Abstract

Topic: Moose management and monitoring

Mapping cervid forage in Sweden using remote sensing and national forest inventory data

Lukas Graf¹, Inka Bohlin², Jonas Dahlgren³, Per-Ola Hedwall⁴, Annika Felton⁴

Southern Swedish Forest Research Centre, SLU, Sweden, Sundsvägen 3, SE
Department of Forest Resource Management, Skogsmarkgränd Umea, Sweden
Department of Forest Resource Management, Skogsmarkgränd Umeå, SLU, Sweden
Southern Swedish Forest Research Centre, Sundsvägen 3, Alnarp, Sweden

Abstract
Browsing of cervids (moose (Alces alces) or roe deer (Capreolus capreolus)) influences forest ecosystems worldwide stressing the need for wildlife management founded in accurate estimates of available forage. In this study, we developed the first national-scale models for Sweden to estimate the abundance of cervid forage using a novel approach combining data from the Swedish National Forest Inventory (NFI) and remote sensing (RS). We focused on six key forage tree species: Scots pine (Pinus sylvestris), birch (Betula spp.), European aspen (Populus tremula), rowan (Sorbus aucuparia), oak (Quercus spp.), and willow (Salix caprea). We combined airborne laser scanning and other auxiliary RS data with NFI data from 2016 to 2022 on small tree abundance from 19.461 plots across Sweden and parametrized generalized linear mixed models using likelihood-ratio tests to predict species-specific forage availability. Our models demonstrated moderate to strong predictive performance, with marginal R² values ranging from 22.6 to 97.3. We validated our models with data from 1.324 independent NFI plots collected in 2023. The model validation showed robust predictive accuracy with low relative root mean square error and root mean square error. The resulting species-specific forage maps provide valuable tools for wildlife management, forestry planning, and ecological research. We provide six species-specific maps both at 1ha and 1km2 spatial scales. Our map products enable stakeholders to assess the spatial distribution of cervid forage, optimize habitat management, and mitigate browsing-related economic losses in forestry.