 (±14) and 59 (±10) mm at Saint-Michel-de-Maurienne and Aussois, respectively. The High Maurienne valley is one of the driest in the Alps; precipitation, temperature and its intra-annual variability and mean annual amplitude are similar to the European continental climate (Fournier 1985).

image

Figure 1. Location of the two study sites in the Northern French Alps.

Download figure to PowerPoint

Above 1500 m, the vegetation in the Vanoise massif at the beginning of the Holocene (8000–5000 bp) was conifer-dominated, with Abies alba, Pinus sylvestris, Pinus cembra forests below 1900 m a.s.l., and Pinus cembra woodlands above 1900 m a.s.l. (David 1995a, 1995b; Carcaillet 1996; Carcaillet & Thinon 1996; Talon et al. 1998). Between 4300 and 3000 bp, the upper tree-line was located c. 2700 m a.s.l. and was composed of Pinus cembra, Pinus cf. uncinata and Larix decidua (Carcaillet 1996; Carcaillet et al. in press). The vegetation in this valley is highly susceptible to natural fires, because of the continental climate and the domination by conifers.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

Sampling

The pattern of burning at various elevations during the Holocene was studied using a number of precisely located sites (Carcaillet 1997). Ten soil samples were collected along an altitudinal transect established at Aussois and 15 at Saint-Michel-de-Maurienne, between 1700 and 2700 m a.s.l. (Fig. 2). The steep and rocky relief and the scarcity of deep soil do not allow sampling above 2700 m a.s.l. Sampled stands were mesic or xero-mesic and were located far from humid zones (peat, pond, riverside). All soils from Saint-Michel-de-Maurienne and those above 2100 m a.s.l. at Aussois were brown acidic on permocarboniferous schist and sandstone. Below 2100 m a.s.l. at Aussois, the sampled soils were calcareous, situated on Würmian moraines composed of gypsum and hard limestone particles. Soil depth is correlated with the slope and the occurrence of moraines but not with altitude, and varies between 60 and 140 cm above bedrock. Stands disturbed by human activities were avoided, as were soils located at the foot of long slopes, and steep, eroded or hydromorphic soils (Carcaillet & Thinon 1996). Sampling was conducted at different depths (10–15 l of dry soil per c. 20 cm in depth) in a pedological trench dug down to the bedrock whenever possible. Each complete soil sample weighed 30–100 kg according to the depth and the organic and sand richness of the soil.

image

Figure 2. Distribution of soil wood charcoal profiles on the altitudinal transects at Saint-Michel-de-Maurienne and Aussois. Underlined profiles did not provide enough big wood charcoal to be dated by AMS.

Download figure to PowerPoint

Laboratory techniques

A flotation process was used to extract charcoal from the total soil sample (Thinon 1992). Dry wood charcoals have a much lower density than water, compared to other mineral soil elements. Flotation, with or without an ascending water current, together with water sieving followed by a manual sorting under a binocular microscope, allowed the separation of charcoal fragments from other soil particles (Carcaillet & Thinon 1996).

Charcoal particles were identified using an incident light microscope equipped with an interference contrast (×200, ×500, ×1000), in reference to published descriptions of wood and charcoal anatomy (Jacquiot 1955; Greguss 1959; Jacquiot et al. 1973b; Queiroz & van der Burgh 1989; Schweingruber 1978, 1990; Anagmost et al. 1994; Talon 1997) and the charred wood reference collection of the Mediterranean Institute of Ecology and Palaeoecology (CNRS and University of Aix-Marseille III, France).

Dating of fire occurence with AMS 14C

The method of Hillaire-Marcel & Occhietti (1977) was applied to determine the most probable date for fire occurrence using the 14C dating of wood charcoal. Charcoal fragments were selected according to their size (>1 mm) and their weight (>1 mg) to avoid small particles that could have been transported for more than a few hundred metres from the source area. Each date represents a mean value for a single charcoal fragment (Carcaillet 1996). The data for each fragment can be represented by a histogram of probabilities with a normal distribution and standard deviation, σ (Olsson 1986), as in Fig. 3. According to Gagnon & Payette (1981), the base of this histogram is the double of the mean standard deviation of 34 measurements, i.e. in this case c. 200 years. The 14C age of a wood charcoal fragment does not necessarily represent the date of the fire, and may need to be adjusted for trees that are a few hundred to thousand years old. In the Alps, it is indeed not rare to observe a number of old living trees, dead stumps or subfossil trunks with more than 500 tree-rings, especially in quite dry area (Serre-Bachet 1978, 1986; Tessier 1986; Schär & Schweingruber 1988; Édouard et al. 1991; Belingard 1996), and most dates in this study were obtained from charcoal fragments of long-lived species such as Pinus cembra, Pinus sylvestris/uncinata and Abies alba (Tables 1 and 2). However, since the probability that the charcoal fragments were derived from the centre of a few hundred-year-old trees is low, it was postulated that the date of the fire lies within the 2 σ range. To ensure that the measured date did encompass the date of the fire, the base of the dating histogram was considered to be quadruple that of the mean standard deviation, i.e. c. 400 years. A statistical weight was computed for each class interval (57 year class) from probability tables for each date (Fig. 3), and the sum was computed for all dates for each altitudinal range.

image

Figure 3. Histogram of probability frequencies of 14C dating. Values from probability tables for the normal distribution are corrected so that the sum of statistical weights is 1 (according to Gagnon & Payette 1981).

