A remote sensing methodology for monitoring lichen cover

Abstract

Land degradation has been recognised in the mountainous areas of Sweden due to increased land use, particularly the intensive grazing and trampling by reindeer, which causes mechanical damage to the vegetation cover. As these areas are often inaccessible, it is valuable to be able to use satellite data to monitor vegetation changes. The Swedish Environmental Protection Agency has proposed an environmental quality objective for mountainous areas of Sweden, which emphasizes the need for monitoring. The aim of this paper is to investigate a method for such monitoring using satellite data. Of the different heath communities above the tree line, the lichen-dominated heath is among the most sensitive to mechanical damage. Hence lichen cover is used as an indicator of change because of its ecological relevance and its spectral characteristics. Landsat-5 thematic mapping (TM) data, hyperspectral imaging scanner data, and spectral characteristics of relevant mountainous plant communities and lichen species were used to study heath vegetation in the southern part of the Swedish mountain range. For comparison, colour-infrared (CIR) aerial photographs at a scale of 1 : 60 000 and field data were used. The changes in lichen cover have been detected by spectral changes using the normalized difference vegetation index (NDVI) differencing technique and thresholding. The results show that mountainous lichen-dominated heath, above the tree line, can be mapped with good accuracy using Landsat TM data, and this heath is proposed for change detection, as it is possible to differentiate lichen cover in the following classes: low (20‐50% cover), moderate (51‐80% cover), and high (>80% cover). A lichen cover <20% cannot be separated from other types of dry heath. The class boundaries have been determined using field observations and CIR aerial photographs. The method indicates that a change in lichen cover can be classified and mapped in three classes: unchanged, moderate decrease, and high decrease. The classes can be regarded as three risk-assessment classes for vegetation degradation and ensuing soil erosion. The major conclusion from this study is that a change in lichen cover, differentiated in three classes, can be used as a tool for monitoring disturbed ecosystems in the Swedish mountain range.