Data from: The effects of moose- and pine density on browsing damage in Swedish pine forests
Data files
Feb 19, 2026 version files 52.12 KB
-
R_code_and_data.zip
48.35 KB
-
README.md
3.77 KB
Abstract
Moose (Alces alces) is a culturally and economically important game species in Sweden, but their browsing on regenerating Scots pine trees (Pinus sylvestris) often causes extensive damage to the production and quality of timber. Forest- and wildlife managers are faced with the dilemma of how to reduce damage to timber trees while also supporting moose populations and hunting opportunities. The proportion of damaged trees can be reduced by decreasing the number of moose, but also by increasing the number of pines. However, the relative effectiveness of these two approaches is debated and has not been conclusively determined. Here we addressed this question by analyzing the effects of moose- and pine density on pine damage based on yearly data from almost all of Sweden’s moose management areas (MMAs) over 10 years, 2015-2024 (718 observations). We developed a mechanistic model to realistically represent the browsing process and used regression with mixed models to account for variable vulnerability (damage at a common number of moose per pine tree) among MMAs in the statistical analysis. The model explained 53% of the variation in the proportion of damaged trees and showed that, on average, the relative damage reduction effect of a decreased moose population was ~1.5x larger (25%) than the effect of increased pine density (17%). Vulnerability to browsing varied substantially among MMAs and between years within each MMA, especially in areas with low pine density. This variability prevents reliable predictions of management effects at the individual MMA level for most MMAs. Such local predictions may be improved in the future by incorporating longer time series of observations and additional variables, such as alternative forage sources, browsing by other deer species, and snow cover and duration.
Dataset DOI: 10.5061/dryad.x3ffbg81f
Description of the data and file structure
The data include reported browsing damages from the Swedish browsing damage inventory and estimated moose densities. Further information in the paper.
Files and variables
File: R_code_and_data.zip
File: Estimates_moosedensity_MMA.txt
- The dataset contains a time series for each moose management area (MMA) in Sweden with estimates of the number of moose per 1,000 hectares in the winter population. The estimates were produced by the Swedish University of Agricultural Sciences on behalf of the Swedish Environmental Protection Agency. The estimates are part of official statistics. The unit is the number per 1,000 hectares. 1 moose per 1,000 ha = 0.1 moose per km².
Variables:
Year: year of observation
ID: identification number for the MMA-County combination
MMA: moose management area
Countynr: County number
MMAnr: MMA number
MooseDensity: Estimated number of moose per 1000 hectares in the winter population
File: Timedata-moosedamageMMA.txt
-
The dataset contains a time series for each MMA in Sweden with results from the national moose browsing inventory (Älgbetesinventeringen in Swedish). Approximately 40,000 randomly selected sample plots are inventoried annually in Sweden within MMA areas. The results are presented for each moose management area. The surveyed population consists of young Scots pine trees in stands with an average height between 1 and meters.
Variables:
Region: not used
Countynr: County number
MMAnr: MMA number
MMA: moose management area
Year: year of observation
PropDamagedPines: Proportion of pines damaged in the year of observation in the MMA
PropUndamagedPines: Proportion of pines with no damage ever in the MMa
NumberOfDamagePines: Number of pines damaged thin e year of observation in the MMA
NumberOfUndamaged: Number of pines with no damage ever in the MMA
NumberOfPines: Number of pines in the MMAFile: MMAData.txt
-
The dataset contains land-use data for each MMA as well as the amount of land registered as hunting land in 2024. The unit is hectares.
The variables are:
ID: identification number for the MMA-County combination
MMA: moose management area
Countynr: County number
MMAnr: MMA number
HuntingArea: The size of the area where moose and pine numbers were measured or estimated.
File: Mooseobs_MMA_Tidsserie.txt
- The dataset contains time-series data on observed moose during the first week of the autumn hunting season each year. The unit is the number of observations per hunter-hour. A hunter-hour is the total number of hours spent observing. The data are part of a citizen science program and constitute official statistics. Moose observations per hunter-hour are used to monitor trends in the development of the moose population.
variables:
ID: identification number for the MMA-County combination
MMA: moose management area
Countynr: County number
MMAnr: MMA number
Columns 2015-2034: Observed number of moose each year
File: algskador 2025 for submission English.R
The R code used for the analysis. This script analyzes how moose density and pine density affect the proportion and number of damaged pines across management areas and years.
It first performs area-specific linear models, then fits and compares spatiotemporal mixed-effects models to identify the best overall model.
Finally, it estimates average ecological effects, evaluates model performance, and generates predictions and visualizations of damage patterns.