Download figure to PowerPoint

Table 1.  AMS radiocarbon dates of wood charcoal in soil profiles from Saint-Michel-de-Maurienne (Savoy, France) 1700–2600 m elevation. Asterisks indicate 14C measurements corresponding to the dates of fires according to the type of charcoal fragment: complete transverse sections with <10 tree rings
Profiles, levels andelevation (m a.s.l.)Charcoal fragment14C age(years bp)Calibrated age (years bc) andprehistoric civilizationsSample code
MAUR 1 A (2360 m)Pinus cembra3415 ± 851940–1520 Bronze AgeAA-12483
MAUR 1 C (2360 m)Pinus cembra3745 ± 752457–1950 Bronze AgeAA-12484
MAUR 4 B (1960 m)Pinus cembra4975 ± 703970–3640 Oldest NeolithicAA-12487
MAUR 4 D (1960 m)Pinus cembra3885 ± 702575–2143 Oldest NeolithicAA-12486
MAUR 6 D (1750 m)Pinus sylvestris/uncinata4980 ± 853990–3546 Oldest NeolithicAA-14775
MAUR 6 F (1750 m)Abies alba5590 ± 655655–4340 Oldest NeolithicAA-14776
MAUR 7 B (1960 m)Alnus viridis2670 ± 55920–720 Bronze/Iron AgeAA-16496
MAUR 7 D (1960 m)Pinus cembra5125 ± 504035–3788 Oldest NeolithicAA-16497
MAUR 8 D (1960 m)Pinus cembra3970 ± 452588–2397 Late NeolithicAA-16495
MAUR 8 F (1960 m)Pinus cembra4495 ± 553360–2930 Middle NeolithicAA-16494
MAUR 9 A (2580 m)Bark5115 ± 245*4460–3370 Oldest NeolithicAA-19115
MAUR 10 A (2510 m)Pinus cembra3680 ± 952451–1781 Bronze AgeAA-19116
MAUR 11 A (2450 m)Pinus cembra3280 ± 801750–1410 Bronze AgeAA-19117
MAUR 12 A (2400 m)Arctostaphyllos uva-ursi6715 ± 95*5750–5428 MesolithicAA-19118
MAUR 12 C (2400 m)Pinus cembra3245 ± 80*1736–1400 Bronze AgeAA-19119
MAUR 13 A (1700 m)Rhododendron3955 ± 652570–2410 Bronze AgeAA-20475
MAUR 13 D (1700 m)Abies alba3580 ± 652130–1742 Bronze AgeLyon-302 (OxA)
MAUR 13 G (1700 m)Abies alba3800 ± 652450–2150 Bronze AgeAA-20476
Table 2.  AMS 14C datings of wood charcoal in soil profiles from Aussois (Savoy, France) 1700–2400 m a.s.l. Asterisks indicate 14C measurements corresponding to dates of fires according to the type of charcoal fragment: transverse sections with <10 tree-rings
Profiles, levels andelevation (m a.s.l.)Type of charcoal fragment14C age(years bp)Calibrated age (years bc/ad)and prehistoric civilizationsSample code
AUSSOIS 1 D (1750 m)Abies alba1025 ± 100900–1150 ad Middle AgeAA-20464
AUSSOIS 1 E (1750 m)Pinus sylvestris/uncinata210 ± 1001550–1810 ad ModernAA-20465
AUSSOIS 2 A (1890 m)Pinus sylvestris/uncinata1230 ± 50689–941 ad Middle AgeLyon-298 (OxA)
AUSSOIS 2 G (1890 m)Pinus sylvestris/uncinata1095 ± 100830–1100 ad Middle AgeAA-20466
AUSSOIS 3 B (1890 m)Pinus sylvestris/uncinata1700 ± 50246–481 ad Roman EmpireLyon-299 (OxA)
AUSSOIS 3 D (1890 m)Juniperus1730 ± 50*144–420 ad Roman EmpireAA-16493
AUSSOIS 4 A (1890 m)cf. Picea75 ± 551810–1930 ad ModernAA-20467
AUSSOIS 4 F (1890 m)Pinus sylvestris/uncinata405 ± 551450–1630 ad ModernAA-20468
AUSSOIS 5 C (2050 m)Pinus sylvestris/uncinata4665 ± 60*3612–3187 bc NeolithicLyon-300 (OxA)
AUSSOIS 5 D (2050 m)Arctostaphyllos uva-ursi4530 ± 603372–2994 bc NeolithicLyon-301 (Oxa)
AUSSOIS 6 B (2180 m)Erica herbacea3170 ± 115*1610–1330 bc Bronze AgeAA-20469
AUSSOIS 6 C (2180 m)Unidentified angiosperm2330 ± 110 510–270 bc Iron AgeAA-20470
AUSSOIS 7 A (2340 m)Pinus sylvestris/uncinata3340 ± 65*2450–2130 bc NeolithicAA-20472
AUSSOIS 7 A (2340 m)Arctostaphyllos uva-ursi3845 ± 751750–1530 bc Bronze AgeAA-20471
AUSSOIS 8 A (2380 m)Pinus cembra3605 ± 652110–1910 bc Bronze AgeAA-20473
AUSSOIS 8 C (2380 m)Pinus cembra4175 ± 1052910–2610 bc NeolithicAA-20474

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

Eighteen AMS 14C dates were available at Saint-Michel-de-Maurienne (Table 1). These measurements spanned 4000 years, between 6715 ± 95 bp and 2670 ± 55 bp. At Aussois, 16 dates spanned a period between 4665 ± 60 bp and 75 ± 55 bp (Table 2).

Saint-Michel-De-Maurienne chronology

Histograms of statistical weight for fire dates at Saint-Michel-de-Maurienne show several fire phases (Fig. 4). Two main phases (5220–4920 bp and 4140–3720 bp) and three less important phases (5820–4520 bp, 4680–4380 bp and 2880–2580 bp) were recognizable in the mid-elevation range (1700–2000 m; Fig. 4). At high elevations (>2000 m), four phases became evident: two major (3900–3660 bp and 3540–3240 bp) and two minor (6900–6600 bp and 5340–5040 bp). It seems that no fires occurred after 2500 bp. Therefore, the Saint-Michel-de-Maurienne chronology is characterized by a Mid-Holocene period of fires that occurred between 6900 and 2500 years bp, with a concentration of fires between 5300 and 3000 years bp.

image

Figure 4. Histograms of cumulative statistical weight of 18 AMS 14C datings of wood charcoal from Saint-Michel-de-Maurienne (eleven dates from 1700 to 2000 m a.s.l. and seven dates from 2000 to 2700 m a.s.l.).

