Pages:     | 1 |   ...   | 43 | 44 || 46 | 47 |   ...   | 49 |

It has been speculated that in minerotrophic mires with a large watershed, ditching may lower the concentrations of DOC in runoff waters (Sallantaus 1988); due to restoration, the reverse is obvious. But in addition, drainage has altered the peat quality, and also the anaerobic and aerobic decomposition of large amounts of fresh organic matter, drowned by rewetting, is initiated. As a result, the concentrations of organic carbon in the restored minerotrophic mires are substantially higher than in corresponding natural mires (cf. Tahvanainen et al. 2002, Laine et al. 2004). Organic matter has acidic characters and is affecting also pH values, which are very important in determining the vegetation.

Break down of organic matter, fine roots, aerobic decomposers etc., releases also phosphorus. The background mean annual leaching rate of phosphorus in Seitseminen, 7-8 kg/km2, or 5 kg/km2 in Nuuksio, is rather typical or slightly higher than what is common in totally undisturbed conditions (Mattsson et al. 2003). An increase in the leaching of nutrients due to restoration was expected. The removal of trees resembles the effects of clear cutting, and clear cutting in peatlands is known to increase especially phosphorus concentrations (Ahtiainen 1992). The wetting caused by blocking the ditches is crucial for all organisms having underground parts and requiring oxic conditions, including forest plants and micro-organisms. In addition to increased decomposition, phosphorus is mobilized due to reduced uptake.

Anoxic conditions are known to mobilize phosphorus due to chemical reactions as well.

The increase in phosphorus leaching was, however, unexpectedly large. The mires in Seitseminen are poor in nutrients, and a significant part of the total phosphorus store must have been mobilized (cf. Westman & Laiho 2003). The increase of phosphorus leaching due to drainage is usually much less (Laine et al. 1995), but clear cuttings may mobilize similar amounts (Ahtiainen 1992). Fertilization of peatland forests has caused even much larger specific loads ( Saura et al. 2000). Previous fertilizations may have exacerbated the leaching problems in the case of Seitseminen, but fertilization is not a prerequisite for the phenomenon.

Similar amounts of P were mobilized from the spruce mire catchments with very different fertility, vegetation, and in which the tree stands were not harvested.

The hydrochemical phenomena behind the observed changes may also affect the establishment of new vegetation to the peatland itself e.g. by favouring nutrient demanding rapid colonizers, instead of the wanted mire specialists, which may still persist in the vegetation. Oxygen consumption by mobilized organic matter is an important question, and not properly studied yet. Some forms of aquatic life may be threatened in the protected area itself, or unwanted changes may occur in the recipient watercourses. Also some plants may be sensitive to either increased acidity or poorer oxygen conditions, either directly or through symbiotic factors. Potential water quality impacts and the processes behind these impacts need therefore to be carefully considered in all mire restoration works. Fortunately, serious threats are rare and aquatic impacts can be minimized by avoiding large restoration areas in the same watershed at the same time.

References Heikkil, H. & Lindholm, T. 1994. Restoration plan for the drained mires in the Seitseminen national park. Finnish Forest and Park Service, Metshallituksen luonnonsuojelujulkaisuja, Ser. B 13. 127 p. (in Finnish).

Kalela, A. 1961. Waldvegetationszonen Finnlands und ihre klimatischen Paralleltypen. Arch. Soc. "Vanamo" 16 (suppl.): 65-83.

Laine, J., Komulainen, V.-M., Laiho, R., Minkkinen, K., Rasinmki, A., Sallantaus, T., Sarkkola, S., Silvan, N., Tolonen, K., Tuittila, E.-S., Vasander, H. & Pivnen, J. 2004. Lakkasuo a guide to mire ecosystem. University of Helsinki, Department of Forest Ecology Publications 31. 123 p.

Laine, J., Vasander, H. & Sallantaus, T. 1995. Ecological effects of peatland drainage. Environmental Reviews Vol. 3. N:o 3-4 p. 286-303.

Leivo, A., Liedenpohja-Ruuhijrvi, M., & Tuominen, S. 1989. The vegetation of the Seitseminen national park. Metshallitus Finnish Forest and Park Service, Series SU 4:96. 50 p. + app. 2. ed. (In Finnish).

Mattsson, T., Finr, L., Kortelainen, P. & Sallantaus, T. 2003. Brook water quality and background leaching from unmanaged forested catchments in Finland. Water, Air, and Soil Pollution 147: 275-297.

Ruuhijrvi, R. & Lindholm, T. 2005. The Finnish mire site type classification system. This volume.

Sallantaus, T. 1988. Water quality of peatlands and man's influence on it. In:

Proceedings of the international symposium on the hydrology of wetlands in temperate and cold regions, Joensuu, Finland 6.-8. June 1988, vol. 2. Helsinki, the Academy of Finland. P. 80-98. The Publications of the Academy of Finland 1988 nro 5.

Sallantaus, T. 2004. Hydrological impacts set constraints on mire restoration.

