Dandelion and vole interaction
Data files
Oct 07, 2025 version files 24.52 MB
-
Colonisation_data.txt
14.21 MB
-
Dandelion_vole_interaction.R
25.37 KB
-
Depletion_and_persistence.txt
10.20 MB
-
Depletion_colony_scale.txt
74.98 KB
-
Quadrat_position.csv
2.04 KB
-
README.md
7.08 KB
Abstract
Like many rodents, the water vole is able to reach high densities in meadows. During outbreaks, voles cause significant changes in plant communities. Although water voles consume a wide variety of plant species, dandelions have a unique position: they are selected by voles year-round and serve as a key resource during winter. Voles harvest all parts of the dandelion and store the roots in almost monospecific food stores. As dandelions are perennial plants that take years to grow, vole activity can significantly affect dandelion populations. Our aim was to estimate the influence of dandelion density on vole space use, particularly habitat selection during natal dispersal. We tested the hypothesis that voles select dandelion-rich plots for settlements. We also measured the variation in dandelion density due to new colony settlements to assess potential feedback effects. We hypothesized that voles decrease dandelion populations. To achieve that, we used a drone to monitor dandelions and voles over two years. We monitored 52 quadrats, each half a hectare, three times a year. We analysed each image using remote sensing to locate voles and dandelions, and then examined the interactions between their locations over time. We found that dandelion-rich plots were more likely to colonize. In plots with low dandelion density, areas denser than the plot average were also more likely to be colonised. We observed a decrease in the number of dandelions after colony settlement. Finally, we found evidence that existing burrows were more likely to be reused by new voles if dandelions were still present. This study demonstrates that dandelion density is a key criterion in habitat selection for water voles and that vole colonies rapidly deplete this resource after establishment. These findings provide insight into plant-herbivore interactions and offer valuable perspectives for further exploration of the plant hypothesis, particularly with respect to the dynamics of resource availability and its role in cyclic population fluctuations.
https://doi.org/10.5061/dryad.8w9ghx3xs
Description of the data and file structure
The aim of this study was to investigate the influence of dandelion density on vole space use, particularly on habitat selection during natal dispersal. Data were collected in the Massif Central (France) in 2021 and 2022 and in the Jura Mountains in 2022. 52 quadrats were monitored by drone 3 times per year to map the locations of water voles and dandelions from aerial pictures (in March and September for vole surface indices and in May for dandelions). Each image is divided into 2 m × 2 m tiles, resulting in a total of 1200 tiles per quadrat. In each tile, we compared the influence of dandelion density on the probability of colonization or reuse of an old colony. At a second time, we observed the depletion of dandelions by voles at broad and fine scales.
Files and variables
File: Dandelion_vole_interaction.R
Description: Script used for the analysis.
File: Depletion_colony_scale.txt
Description: Database of information obtained by remote sensing in 46 subplots (20 x 20 m). Each mound were surrounded by 20 buffers at a distanc between 5 cm and 5 m. In each buffer the number of dandelion flowers was counted. This database is used for dandelion depletion by vole at fine scale analysis.
Variables
- Modality: Code with the region of the quadrat and the year of the picture ( MC21 = Massif Central in 2021, MC22 = Massif Central in 2022, D22 = Doubs in 2022)
- Quadrat: ID of the quadrat use for the extraction of the subplot
- Extraction: ID of the extraction (number of the subplot)
- Distance: radius of the buffer around the mound
- Density: mean density of dandelion flower heads within the buffer area
- Mean: mean density of flower heads of the subplot
- Density_comp: comparison of the dandelion density whithin the buffer and the all subplot
File: Colonisation_data.txt
Description: Database for the analysis of the influence of dandelion density on new colony etablishment. This database obtained by remote sensing of 52 plots in 2021 and 2022 in Doubs and Massif central. Each plot was subdivided in 1200 tiles of 4 m².
