Master Thesis

Assessing the forest response along treelines to an Epirrita autumnata outbreak in Abisko, using a combination of fieldwork and remote sensing

Frank Weiser (09/2017-03/2017)

Support: Carl Beierkuhnlein, Martin Wegmann

Outbreaks E. autumnata have increased in frequency all over Northern Scandinavia due to climate change. The outbreaks are able to defoliate and severely damage the Betula pubescens forests of the area. While the birch forests profit from climate change and are able to both move upward and laterally increase their extent, the pressure by more frequent outbreak damage might hinder this increase or even push back treelines. Most research regarding the outbreaks is based on very time consuming field methods, remote sensing techniques are severely underused. Therefore, one aim of this study was to assess the potential of RapidEye data due to its 5x5m spatial resolution and its red edge band, which is very sensitive to changes of the vegetation, for detecting treeline shifts due to E. autumnata damage. The expectations were not completely fulfilled. Supervised classifications proofed difficult due to low canopy cover of some forests and gradual shifts from forests to brushland. A method based on the standard deviation of the Normalized Difference Red Edge Index (NDRE) of an image stack showed better results at detecting forest borders but still had some problems to solve. Field work data collected in order to understand the influence of forest structure on the susceptibility to outbreak damage showed elevation of the treeline to be the most important predictor for the 2012/2013 outbreak damage. The relation showed that low elevation treelines suffered more damage during the outbreak. This suggests, that not all treelines are influenced to the same extent by Geometridae feeding and implies that low elevation forest shifts could be hindered more than high elevation shifts.