While the wolf is away: Modelling the dynamics of a moose population in a protected area
Data files
Mar 04, 2026 version files 938.86 KB
-
climate_GEE_final.xlsx
779.24 KB
-
Main_code_POM.R
145.03 KB
-
Modules_functions.R
11.31 KB
-
README.md
3.29 KB
Abstract
Populations at high density can threaten the ecological integrity of ecosystems through cascading effects. When such situations arise, management practices must be guided by sufficient knowledge of the biological mechanisms at play. Simulation models are powerful tools for acquiring such knowledge. The moose (Alces alces americana) is a species that recently became overabundant in some areas of eastern North America, sometimes requiring specific management measures. While numerous models exist for moose population dynamics, few of them are adapted to high density populations like the one in Forillon National Park (Quebec, Canada), a protected area in which the moose's apical predator (grey wolf Canis lupus) is absent. We developed a sex- and age-structured population model respecting these conditions that we parameterized using pattern-oriented modelling. The most plausible sequence of vital rates identified exhibited strong negative density dependence in survival, reproduction and dispersal. Predation by alternative predators, black bears (Ursus americanus) and coyotes (Canis latrans), caused substantial mortality of calves each year. Contrary to other areas in northeastern North America, winter tick only had a slight effect on calf survival, except when moose density approached carrying capacity. The variations in the population’s sex ratio were mainly explained by a sex-biased dispersal. Our study provides new insights concerning the dynamics of high-density ungulate populations in the absence of their apical predator, and our modelling approach helped to shed light on new methodological challenges and opportunities. We also present a comprehensive process to build a complex population model and parameterize it while using scarce data.
Access this dataset on Dryad at DOI: 10.5061/dryad.mpg4f4rdj
Description of the data and file structure
This Excel database contains climate variables extracted from Google Earth Engine for our study area. This dataset was used to inform our moose population dynamics model about the interannual variations of some climate variables in our study area for all the modelled period (1982–2020). The variables were selected based on their importance in the moose life cycle (see section 2.2.1 in the main text). These data were downloaded from the ERA5-Land Hourly dataset on Google Earth Engine (GEE), a reanalysis obtained using satellite imagery (Munõz Sabater, 2019).
Files and variables
File: climate_GEE_final.xlsx
Columns A, B and C: Date (Year, Month and Day, respectively). Our dataset spans the period from January 1, 1981, to December 31, 2020.
Column D (snow_d): Average snow depth (m) on the ground in the study area (245 km²) for the day given in columns A to C. These values were obtained from GEE.
Column E (temperature_2m): Average air temperature (K) 2 meters above the ground in the study area (245 km²) for the day given in columns A to C. These values were obtained from GEE.
Column F (Snow_depth): Values of column D converted in cm.
Column G (Mean_T): Values of column E converted in °C.
Column H (no_snow_days): The number of snow-days (days-cm) is a snow accumulation metric defined as « the cumulative daily snow cover of a winter for the whole period when there is snow on the ground » (Potvin & Breton, 1992). For a given day d, the number of snow-days is calculated as Snow_daysd = Snow_daysd-1 + Snow_depthd (formula in our Excel file: {H(x+1) = H(x) + F(x+1)}).
Column I (NIVA_list): This column summarizes all the final values from column H (the number of snow-days) for each winter between 1982 and 2020 (included).
Literature cited
Munõz Sabater, J. 2019. ERA5-Land hourly data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). 10.24381/ods.e2161bao (accessed 25 August 2022).
Potvin, F., Breton, L. 1992. Rigueur de l'hiver pour le cerf au Québec : description de l'indicateur prévisionnel NIVA et présentation d'un logiciel approprié. Ministère du Loisir, de la Chasse et de la Pêche du Québec, Direction de la gestion des espèces et des habitats, Publ. 1936.
Main codes and R scripts
The main code for our model parameterised with the pattern-oriented modelling is presented in the form of an R script (Main_code_POM), as are each of the different modules for calculating vital rates (Modules_functions).
File: Modules_functions.R
Description: This is the R script presenting the code for each of the different modules for calculating vital rates.
File: Main_code_POM.R
Description: This is the R script containing the main code for our model parameterised with the pattern-oriented modelling.
Code/software
Excel for the file entitled "climate_GEE".
R for the files entitled "Main_code_POM" and "Modules_functions".
Access information
Other publicly accessible locations of the data:
This is the only location for the data.