Skip to main content
SLU publication database (SLUpub)
Research article - Peer-reviewed, 2023

From simple metrics to cervid forage: Improving predictions of ericaceous shrub biomass

Juvany, Laura; Hedwall, Per-Ola; Felton, Adam; Ohman, Karin; Wallgren, Martha; Kalen, Christer; Jarnemo, Anders; Johansen, Henrik; Felton, Annika


Common understory vegetation species such as the ericaceous shrubs bilberry (Vaccinium myrtillus), cowberry (V. vitis-idaea) and heather (Calluna vulgaris), are key forage plant species for moose and other large herbivores, as well as fulfilling many additional ecosystem functions and services. Here we developed models to predict above-ground biomass of these ericaceous species in coniferous forests, using data on their percentage cover, height, and different stand characteristics. We also built models to understand how the aforementioned variables affect the proportion of the shrubs commonly utilized as forage by large herbivores. We found that the per-centage cover of shrubs was the most important explanatory variable when predicting above-ground biomass, explaining 51%, 47% and 71% of the variation (marginal R2) in bilberry, cowberry and heather biomass, respectively. By adding ramet height to the model with percentage cover, the variation explained increased to 77% for bilberry, 75% for cowberry and 87% for heather. The best outcome for candidate models was obtained by adding stand site index and spruce basal area to the model, improving the variation explained in bilberry to 83%, to 81% for cowberry, and 91% for heather. When modelling the proportion of the shrubs commonly uti-lized as forage by large herbivores, stand site index and spruce basal area often played important roles. Some of the best fitting models for forage biomass explained 51% of the variation in bilberry, 59% in cowberry and 30% in heather. Site location did not have a major role in improving the variability explained in either type of model, which indicated the applicability of the models regardless of study location. Our models therefore have a high potential to be implemented in forestry decision support systems. Their inclusion should provide better large-scale estimations of forage resources, aiding forest management, and thereby taking an important step for-ward to determine the ecosystem carrying capacity of large herbivores.


Biomass modelling; Vaccinium myrtillus; Vaccinium vitis-idaea; Calluna vulgaris; Boreal forest; Ungulate forage

Published in

Forest Ecology and Management
2023, Volume: 544, article number: 121120
Publisher: ELSEVIER