Metla Project 830701
Sustainable development of the Pechora region in a changing environment and society SPICE
Keywords: Boreal forests, GIS, North-West Russia, Pechora, biodiversity, forestry, global change
Objectives The project is part of the EC-funded (5th Framework Programme, COPERNICUS2) SPICE-project (Sustainable development of the Pechora region In a Changing Environment). SPICE assesses alternative scenarios for the sustainable development of the Pechora region, North-east European Russia. SPICE is coordinated by Arctic Centre, University of Lapland. The Pechora region, which politically includes the North and east of the Komi Republic and the East of the Nenets Autonomous Region, faces considerable challenges both in terms of socio-economic development and environmental conditions. The region has extraordinary natural resource, both renewable (e.g. forests) and non-renewable (e.g. coal, oil and gas). In addition, traditional livelihoods (reindeer herding, fishing) are important subsistence sectors. In terms of the environment, the region is unique in continental Europe with extensive lowland tundra and permafrost in the north, and the largest contiguous oldgrowth taiga forest in the Ural mountains (an area of ca. 25000 km2 is protected). More information of SPICE at http://kaares.urova.fi/home/arktinen/spice/spice.htm.
The aim of the projectWorkpackage 'Habitat classification and the analysis of landscape structure' (Metla, Rovaniemi Research station) aims at
Four comparison areas for boreal taiga forests and four for tundra areas are selected based on the fact that they fulfill two criterias:
- generating land use and cover classification for selected study areas utilizing Landsat TM5 satellite images, ground truth data and digital maps,
- calculating indices of landscape structure for these areas and
- linking measures of biodiversity in the same areas to landscape structure indices, together with other workpackages in SPICE-project.
- they originally have represented similar habitat composition on each study area and
- the other area has been changed due to human activities where as the other one represents untouched reference area.
Results In each research area eight 2 km transects were inventoried and data about forest structure and other vegetation was recorded every 100 m interval. Transects, as well as, ground truth plots along transects were located and marked on satellite image printouts.
Each study area was classified to about ten habitat and land use and cover classes. The overall accuracy of classification for four taiga-sites ranged from 51.5 % to 71.3 %, and weighted Kappa statistics from 0.41 to 0.65, respectively. Pine-lichen type dominated forests as well as deciduous tree dominated young forests could be separated relatively well, but there was more confusion between the classes spruce- and spruce-fir dominated and mixed forests. However, also the field-observations showed that these types create such continuum that defining borderlines between them is very arbitrary.
The accuracy of classification in tundra-sites ranged from 59.7 % to 64.6 % and weighted Kappa statistics from 0.47 to 0.58, respectively. Misclassifications were found especially between open bog and shrub-moss tundra. In tundra areas, also the coarse spatial resolution of Landsat images causes problems, as the size of vegetation patches is often only some tens or hundreds os square meters.
The comparisons of the landscape structure in taiga areas between human impact and pristine areas showed that there were some obvious differences in the vegetation and landscape structure between areas, but all differences were not totally due to different logging history. The differences were explained partly also due to different edaphic related growth conditions and the intensity of the earlier forest fires. Most loggings had been made several decades ago and after cutting forests have regenerated naturally. Today, earlier spruce dominated or spruce-fir mixed forests are now dominated by birch (Betula pubescens Ehrh.), often with some spruce undergrowth. Regeneration has not been very good at all locations, at least partly due to excessive moisture. In pristine areas, like in the National park Yugud-Va, there has been no forest loggings and some young stands observed in the area were caused especially by wind and snow damage, and probably partly also due to forest fires.
In tundra areas, the signs of human impact could be related to oil and gasfields, but the proportion of human impact areas of the study areas (about 15 km × 15 km) was generally low, only some few per cents of the area.
The analysis of the effect of human induced changes in study areas on bird diversity are under preparation jointly with researchers in Arctic Centre.
The Finnish Forest Research Institute,
PL 16, FI-96301 ROVANIEMI, FINLAND
Phone: +358 29 532 5568
Mikkola, Kari, RO (2001-02), Virtanen, Tarmo, RO (2000-01)
Top of page