Winter survival of a small predator is determined by the amount of food in hoards
Data files
Jan 07, 2025 version files 406.42 KB
-
capture-histories.Rdata
3.72 KB
-
Kauhava_lentokentta_1.11.2003_-_1.4.2023_lumensyvyys.xlsx
184.61 KB
-
Kauhava_lentokentta_1.11.2003_-_1.4.2023_paivan_keskilampotila.xlsx
206.41 KB
-
README.md
4.84 KB
-
storage-sizes.Rdata
6.43 KB
-
vole-index.Rdata
415 B
Abstract
The hoarding behaviour of animals has evolved to reduce starvation risk when food resources are scarce, but food limitation on survival of hoarding animals is poorly understood.
Eurasian pygmy owls (Glaucidium passerinum) hoard small mammals and birds in natural cavities and nest-boxes in late autumn for later use in the following winter. We studied the relative influence of the food biomass in hoards of pygmy owls on their over-winter and over-summer apparent survival. We also tested whether this influence is modulated by intrinsic (age, sex) traits or extrinsic factors (winter temperature, snow depth).
We measured biomass of prey items in pygmy owl food-hoards during autumns 2003 – 2023 in west-central Finland. We individually marked and recaptured pygmy owls both at nests in the breeding season and at food-hoards. Our dataset included a total of 407 pygmy owls which were all captured from a food-hoard at least once during their capture history. The mean biomass of the annual food-hoards associated with one individual was 443 g (sd = 523 g, range from 3.5 to 4505 g) and was markedly higher in autumns of vole abundance than in those of vole scarcity.
Hoard size had a positive effect on apparent survival of owls over consecutive winter, whereas it did not affect apparent survival over next summer. Hoard size was a better predictor of apparent survival than vole abundance (main food of pygmy owls) in the field. Male owls had higher overall apparent survival rates than female owls, particularly when food-hoards were small.
That hoard size was a better predictor of apparent survival than vole abundance indicates that the hoards are critical for pygmy owls during winter, likely because they are unable to hunt voles below deep snow cover. The positive relationship between apparent survival of owl individuals and their hoard size during winter (when the hoard is being consumed), but not summer, indicates that the hoard size has a true positive effect on survival, and does not only reflect latent inter-individual differences and/or dissimilarities in their environments. We conclude that food limitation during hoarding essentially regulates apparent over-winter survival of pygmy owl individuals.
README: Winter survival of a small predator is determined by the amount of food in hoards
https://doi.org/10.5061/dryad.ksn02v7fk
Description of the data and file structure
This dataset contains data used in the following publication: Korpimaki, E., Piironen, A. & Laaksonen, T. 2025: Winter survival of a small predator is determined by the amount of food in hoards. Journal of Animal Ecology.
The data is a capture-mark-recapture data collected by catching pygmy owls from the nest boxes during spring and fall. Also, the biomass of prey items in the food hoards was recorded as well as relative vole density in the field in every spring and fall. See a detailed description of the methods from the manuscript.
Author information for the data:
Name: Antti Piironen
Institution: University of Saskatchewan, University of Turku
Email: antti.p.piironen@utu.fi, antti.p.piironen@gmail.com
Name: Erkki Korpimaki
Institution: University of Turku
Email: ekorpi@utu.fi
Files and variables
Date of data collection: Years 2003-2022
Geographic location of data collection: Finland
SHARING/ACCESS INFORMATION
1. Links to publications that cite or use the data:
Korpimaki, E., Piironen, A. & Laaksonen, T. 2025: Winter survival of a small predator is determined by the amount of food in hoards. Journal of Animal Ecology.
2. Links to other publicly accessible locations of the data: None
4. Links/relationships to ancillary data sets: None
5. Was data derived from another source: No
FILE OVERVIEW
1. File List:
capture-histories.Rdata
storage-sizes.Rdata
vole-index.Rdata
Kauhava lentokentta_1.11.2003 - 1.4.2023_lumensyvyys.csv
Kauhava lentokentta_1.11.2003 - 1.4.2023_paivan_keskilampotila.csv
2. Relationship between files, if important: No
3. Additional related data collected that was not included in the current data package: None
1. Number of variables: 44
2. Number of cases rows: 555
3. Variable List:
* ring: individual identifier
* sex: sex of the individual
* age: age of the individual (at banding)
* columns from '2003' to '2023': individual encounter histories in spring (1 = recaptured/banded, 0 = not recaptured). The first '1' in each row represents banding.
* columns from '2003-04' to '2022-23': individual encounter histories in fall (1 = recaptured/banded, 0 = not recaptured). The first '1' in each row represents banding.
1. Number of variables: 44
2. Number of cases rows: 555
3. Variable List:
* ring: individual identifier
* sex: sex of the individual
* age: age of the individual (at banding)
* columns from '2003' to '2023': Redundant (all NA).
* columns from '2003-04' to '2022-23': the total biomass of prey items in hoards associated with the individual (in grams) in the fall.
4. Missing data codes: In this file, missing values (=empty cells) represents cases where an individual was not associated to any hoard in a given sampling occasion. As the file is in .Rdata-format, R will automatically correctly read the missing values.
1. Number of variables: 41
2. Number of cases rows: 1
3. Variable List:
* vole-idx: Z-standardized value of relative vole abundance in spring and fall in the 2003-2023. The first value represents the spring 2003 and the last value the spring 2023.
1. Number of variables: 6
2. Number of cases/rows: 6996
3. Variable List:
* weather station: The name of the weather station
* year: Year
* month: Month
* day: Day of month
* time: Local time (HH:MM)
* snow depth [cm]: Snow depth (in centimeters)
4. Missing data codes: Empty cell. Empty cells only occur in the column "snow depth [cm]" and represent missing snow depth measurements.
1. Number of variables: 6
2. Number of cases/rows: 7091
3. Variable List:
* weather station: The name of the weather station
* year: Year
* month: Month
* day: Day of month
* time: Local time (HH:MM)
* temperature [C]: Daily mean air temperature (in Celsius degrees)
4. Missing data codes: Empty cell. Empty cells only occur in the column "temperature [C]" and represent missing temperature measurements.
Code/software
To be able to run the analysis, R software version 4.3.3 or newer is needed. All needed R packages are mentioned in the attached R script (glapas-code.R).