Variables
- Modality: Code with the region of the quadrat and the year of the picture ( MC21 = Massif Central in 2021, MC22 = Massif Central in 2022, D22 = Doubs in 2022)
- Quadrat: ID of the quadrat
- num_tile: ID of the tile after the 4 m² grid was applied
- Plot_yellow_rate: rate of yellow pixel in the plot in May
- Plot_brown_rate_march: rate of brown/ground color pixel in the plot in March
- Sum_occ_ tile_march: Number of tiles considered as colonized by vole in March
- Stade_physio: mean number of flower heads of dandelion by plant
- Plot_quality_march: degree of confidence of the observation in March from A = the best to C= the worst
- Plot_quality_may: degree of confidence of the observation in May from A = the best to C= the worst
- Tile_brown_rate_march: rate of brown/ground color pixel in the tile in March
- Tile_yellow_rate: rate of yellow pixel in the tile in May
- Plot_brown_rate_sept: rate of brown/ground color pixel in the plot in September
- Sum_occ_tile_sept: Number of tiles considered as colonized by vole in September
- Plot_quality_sept: Degree of confidence of the observation in September from A = the best to C= the worst
- Tile_brown_rate_sept: rate of brown/ground color pixel in the tile in September
File: Quadrat_position.csv
Description: Geographical coordinates of the centroids of each plot
Variables
- Name_Quadrat: Quadrat ID
- Coord_X: X coordinate
- Coord_Y: Y coordinate
File: Depletion_and_persistence.txt
Description: Database for the analysis of the influence of dandelion density on persistence and depletion (broad scale analysis). This database is a time serie 2021 and 2022 of the 35 plots in Massif central. Each quadrat was subdivided in 1200 tiles of 4 m².
Variables
- Quadrat: ID of the quadrat
- num_tile: ID of the tile after the 4m² grid was applied
- Plot_yellow_rate_21: rate of yellow pixel in the plot in May 2021
- Plot_brown_rate_march_21:rate of brown/ground color pixel in the plot in March 2021
- Sum_occ_tile_march_21: Number of tiles considered as colonized by vole in March 2021
- Stade_physio_21: mean number of flower heads of dandelion by plant in May 2021
- Plot_quality_march_21: degree of confidence of the observation in March 2021 from A = the best to C= the worst
- Plot_quality_may_21: degree of confidence of the observation in May 2021 from A = the best to C= the worst
- Tile_brown_rate_march_21: rate of brown/ground color pixel in the tile in March 2021
- Tile_yellow_rate_21: rate of yellow pixel in the tile in May 2021
- Plot_brown_rate_sept_21: rate of brown/ground color pixel in the plot in September 2021
- Sum_occ_tile_sept_21: Number of tiles considered as colonized by vole in September 2021
- Plot_quality_sept_21: Degree of confidence of the observation in September 2021 from A = the best to C= the worst
- Tile_brown_rate_sept_21:rate of brown/ground color pixel in the tile in September 2021
- Plot_brown_rate_march_22: rate of brown/ground color pixel in the plot in March 2022
- Sum_occ_tile_march_22: Number of tiles considered as colonized by vole in March 2022
- Plot_quality_march_22: degree of confidence of the observation in March 2022 from A = the best to C= the worst
- Tile_brown_rate_march_22: rate of brown/ground color pixel in the tile in March 2022
- Plot_yellow_rate_22: rate of yellow pixel in the plot in May 2022
- Stade_physio_22: mean number of flower heads of dandelion by plant in May 2022
- Plot_quality_may_22: degree of confidence of the observation in May 2022 from A = the best to C= the worst
- Tile_yellow_rate_22: rate of yellow pixel in the tile in May 2022
- Plot_brown_rate_sept_22: rate of brown/ground color pixel in the plot in September 2022
- Sum_occ_tile_sept_22 : Number of tiles considered as colonized by vole in September 2022
- Plot_quality_sept_22: Degree of confidence of the observation in September 2022 from A = the best to C= the worst
- Tile_brown_rate_sept_22: rate of brown/ground color pixel in the tile in September 2022
Code/software
All analyses were performed using the R environment (version 4.2.2).
Packages to download to use the script :
library(lme4)
library(MuMIn)
library(boot)
library(tidyverse)
library(ggpubr)
library(mgcv)
library(itsadug)
library(lmerTest)
library(visreg)
library(extrafont)
library(wesanderson)
The script is divided into 4 parts and each part contains:
- importating and formating data
- checking the data
- the models + choice of reference quadrat
- creating graphics
Run the part # COLONISATION to find the figure 3A
Run the part # REUSE to find the figure 3B
Run the part # DEPLETION: population scale ---- to find the figure 4A
Run the part # DEPLETION: subplot scale ---- to find the figure 4B
All model are summarize in table 1