In: Pivnen, J. (Ed.). Wise use of peatlands. Proc. 12th Int. Peat Congr., Tampere, Finland Vol. 1 p. 68-73. Int. Peat Soc., Jyvskyl.

Saura, M., Frisk, T., Sallantaus, T. & Bilaletdin,. 2000. The effects of forest fertilization on a small polyhumic lake. Verh. Internat. Verein. Limnol. 27:


Tahvanainen, T., Sallantaus, T., Heikkil, R. & Tolonen, K. 2002. Spatial and seasonal variation in surface water chemistry in relation to vegetation in two fens, north-eastern Finland. Annales Botanici Fennici 39,3: 235-Westman, C.J. & Laiho, R. 2003. Nutrient dynamics of drained peatland forests. Biogeochemistry 63: 269-298.

RELIABLE POSTGLACIAL FIRE RECORD OF BOREAL FORESTS FROM PEAT STRATIGRAPHICAL DATA K. TOLONEN & A. K. PITKNEN University of Joensuu, Finland kimmo.tolonen@dnainternet.net, aki.pitkanen@joensuu.fi The long-term fire record of forests surrounding certain small mire basins is preserved as thin charcoal layers in the peat. The fires typically advance only a few metres on mire surface, or are blocked about the margin of the mineral soil. Since the peat deposits have expanded (both vertically and horizontally) in the course of millennia, the number and age of the charcoal layers varies along a transect between the margin and the centre of the basin. A reliable reconstruction of history of the forest fires is obtained, when the charcoal layers are clarified and dated in numerous profiles along the transect (the columns denote the location of suitable coring sites at this hypothethic site).

The principle of the method is simple (fig. 1.). Since fire intervals estimated by our method were in agreement with dendrochronological fire scar records extending about 500 years back, we believe that the new peat stratigraphical method provides valid results for the older peat strata in the the same basins as well. Before intensive land use the natural average fire interval in middle boreal pine forests of Eastern Finland varied as follows: (i) during the whole abiegnic time the average fire interval was between 170 and 240 years, (ii) around the (second) arrival of spruce about 100 years, (iii) during the Atlantic Chronozone fires recurred at an interval of 600 - 900 years, and (iv) during the Boreal Chronozone the average fire interval was some 200 - 300 years.

During the Preboreal Chronozone fire interval was on average about 100 - years (v). At one pine dominated esker forest in western Finland (southern boreal zone) the average natural fire interval varied as follows:(i) during the abiegnic time (3300 B. C. - AD 1020) the average fire interval was 60 - years (ii) but during the remarkable forest use (AD 1020 - AD 1845) only 35 - 45 years.

Certain spruce mires were true fire refugias during the Holocene, and spruce forests on mineral soil burned naturally with very long intervals, and many forest patches in fragmented lanscapes with mires, lakes and streams did possibly not meet fire during the Postglacial. In Kuhmo (middle boreal zone) some of the studied mineral soil forests burned with following intervals: (i) thru the preabiegnic time period some 1000 years on average, (ii) before B.C. at some 300 - 400 years, (iii) during the abiegnic time prior to beginning of the swidden agriculture in the area, the average fire interval was about 320 - 520 years.

Figure 1. The scheme of fire layeres inside a peat deposits.

The results obtaided fit well with the knowledge about the density of lighting induced fires in boreal European forests. The fire regime was found to control the structure of the forests in terms of the dominance of the trees, for example. Our new results arouse serious concern about the present forestry practice in coniferous forests, and in spruce forests in particular, in which intensive forestry has strongly distorted the rotation time from that of the natural state.

According to historical sources the forest fires have been very common in boreal forests (eg.Vakurov 1955, Goldammer and Furyaev 1996) but their precise occurrence in time and space often remains unclear. Reliable and systematically collected fire data are available from the early 19th century onwards. The fire scar method of pine (and sometimes of other trees) is superior among the methods both in temporal resolution (in some cases, in addition to the fire year, even the season can be determined), and in the spatial accuracy (eg. Lehtonen 1998). Since wood material older than some 400 - 500 years is rather scarce (oldest material available being of about 750 years of age) the tree ring method covers only a short, and relatively recent period of the Holocene. The results by this method show that forest fires recurred in dry pine forests in eastern Finland at an average interval of 30 to 40 years or even less, during the most intensive swidden agriculture about 1700 - 1850 AD.

Outside the slash- and burn province, in Northern Norrland (Sweden) (Zackrisson 1977) and in Viena Karelia (Russia) (Lehtonen and Kolstrm 2000), the fire frequency was lower (average fire return interval between 50 and 100 years).

Quite a lot data has been accumulated from charcoal analysis of lake sediments (and occasionally from peat deposits) (e.g. Tolonen 1983), and some of the data cover the whole Holocene period and extend even longer. Although the charcoal particle data from lake sediments and peat can potentially provide an opportunity to detect, and date accurately even single fire events of the past, the exact location of fires remains usually open. Owing to taphonomic processes in lakes, interpretation of lake sediment charcoal particle data is often difficult, so reliable and unequivocal reconstructions of the fire history are in many cases not possible.