Download figure to PowerPoint

Aussois chronology

At high altitudes, 2000–2500 m a.s.l., as at Saint-Michel-de-Maurienne, fire dates at Aussois were concentrated during the Mid- to Late-Holocene period, 4800–2200 bp. No measurements were older than 4800 bp or younger than 2000 bp (Fig. 5). However, at mid-altitudes (1700–2000 m a.s.l.), three fire phases were emphasized and these occurred during the last two millennia, during the Late Holocene (Fig. 5). Fire occurrence at Aussois was not synchronous along the altitudinal gradient. Moreover, at mid-altitudes, Figs 4 and 5 show that fire occurences were clearly not synchronous at the two sites despite the short distance (10 km) between them.

image

Figure 5. Histograms of cumulative statistical weight of 16 AMS 14C dates of wood charcoal from Aussois (eight dates from 1700 to 2000 m a.s.l. and eight dates from 2000 to 2500 m a.s.l.).

Download figure to PowerPoint

Fire intervals

Figures 6 and 7 show the detailed altitudinal distribution of dates. Evidence was obtained that fires occurred at different times in each stand. The interval between two fires (Table 3) is a characteristic of a given stand and is not an artifact due to the compilation of dates per elevation range and per site (Figs 4 and 5). The time interval between fire phases for a given site and a given elevation range are indicators of the time necessary for woody vegetation to recover and to build up fuel for a new fire event (Table 3). Fire intervals varied between 390 and 1520 radiocarbon years (i.e. not corrected to their solar equivalents) but most were c. 500–1000 years.

image

Figure 6. Distribution of the 18 AMS 14C datings (±2 σ) with sampling elevation at Saint-Michel-de-Maurienne.

Download figure to PowerPoint

image

Figure 7. Distribution of the 16 AMS 14C datings (±2 σ) with sampling elevation at Aussois.

Download figure to PowerPoint

Table 3.  Fire dates and standard deviations calculated from the histograms in Figs 1 and 2 (using the method proposed by Gagnon & Payette 1981). Intervals between fires are indications of the times necessary for woody vegetation to recover, fuel to build up and allow a new fire event
Saint-Michel 1700-2000 m a.s.l.Saint-Michel 2100-2700 m a.s.l.
Fire dates (year bp)Standard deviation(year bp)Interval(year)Fire dates (year bp)Standard deviation(year bp)Interval(year)
27302580-2880 33903240-3540 
  1260  390
39903720-4140 37803660-3900 
  540  1450
45304380-4680 52305040-5340 
  500  1520
50304920-522064067506600-6900 
56704520-5820    
Aussois 1700-2000 m a.s.l.Aussois 2000-2500 m a.s.l.
Fire dates (bp)Standard deviation(year bp)Interval(year)Fire dates (year bp)Standard deviation(year bp)Interval(year)
21060-480 23702220-2520 
  960  930
11701020-1320 33003120-3480 
  540  720
17101620-1920 4020 47503600-4200 4500-4800730

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

Fire regimes: human or climate dependent?

Sources of ignition and factors influencing the temporal regime of fires may be due to natural or anthropogenic factors. Natural ignition sources are mainly volcanic activity and lightning, and the natural temporal pattern of fires is controlled by moisture conditions (Clark 1989; Tarrega & Luis-Calabuig 1990; Johnson 1992; Whelan 1995; Bond & van Wilgen 1996). Fires are also more frequent when vegetation is more susceptible to ignition because of high fuel availability and continuity (Suffling 1993). Fire frequency is also dependent on the flammability of species (Trabaud 1976; Forgeard 1989; Clark 1990; Forgeard & Lebouvier 1991).

Within the Alps, there are no volcanoes, but lightning is an important component of the climate, especially in dry and continental inner valleys such as the High Maurienne (Fournier 1985). High elevation woodlands have been dominated by relatively flammable conifers since 8000 bp (David 1995a, 1995b; Carcaillet 1996; Carcaillet et al. in press). Generally, then, conditions have been favourable for natural fires within the High-Maurienne valley since the beginning of the Holocene, and this kind of disturbance should have been a key determinant of the forest mosaic. If, however, lightning and moisture control the alpine fire regime, temporal patterns should be similar, during the Holocene, for two study sites 10 km apart. The fire phases were not synchronous for the mid-elevation range of 1700–2000 m (Figs 4 and 5): all fire dates at Saint-Michel-de-Maurienne were between 6900 and 2500 bp, but at Aussois all occurred after 2000 bp. Therefore, non-climatic factors controlled the temporal and spatial pattern of fires in the High Maurienne valley at this scale of study. The results show that the wood charcoal data buried in soils are highly dependent on local- or stand-scale fire history. Consequently, a single soil sample cannot be used for a discussion on climatic forcing. The palaeoclimatic interpretations of Berli et al. (1994) and Vernet et al. (1994), who based their conclusions on a single soil profile in Southern Swiss and in Brazil, respectively, are therefore invalidated.

The absence of a large scale (inner Maurienne valley) geographical pattern determining fires at Aussois and Saint-Michel-de-Maurienne emphasizes that local processes are the main sources of ignition. Human settlement dates back to the Neolithic and the Bronze Age in the inner valleys of the Savoy such as the High-Maurienne (Bocquet 1988). Fires between 1700 and 2600 m a.s.l. at Saint-Michel-de-Maurienne and 2000–2500 m a.s.l. at Aussois occurred during the Neolithic and Bronze Age (Tables 1 and 2); archaeological evidence of these civilizations has been discovered near these two study sites (Bocquet & Prieur 1983). Therefore, most fires at Aussois and Saint-Michel-de-Maurienne probably originated from slash-and-burn agriculture. Because there is no record of Mesolithic settlement within the inner Alps in the Savoy (Bocquet 1988), the fire dating from 6715 ± 95 bp (Table 1) could be of natural origin. However, in the Italian Alps, Wick (1994) obtained evidence of fires that occurred before forest vegetation changes during the Mesolithic, and ignition could therefore have been the result of hunting activities near the upper tree-line.

Deforestation processes

At Aussois, between 1700 and 2000 m a.s.l., the absence of wood charcoal dates before 2000 bp suggests that no local deforestation by fire occurred before the colonization of the Roman Empire epoch (2000–1500 bp). Woodlands occupied this elevation range before 2000 bp, while meadows at higher elevations were used for grazing from 4700 bp. This land use interpretation conforms to the organization of the cultural landscape, which is understood to have occurred until the 19th century in this valley (Desmaris 1991): food crops and wintering villages at low elevations (<1500 m), forests for firewood and timber at middle elevations, and meadows for grazing and hay harvesting at high elevations (>2000–2200 m). Currently, a fragmented forest matrix occurs between 1600 and 2100 m a.s.l., showing occasional patches and corridors of meadows (for hay, autumn and spring grazing or ski tracks). The asynchronous pattern of fires at middle and high elevations at Aussois (Fig. 5) suggests that prehistoric human populations tried very early to optimize their territory. This hierarchical land use along an altitudinal gradient dates back to the Neolithic and the Bronze Age in this area of the Alps.

