SEARCH

SEARCH BY CITATION

Keywords:

  • ArcGIS;
  • Arctic;
  • boreal;
  • bioclimatology;
  • bioclimatic map;
  • index of thermophily;
  • Landsat TM;
  • vegetation map;
  • vegetation types

Abstract

Aim  To develop a new method for bioclimate mapping where the vegetation layer is the main source of climate information.

Location  The study area includes four subareas, all situated on the Varangerhalvøya peninsula in Finnmark, north-easternmost Norway (70–71° N). The four subareas were chosen to represent most of the climatic, topographic, geomorphologic and botanic diversity along the arctic–boreal gradient in the area. The four meteorological stations in the area show a climatic gradient with mean July temperature ranging from 10.1 to 12.3 °C.

Methods  The new vegetation-based method is based on the fact that most plant species and plant communities both in the Arctic and adjacent areas have a distribution pattern limited by temperature to some extent. The vegetation is mapped using Landsat TM data and a contextual correction process in a geographic information system. The mapped vegetation units are defined as temperature indicators based on their total distribution patterns and the temperature indicator value of their high frequency and dominant species. The indicator value and degree of cover of all thermophilous vegetation units, within each 500 × 500 m study unit, are combined in a Vegetation-based Index of Thermophily, VItm. This new vegetation-based method is based on the same basic idea as a recently published floristic-based method for calculating a Floristic-based Index of Thermophily, FItm. The VItm values are tested by comparison with the FItm values, and temperature data collected in the field during two growing seasons, and the differences are interpreted ecologically.

Results  Twenty-one of the mapped vegetation units were defined as thermophilous and categorized in five groups of temperature indicators. The VItm values showed a strong positive linear relationship with the temperatures measured during the years 2001 and 2002, with r2 values of 0.79 and 0.85, respectively. The VItm values show a high linear relationship (r2 = 0.76) with the 71 study units where the FItm values were calculated. As interpreted from the relationship with temperature measurements and FItm values, the vegetation-based method seems to work at a broad range of ecological conditions, with very dry, acidic sites being the most important exception. The VItm values are related to growing degree-days of a normal year, and the four subareas are mapped, showing a diversity of 13 bioclimatic classes. The birch forest line is estimated to occur at about 980 °C-days. The results show climatic gradients with temperatures increasing from the cold coast towards the interior, from wind-exposed convex hills towards wind-protected valleys, and from mountain plateaux towards south-facing lowlands. The north-easternmost study site at the coast is positioned within the arctic shrub tundra zone.

Main conclusions  The vegetation-based method shows a strong positive correlation both with measured temperatures and the floristic-based method within a broad range of different ecological conditions. The vegetation-based method has the potential for bioclimatic mapping of large areas in a cost-effective way. The floristic-based method has higher accuracy and is more flexible than the vegetation-based method, and the two methods seem to complement each other.