The recent improved peat stratigraphical method (Pitknen et al.

2001) enables, we believe, a reliable, spatially precise and temporaly sufficient record for the postglacial fire history of boreal forests to be achieved. The method is simple (fig. 1). We have used the method in oner small mire basin nearby the northern shore of Lake Suomunjrvi in Patvinsuo National Park, eastern North-Karelia (fig. 2: site A). The mire represented a pine bog with tall dwarf shrubs (Betula nana, Ledum, Vaccinium uliginosum and Chamaedaphne etc.) The mire basin was surrounded by dry pine forests of Empetrum - Vaccinium (EVT) type in the Finnish forest type classification.

Dendrochronologically dated forest fire data from similar dry pine forest at the southern shore of the lake, about three km of the current site (FIG 2: site C) indicated eight fires between 1531 AD and 1832 AD corresponding to an average fire interval of about 50 years. Accordingly, we assume that the charcoal layers above the decline of spruce in the pollen diagram (the event indicating the beginning of slash-and-burn cultivation in the area) correspond fires between AD 1500 and AD 1900.

Fig. 2. Location of the studied mire basins ( A and B) near Lake Suomunjrvi in Patvinsuo National Park. In the index map are given: the boundary line between the southern boreal and the middle boreal zones, Ulvinsalo = U, Patvinsuo = P and Orivesi = O.

Since the peat type indicated fairly constant mire hydrology through the whole peat strata at the mire margin, we believe that the peat has reliably registrated the fire events during the whole postglacial period at the site studied. However, for several technical and taphonomical reasons, the peat stratigraphical fire record is always somewhat conservative. This issue is discussed in Pitknen et al (2003). The obtained peat cores were dated by numerous conventional 14 C- datings, but the comparison with pollen analytical marker horizons with known average ages in the region led us to the conclusion that many of those radiocarbon ages were "erronously too young". Therefore, our event chronology was primarily based on the pollen analytical marker levels. The dating may diverge a little from right radiocarbon ages but considering the long periods used for averaging the fire interval estimates, it is not so crucial to attain an exactly accurate dating. The charcoal stratigraphy of a peat basin near the southern shore of the lake (FIG 2: site B) was also studied (TABLE 1).

Table 1.

Charcoal layer stratigraphy, radiocarbon and pollen ages from coring points of sites near northern (A) and southern shore (B) of Lake Suomunjrvi.

distance number of charcoal layers during four periods from (cal.BP) dept pollen radiocarbon mineral h, age, soil, age BP present- 450- 6300- 9000cm cal.BP total 450 6300 9000 m site A 0.5 30 - 4 1 3 - 1 50 - 13 3 10 - 1.5 53 388080 4320 12 0 12 - 2 67 469090 6300 21 1 19 1 3 85 6220100 >9000 32 2 26-27 1-2 10 140 7910110 10500 32 0 23 5 site B 4 37 252080 2715 9 6-7 3-2 - 4.5 40 - 6300 21 4 16 1 5.3 57 5700100 6480 35 6 28 1 6470100 20-7.7 62 10500 29 1 3-4 One can briefly sum up the average fire interval estimates at the two sites and at two additonal sites around Lake Suomunjrvi (Pitknen et al. 2002, 2003) as follows:

1. during the "Betula- period" (9000 - 8000 cal yrs. B.C.) about 100 - years 2. during the "Pine- period" (8000 - 7000 B.C.) some 200 - 300 years 3. during the Atlantic Chronozone (7000 - 4300 B.C.) some 600 - years 4. around the (second) postglacial arrival of spruce (4500 - 4000 B.C.) about 100 years 5. during the abiegnic time, but before the swidden culture in the area (4500 - 4000 B.C. - AD 1500), 170 - 240 years 6. during the period of swidden agriculture (AD 1500 - AD 1900) about 50 years It seems likely that within the general spread of spruce (Picea abies) the summers might have turned drier, and the frequent fires by one way or another way facilitated spruce invasion and its competitiveness in then prevailing closed mesic forests with plenty of deciduos trees. At one site in southern Tavastia, in southern boreal Finland, the forests of Atlantic period, prior to the invasion of spruce, had a clear dominance of lime (Tilia cordata) (M.Tolonen, 1978). It has been considered (Huntley 1988) that likewise the forest fires probably contributed to the spread of beech (Fagus silvatica) in Cental Europe in the late Holocene.

Another location, where we done a similar study on the past forest fires is Orivesi, southern boreal Finland. At present pine forests of Vaccinium type (VT) surround the small mire basin in an esker landscape.

The "basin based approch" revealed an average fire interval of 60 - years during the abiegnic period (3300 B.C. - AD 1200), shortening into 35 - 45 years during the period 1200 AD - 1845 AD, or the time when traces of the human impact to the forests in the area appear in the pollen analyses and otherwise. Since AD 1845 there have been no fires on that esker.

Pages:     | 1 |   ...   | 43 | 44 || 46 | 47 |   ...   | 49 |

2011 www.dissers.ru -

, .
, , , , 1-2 .