In contrast to Aussois, the Saint-Michel-de-Maurienne study site shows a synchronous pattern of fires and deforestation along the altitudinal gradient, between 1700 and 2700 m. All fires occurred between 6900 and 2500 bp. Since the 18th century, the site has been completely deforested (Desmaris 1991; Delcros 1994). There are close coincidences, first between the old phase of slash-and-burn activity (Mid-Holocene) and the deforested catchment area of Saint-Michel-de-Maurienne, and second between the recent fire phase (Late-Holocene) and the occurrence of a forested matrix at Aussois. On the basis of wood charcoal dates at Saint-Michel-de-Maurienne, it is suggested that fires before 2500 bp and extensive agriculture after this date prevented the recovery of woodland communities, or that after 2500 bp woodlands recovered in the area but there was no further deforestation using fires. Wood charcoal dates do not allow resolution of the relative importance of the two processes. Thus, it is important to use another palaeoenvironmental approaches to reconstruct the land use history since 2500 bp. This approach must be independent of soil charcoal analysis and based on pollen and plant and insect macrofossil analysis from well-stratified sediments in small-sized (< 1 ha) ponds or peat. But, peats and ponds are rare in this study area, as in a large part of the inner Alps (David 1995a,b). No pond or peat has been discovered at Aussois and only one pond at Saint-Michel-de-Maurienne (Carcaillet 1997). Palaeoecological analysis of this pond, located at 2050 m a.s.l. within the present elevation transect at Saint-Michel-de-Maurienne, is in process.

The Saint-Michel-de-Maurienne study site allows an insight into deforestation processes implemented by farmers since the Neolithic and the Bronze Age. At both middle and high elevations (Figs 4 and 6), deforestation occurred more than once and certainly involved many stand-scale fires along the altitudinal gradient during this period. After c. 2500 bp, land use was modified, with or without the recovery of woody plant communities.

At Aussois (Figs 5 and 7) between 2000 and 2500 m a.s.l., the same pattern of land use seems to have occurred as at the same elevation at Saint-Michel-de-Maurienne; this long period of 2600 years (4800–2200 bp) precedes a long period without fire. However, between 1700 and 2000 m a.s.l., fires occurred only recently at Aussois, and destruction of the forest matrix is still not complete. It seems that slash-and-burn cultivation and stock breeding was in progress at Aussois before land abandonment, which dates to the 20th century. Indeed, soil samples between 1700 and 2000 m a.s.l. were located in woodlands a few tens of metres from meadows for hay and grazing. Settlement, followed by expansion of these agricultural patches within the forest matrix, occurred after 2000 bp (Figs 5 and 7). Periods of lesser impact of agriculture, and even land abandonment, allowed woody plants to recover at the margins of agricultural patches, before they were once more cultivated.

Fire patterns and consequences for forest communities

The study of spatial and temporal pattern of fires within the Maurienne valley emphasizes that slash-and-burn deforestation of pristine forests occurred in stages during a period of 2000–4000 years depending on the site (Fig. 8). During this period, fire phases were 300–1500 years apart, with a mean interval of 500–900 years (Table 3). These intervals are longer than fire intervals reported in Europe by Tolonen (1978), Clark et al. (1989) and Delarze et al. (1992), but shorter than intervals hypothesized by Chandler et al. (1983). It is important to consider that between any two fires, the stand was likely to be have been used for grazing or hay-making and this increases the recovery time of woody plants depending on the duration and the intensity of the land use. Moreover, at high elevations, species dominating woodlands are long-lived: Abies alba, Larix decidua, Picea abies, Pinus cembra and Pinus uncinata (Serre-Bachet 1978; Tessier 1986; Schär & Schweingruber 1988; Édouard et al. 1991; Belingard 1996). Some species, such as Pinus cembra, show a very slow rate of growth during the first 30–50 years (Contini & Lavarelo 1982). Growth of woody plants to reach an old-growth forest stage with high fuel stock is therefore always slower at high elevations than at low elevations.

image

Figure 8. Model of changes in woody vegetation following slash-and-burn activity. A = Initial stage. B and C = Microscale distributed stage on small areas, with a possible return to the initial stage after 300–1500 years; the recovery of deforested patches is possible for all species. D = Large-scale distributed stage; the return-time to previous stages involves catastrophic economic, sociological or demographic processes in human society. The recovery of the area occurs through long-distance scattering of diaspores; the composition of communities depends on which species have been maintained in the landscape on the margins of fields for fodder (Fraxinus excelsior, Acer, Ulmus, Betula) and human food (Prunus spp.) or in riparian woodlands (Alnus, Fraxinus excelsior, Betula, Acer).

Download figure to PowerPoint

If an area was not completely deforested, but rather formed a mosaic of woodlands and agricultural patches, regeneration of woody plants may have been limited by the short distance dispersal of diaspores. At the end of such a process, an agricultural matrix within the primary forested matrix would result. In such a case, forest dynamics are inhibited and fires are no longer necessary deforestation agents. If a short period of economic, demographic or social catastrophe occurs in human communities (Emanuelsson 1988), woody plants can re-establish but require long distance dispersal of seeds due to the fragmentation and the isolation of woodlands.

The expected plant dynamics after land abandonment of a completely deforested area would favour species capable of long-distance dispersal of seeds by wing (Pinus sylvestris, Larix decidua, Acer spp.). This would be combined with species remnants of the agricultural matrix, especially on the margins of the fields for fodder (Fraxinus excelsior, Acer, Ulmus, Betula) and human food (Prunus spp.) (Delcros 1994). This post-land abandonment succession is necessarily different from post-fire succession observed in areas dominated by primary coniferous forests.

