From pixels to pastures: methods for multi-scale assessment of ground lichen as a reindeer grazing resource

Ground lichens are a critical winter food source for reindeer, yet lichen resources have declined substantially over recent decades in Sweden. Pressure from competing land uses and climate change further affects their accessibility, determining whether reindeer can reach and utilise the lichen that remains. Reliable and transparent methods to map and measure lichen and assess accessibility of lichen are therefore much needed to support the management and planning of reindeer winter grazing areas. This thesis develops and evaluates such methods across three spatial scales, and demonstrates how reindeer movement data can be used to analyse the accessibility of lichen pastures.At the broad scale, a method was developed that uses reindeer GPS locations to validate a satellite-based lichen cover map of the Swedish reindeer winter grazing area. At the local scale, a drone-based method was developed to estimate lichen cover and biomass within specific areas. At the plot scale, field methods for estimating lichen cover were evaluated and compared, and a deep learning image segmentation model was developed to classify lichen in ground-level photographs, providing a more objective and consistent alternative to visual assessment. To link lichen estimates to reindeer, a simple approach to estimate reindeer grazing days from lichen biomass is also presented.The thesis also demonstrates methods for analysing how access to lichen pastures is affected by industrial development and other disturbances, accounting for cumulative impacts from infrastructure such as wind power plants, mining, and roads.Together, these methods form a complementary toolkit that can be applied independently or in combination to map and quantify lichen resources across multiple scales in a transparent and reproducible way.