Natural fire disturbance during the Holocene seems to be a negligible factor in organizing the landscape mosaic in the High Maurienne valley. This differs from observations in the Scandes mountains in Sweden (Zackrisson 1977; Engelmark 1984, 1987; Steijen & Zackrisson 1987; Bradshaw & Zackrisson 1990; Engelmark et al. 1994; Zackrisson et al. 1996) and in the Rocky Mountains of the USA (Billings 1969; Mehringer et al. 1977; Romme 1982; Romme & Knight 1982; Baker 1992; Suffling 1993; White & Vankat 1993; Turner et al. 1994; Veblen et al. 1994; Romme et al. 1995). The processes that determine natural landscape mosaics cannot be directly transferred from similar landscapes with the same climate or topographic constraints simply because similar conifer forests dominate them.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

By using the present approach, based on large charcoal particles buried in soil, it can be shown that temporal and spatial patterns of fires within the Maurienne valley are controlled by local processes; any direct relationship with climate is rejected. Slash-and-burn practices since the Neolithic (c. 6000 bp) are the probable origin of fires at elevations between 1700 and 2700 m. Since the Neolithic, cultural landscapes have changed in a step-wise fashion. The first stage corresponds to a deforestation/land use/land abandonment cyclic phase on small areas during a 2000–4000 years period. The second stage is the maintenance of an agricultural landscape with open fields over a long period. At Saint-Michel-de-Maurienne, the entire altitudinal gradient >1700 m a.s.l. was devoted to agriculture, but at Aussois middle elevations were reserved for forests and high elevations for meadows. Since 2000 bp, the forest between 1700 and 2000 m a.s.l. at Aussois has changed through settlement and expansion of small agricultural patches within the pristine forest matrix.

Aknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References

This research was funded by the French Environmental Office (Ministère de l’Environnement, comité EGPN), programmes DRAEI 92184, DRAEI 93019 and DGAD 95227, and by a PhD thesis grant from the French National Education Office (Ministère de l’Éducation Nationale, de l’Enseignement Supérieur et de la Recherche). The author thanks D. Kneeshaw (UQAM, Québec), B. E. McLaren (Wildlife Division, St John’s, New Founland), P. J. H. Richard (U. Montreal, Québec), P. Roche, B. Talon and M. Thinon (CNRS-IMEP, Marseille, France) for their advice and comments on the manuscript. The author is also grateful to M. Barbero, J.-L. de Beaulieu (CNRS-IMEP, Marseille, France), J.-J. Brun (Cemagref, Grenoble, France), S. Payette (CEN U. Laval, Québec), P. Trehen (U. Rennes 1, France) and B. van Vliet-Lanoë (CNRS-Géosciences, Rennes, France). The author expresses his appreciation to L. Haddon, P.D. Moore and two anonymous referees for reading and commenting on the manuscript and for improving the English. The 14C datings by AMS were carried out at the NSF Arizona Facility Laboratory (USA) and at the Centre de Datation par le Radiocarbone (Lyon, France).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Study area
  5. Methods
  6. Results
  7. Discussion
  8. Conclusion
  9. Aknowledgements
  10. References
  • Anagmost, S.E., Meyer, R.W. & De Zeeuw, C. 1994 Confirmation and significance of Bartholin's method for the identification of the wood of Picea and Larix. IAWA Journal, 15, 171184.
  • Baker, W.L. 1992 Structure, disturbance and change in the Bristlecone pine forests of Colorado, USA. Arctic and Alpine Research, 24, 1726.
  • Behre, K.-E. 1988 The role of man in European vegetation history. Vegetation History (eds B.Huntley & T.Webb, III), pp. 633672. Kluwer Academic Publishers, Dordrecht.
  • Belingard, C. 1996Étude Dendroécologique de la Dynamique de la Limite Supérieure de la Forêt Dans les Alpes Du Sud, En Relation Avec les Facteurs Climatique et Anthropique. PhD Thesis, University of Aix-Marseille III, Marseille, France.
  • Berglund, B.E, ed.. 1991 The Cultural Landscape During 6000 Years in Southern Sweden. Ecological Bulletins 41.
  • Berli, S., Cherubini, P. & Schoch, W. 1994 Rekonstruktion von Bestandesfluktuationen, Bodenmächtigkeit und Feuergeschichte über 7, 000 Jahre BP mittels Holzkohle-Analysen. Botanica Helvetica, 104, 1730.
  • Bhiry, N. & Filion, L. 1996 Mid-Holocene Hemlock decline in Eastern North America linked with phytophagous insect activity. Quaternary Research, 45, 312320.
  • Billings, W.D. 1969 Vegetational pattern near alpine timberline as affected by fire–snowdrift interactions. Vegetatio, 19, 192207.
  • Birks, H.H., Birks, H.J.B., Kaland, P.E. & Moe, D. 1988The Cultural Landscape, Past, Present and Future. Cambridge University Press, Cambridge.
  • Björkman, L. & Bradshaw, R. 1996 The immigration of Fagus sylvatica L. ans Picea abies (L.) Karst. into a natural forest stand in southern Sweden during the last 2000 years. Journal of Biogeography, 23, 235244.
  • Bocquet, A. 1988 La préhistoire. Histoire de la Savoie (ed. P.Guichonet), pp. 1557. Édition Privat. Coll. Univers de la France, Toulouse.
  • Bocquet, A. & Prieur, J. 1983 La Préhistoire et le peuplement de la Savoie. La Savoie Des Origines À L’an Mil (ed. P.Guichonet), pp. 159. Ouest-France, Rennes.
  • Bond, W.J. & Van Wilgen, B.W. 1996Fire and Plants. Chapman & Hall, London.
  • Bradshaw, R.H.W. 1993 Tree species dynamics and disturbance in three Swedish boreal forest stands during the last two thousand years. Journal of Vegetation Science, 4, 759764.
  • Bradshaw, R.H.W. & Zackrisson, O. 1990 A two thousand year history of a northern Swedish boreal forest stand. Journal of Vegetation Science, 1, 519528.
  • Burga, C.A. 1976 Frühe menschliche Spuren in der subalpinen Stufe des Hinterrheins. Geographica Helvetica, 2, 9396.
  • Bussière, B., Payette, S. & Filion, L. 1996 Déboisement et entourbement des hauts sommets de Charlevoix à l’Holocène supérieur: origine des étages alpins et subalpins. Géographie Physique et Quaternaire, 50, 257269.
  • Carcaillet, C. 1996Évolution de L’organisation Spatiale Des Communautés Végétales D’altitude Depuis 7000 Ans BP Dans la Vallée de la Maurienne (Alpes de Savoie, France): une Analyse Pédoanthracologique. PhD Thesis, University of Aix-Marseille III, Marseille, France.
  • Carcaillet, C. 1997 Évolution de la structure du paysage pendant l’Holocène dans la vallée de la Maurienne (Alpes du nord-ouest): présentation d’un programme de paléoécologie du paysage. Ecologia Mediterranea, 23, 131143.
  • Carcaillet, C., Barakat, H.N., Panaǐotis, C. & Loisel, R. 1997 Fire and late Holocene expansion of Quercus ilex and Pinus pinaster in Corsica. Journal of Vegetation Science, 8, 8594.
  • Carcaillet, C. & Talon, B. 1996 Stratigraphie et datations de charbons de bois dans les Alpes: quelques aspects taphonomiques. Géographie Physique et Quaternaire, 50, 233244.
  • Carcaillet, C., Talon, B. & Barbero, M. in press Pinus cembra et incendies au cours de l’Holocène, 300 m au-dessus de la limite actuelle des arbres dans les Alpes du nord-ouest. Écologie, in press.
  • Carcaillet, C. & Thinon, M. 1996 Pedoanthracological contribution to the evolution of the upper treeline in the Maurienne Valley (North French Alps): methodology and preliminary data. Review of Palaeobotany and Palynology, 91, 399416.
  • Chandler, C., Cheney, P., Thomas, P., Trabaud, L. & Williams, D. 1983Fire in Forestry. John Wiley, New York.
  • Clark, J.S. 1988 Particle motion and the theory of charcoal analysis: source area, transport, deposition, and sampling. Quaternary Research, 30, 6780.
  • Clark, J.S. 1989 Effects of long-term water balances on fire regime, North-Western Minnesota. Journal of Ecology, 77, 9891004.
  • Clark, J.S. 1990 Twentieth-century climate change, fire suppression, and forest production and decomposition in northwestern Minnesota. Canadian Journal Forest Research, 20, 219232.
  • Clark, J.S., Merkt, J. & Müller, H. 1989 Post-glacial fire, vegetation, and cultural history of the northern alpine forelands southwest Germany. Journal of Ecology, 77, 897925.
  • Clark, J.S. & Royall, P.D. 1995 Particle-size evidence for source areas of charcoal accumulation in late Holocene sediments of Eastern North American lakes. Quaternary Research, 43, 8089.
  • Clark, J.S. & Royall, P.D. 1996 Local and regional sediment charcoal evidence for fire regimes in presettlement north-eastern North America. Journal of Ecology, 84, 365382.
  • Contini, L. & Lavarelo, Y. 1982 Le Pin Cembro, Pinus cembra L., Rèpartition, Écologie, Sylviculture, Production. INRA, Paris.
  • David, F. 1995a Mise en place des forêts d’altitude en Vanoise et périphérie. Travaux Scientifique Du Parc Nationale de la Vanoise, 19, 91106.
  • David, F. 1995b Vegetation dynamics in the northern French Alps. Historical Biology, 9, 269295.
  • Delarze, R., Caldelari, D. & Hainard, P. 1992 Effects of fire on forest dynamics in southern Switzerland. Journal of Vegetation Science, 3, 5560.
  • Delcros, P. 1994Écologie Du Paysage et Dynamique Végétale Post-Culturale En Zone de Montagne. Études du CEMAGREF, série Gestion des Territoire no. 13, Grenoble.
  • Desmaris, C. 1991Analyse Historique et Spatiale Des Mutations Socio-Économiques D’un Adret de Moyenne Maurienne. Master of Geography Dissertation, University of J. Fournier, Grenoble, France.
  • Desponts, M. & Payette, S. 1993 The Holocene dynamics of Jack pine at its nothern range limit in Quebec. Journal of Ecology, 81, 719727.
  • Earle, C.J., Brubaker, L.B. & Anderson, P.M. 1996 Charcoal in North-central Alaskan lake sediments: relationships to fire and late-Quaternary vegetation history. Review of Palaeobotany and Palynology, 92, 8395.
  • Édouard, J.-L., Tessier, L. & Thomas, A. 1991 Limite supérieure de la forêt au cours de l’Holocène dans les Alpes françaises. Dendrochronologia, 9, 125142.
  • Emanuelsson, U. 1988 A model for describing the development of the cultural landscape. The Cultural Landscape, Past, Present and Future (eds Birks, H.H. Birks, H.J.B. Kaland, P.E. & Moe, D. ), pp. 111121. Cambridge University Press, Cambridge.
  • Engelmark, O. 1984 Forest fires in the Muddus National Park (northern Sweden) during the past 600 years. Canadian Journal of Botany, 6, 893898.
  • Engelmark, O. 1987 Fire history correlations to forest type and topography in northern Sweden Acta Botanica Fennica, 24, 317324.
  • Engelmark, O., Kullman, L. & Bergeron, Y. 1994 Fire and age structure of Scots pine and Norway spruce in northern Sweden during the past 700 years. The New Phytologist, 126, 163168.
  • Filion, L. 1984 A relationship between dunes, fire and climate recorded in the Holocene deposits of Quebec. Nature, 309, 543546.
  • Filion, L. & Quinty, F. 1993 Macrofossil and tree-ring evidence for a long-term forest succession and Mid-Holocene hemlock decline. Quaternary Research, 40, 8997.
  • Forgeard, F. 1989 Étude expérimentale du rôle de la structure et de la biomasse végétale sur le comportement du feu dans les landes de Bretagne. Acta Ecologica, 10, 273294.
  • Forgeard, F. & Lebouvier, M. 1991 Inflammabilité et combustibilité des principales espèces végétales des landes de Bretagne. Bulletin D’écologie, 22, 389404.
  • Fournier, J. 1985Contribution À L’étude Des Alpes Intermédiaires Françaises: la Moyenne Maurienne; Bioclimatologie, Groupements Végétaux Forestiers et Impacts Humains. PhD Thesis, Université Scientifique et Médicale de Grenoble, Grenoble, France.
  • Gagnon, R. & Payette, S. 1981 Fluctuations holocènes de la limite des forêts de mélèzes, Rivière aux Feuilles, Nouveau-Québec: une analyse macrofossile en milieu tourbeux. Géographie Physique Quaternaire, 35, 5772.
  • Greguss, P. 1959Holzanatomie der Europaïschen Laubhölzer und Sträucher. Akadémiai Kiado, Budapest.
  • Hillaire-Marcel, C. & Occhietti, S. 1977 Fréquence des datations au 14C de faunes marines post-glaciaires de l’est du Canada et variations paléoclimatiques. Palaeogeography, Palaeoclimatology and Palaeoecology, 21, 1754.
  • Hopkins, M.S., Ash, J., Graham, A.W., Head, J. & Hewett, R.K. 1993 Charcoal evidence of the spatial extent of the Eucalyptus woodland expansions and rainforest contractions in North Queensland during the late Pleistocene. Journal of Biogeographie, 20, 357372.
  • Huttunen, P. 1980 Early land-use, especially the slash-and-burn cultivation in the commune of Lammi, Southern Finland, interpreted mainly using pollen and charcoal analyses. Acta Botanica Fennica, 113, 145.
  • Iversen, J. 1949 The influence of prehistoric man on vegetation. Danmarks Geologiske Undersøgelse, R., 3, 125.
  • Jacquiot, C. 1955Atlas D’anatomie Des Bois de Coniféres. Centre Technique du Bois, Paris.
  • Jacquiot, C., Robin, A.-M. & Bedeneau, M. 1973a Reconstitution d’un ancien peuplement forestier en forêt de Fontainebleau par l’étude anatomique de charbons de bois et leur datation par le 14C. Bulletin de la Société Botanique de France, 120, 231234.
  • Jacquiot, C., Trenard, Y. & Dirol, D. 1973bAtlas d’Anatomie Des Bois Des Angiospermes (Essences Feuillues). Centre Technique du Bois, Paris.
  • Johnson, E.A. 1992 Fire and Vegetation Dynamics, Studies from the North American Boreal Forest. Cambridge Studies in Ecology, Cambridge University Press, Cambridge.
  • Lavoie, C. & Payette, S. 1996 The long-term stability of the boreal forest limit in subarctic Québéc. Ecology, 77, 12261233.
  • Litt, T. 1992 Fresh investigations into the natural and anthropogenically influenced vegetation of the earlier Holocene in the Elbe-Saale region, Central Germany. Vegetation History and Archaeobotany, 1, 6974.
  • Mathieu, C., Payette, S. & Morin, H. 1987 Chronologie 14C et développement des combes à neige du lac à l’Eau Claire, Québec nordique. Géographie Physique Quaternaire, 41, 97108.
  • Mehringer, P.J. Jr, Arno, S.F. & Petersen, K.L. 1977 Postglacial history of Lost Trail Pass Bog, Bitterroot Mountains, Montana. Arctic and Alpine Research, 9, 345368.
  • Meyer, G.A., Wells, S.G., Balling, R.C. Jr & Jull, A.J.T. 1992 Response of alluvial systems to fire and climate change in Yellowstone National Park. Nature, 357, 147150.
  • Millet, J. & Payette, S. 1987 Influence des feux sur la déforestation des iles centrales du Lac à l’Eau Clair, Québec nordique. Géographie Physique Quaternaire, 41, 7980.
  • Occhietti, S. & Hillaire-Marcel, C. 1977 Chronologie 14C des évênements paléogéographiques du Québec depuis 14000 ans. Géographie Physique Quaternaire, 31, 123133.
  • Olsson, I.U. 1986 Radiometric dating. Handbook of Holocene Palaeoecology and Palynology (ed. B. E.Berglund), pp. 273312. John Wiley & Sons, Chichester.
  • Payette, S. & Gagnon, R. 1985 Late Holocene deforestation and tree regeneration in the forest-tundra of Québec. Nature, 313, 570572.
  • Payette, S. & Morneau, C. 1993 Holocene relict woodlands at the Eastern Canadian treeline. Quaternary Research, 39, 8489.
  • Pons, A. 1995 le développement de la paléoécologie continentale. La Révolution Technologique En Écologie (eds J.-M.Legay & R.Barbault), pp. 6387. Coll. d’Écologie no. 28, Masson Éditeur, Paris.
  • Pons, A. & Thinon, M. 1987 The role of fire from palaeoecological data. Ecologia Mediterranea, 13, 111.
  • Queiroz, P.F. & Van Der Burgh, J. 1989 Wood anatomy of Iberian Ericales. Revista de Biologia, 14, 95134.
  • Romme, W.H. 1982 Fire and landscape diversity in subalpine forests of Yellowstone National Park. Ecological Monographs, 52, 199221.
  • Romme, W.H. & Knight, D.H. 1982 Landscape diversity: the concept applied to Yellowstone Park. Bioscience, 32, 664670.
  • Romme, W.H., Turner, M.G., Wallace, L.L. & Walker, J.S. 1995 Aspen, Elk and fire in Northern Yellowstone National Park. Ecology, 76, 20972106.
  • Schär, E. & Schweingruber, F.H. 1988 4000 years forest development in pre-Christian times near a mountain lake in the Alps. Dendrochronologia, 6, 131140.
  • Schweingruber, F.H. 1978Mikroscopische Holzanatomie. Anatomie Microscopique Du Bois. Microscopical Wood Anatomy. Inst. féd. rech. for. Birmensdorf. Zürcher AG, Zug.
  • Schweingruber, F.H. 1990 Anatomie Europaïscher Hölzer. Anatomy of European Woods. Eidgenöss, Forschungsanst, f. Wald, Schnee U. Landschaft Birmensdorf-Verlag, Paul Haupt, Bern.
  • Serre-Bachet, F. 1978 The dendroclimatological value of the European Larch (Larix decidua Mill.) in the French Maritime Alps. Tree-Ring Bulletin, 38, 2534.
  • Serre-Bachet, F. 1986 Une chronologie maitresse du sapin (Abies alba Mill.) du Mont ventoux (France). Dendrochronologia, 4, 8796.
  • Steijen, I. & Zackrisson, O. 1987 Forest fires in the Muddus National Park (northern Sweden) during the past 600 years. Canadian Journal of Botany, 65, 839848.
  • Stern, R. 1983 Human impact on tree borderlines. Man's Impact on Vegetation (eds W.Holzner, M. J. A.Werger & I.Ikusima), pp. 227236. Dr Dr W. Junk, Publishers, The Hague.
  • Suffling, R. 1993 Induction of vertical zones in sub-alpine valley forests by avalanche-formed fuel breaks. Landscape Ecology, 8, 127138.
  • Talon, B. 1997 Étude anatomique et comparative de charbons de bois de Larix decidua Mill. et Picea abies (L.) Karst. Comptes Rendues de l’Académie Des Sciences Paris, Série Sciences de la Vie/Life Sciences, 320, 581588.
  • Talon, B., Carcaillet, C. & Thinon, M. 1998 Apport de la pédoanthracologie à l’étude des variations de la limite supérieure des arbres au cours des derniers millénaires dans les Alpes françaises. Géographie Physique Quaternaire, 52, in press.
  • Tarrega, R. & Luis-Calabuig, E. 1990 Forest fires and climatic features in Leon Province (Spain) – fire effects on Quercus pyrenaica ecosystems. Fire in Ecosystem Dynamics (eds J. G.Goldammer & M. J.Jenkins), pp. 6369. SPB Academic Publishing, The Hague.
  • Tatoni, T., Magnin, F., Bonin, G. & Vaudour, J. 1994 Secondary succession on abandoned cultivation terraces in calcareous Provence. I. Vegetation and soil. Acta Ecologica, 15, 431447.
  • Tessier, L. 1986 Chronologie de mélèzes des Alpes et petit age glaciaire. Dendrochronologia, 4, 97113.
  • Thinon, M. 1978 La pédoanthracologie: une nouvelle méthode d’analyse phytochronologique depuis le Néolithique. Comptes Rendues de l’Académie Des Sciences Paris, Série D, 287, 12031206.
  • Thinon, M. 1992L’analyse Pédoanthracologique: Aspects Méthodologiques et Applications. PhD Thesis, University of Aix-Marseille III, Marseille, France.
  • Tinner, W., Ammann, B. & Germann, P. 1996 Treeline fluctuations recorded for 12500 years by soil profiles, pollen, and plant macrofossils in the Central Swiss Alps. Arctic and Alpine Research, 28, 131147.
  • Tolonen, M. 1978 Palaeoecology of laminated sediment in Lake Ahvenainen, South Finland. I. Pollen and charcoal analyses and their relation to human impact. Acta Botanica Fennica, 15, 155208.
  • Trabaud, L. 1976 Inflammabilité et combustibilité des principales espèces des garrigues de la région méditerranéenne. Ecologia Plantarum, 11, 117136.
  • Turner, M.G., Hargrove, W.W., Gardner, R.H. & Romme, W.H. 1994 Effects of fire on landscape heterogeneity in Yellowstone National Park, Wyoming. Journal of Vegetation Science, 5, 731742.
  • Veblen, T.T., Hardley, K.S., Nel, E.M., Kitzberger, T., Reid, M. & Villalbo, R. 1994 Disturbance regime and disturbance in a Rocky Mountain subalpine forest. Journal of Ecology, 82, 125135.
  • Vernet, J.-L., Wengler, L., Solari, M.-E., Ceccanti, G., Fournier, M., Ledru, M.-P. & Soubiès, F. 1994 Feux, climats et végétations au Brésil central durant l’Holocène: les données d’un profil de sol à charbons de bois (Salitre, Minas Gerais). Comptes Rendus de l’Académie Des Sciences Paris, Série II, 319, 13911397.
  • Van Vliet-Lanoë, B. 1988 Impact pédomorphologique du feu en contexte de début et de fin glaciaire. Bulletin Du Centre de Géomorphologie, 35, 2186.
  • Vorren, K.-D., Mørkved, B. & Bortenschlager, S. 1993 Human impact on the Holocene forest line in the Central Alps. Vegetation History and Archaeobotany, 2, 145156.
  • Vuorela, I. 1970 The indication of farming in pollen diagrams from southern Finland. Acta Botanica Fennica, 113, 140.
  • Vuorela, I. 1986 Palynological and historical evidence of slash-and-burn cultivation in Southern Finland. Anthropogenic Indicators in Pollen Diagrams (ed. K.-E.Behre), pp. 5364. Balkema, Rotterdam/Boston.
  • Wein, R.W., Burzinski, M.P., Sreenivasa, B.A. & Tolonen, K. 1987 Bog profile evidence of fire and vegetation dynamics since 3000 years BP in the Acadian forest. Canadian Journal of Botany, 65, 11801186.
  • Whelan, R.J. 1995 The Ecology of Fire. Cambridge Studies in Ecology. Cambridge University Press, Cambridge.
  • White, M.A. & Vankat, J.L. 1993 Middle and high elevation coniferous forest communities of the North Rim region of Grand Canyon National Park, Arizona, USA. Vegetatio, 109, 161174.
  • Whitlock, C. & Millspaugh, S.H. 1996 Testing the assumptions of fire-history studies: an examination of modern charcoal accumulation in Yellowstone National Park, USA. The Holocene, 6, 715.
  • Wick, L. 1994 Early Holocene reforestation and vegetation change at a lake near the alpine forest limit: Lago Basso (2250 m a.s.l.), northern Italy. Dissertationes Botanicae, 234, 555563.
  • Zackrisson, O. 1977 Influence of forest fires on the North Swedish boreal forest. Oikos, 29, 2232.
  • Zackrisson, O., Nilsson, M.-C. & Wardle, D.A. 1996 Key ecological function of charcoal from wildfire in the boreal forest. Oikos, 77, 1019.
Footnotes
  1. Present address and correspondence: C. Carcaillet, Laboratoire de Paléobiogéographie et de Palynologie, Département de Géographie, Université de Montréal, CP 6128 succ. ‘Centre-Ville’, Montréal (Qc), H3C 3 J7 Canada (tel. + 1 (514) 343 8048; fax + 1 (514) 343 8008; e-mail carcailc@magellan.umontreal.c).

Received 28 February 1997revision accepted 18 August 1997