Intrinsic factors influence a physiological measure across a forest bird community
Data files
Mar 04, 2025 version files 1.59 MB
-
allbirds.nex
38.80 KB
-
Intr-stress_dataset.csv
135.68 KB
-
phylotrees.nex
1.41 MB
-
README.md
3.36 KB
Abstract
Physiological stress parameters have the potential to serve as valuable early warning indicators for the conservation of animal populations. However, measuring stress in wildlife is often challenging, due to the lack of knowledge about baseline levels, and intrinsic differences between individuals across species. This study is aimed at filling this gap by investigating the influence of intrinsic factors, including sex, age, body condition, and reproductive status on the physiological stress of a forest bird community. For measuring stress levels, we used the heterophil to lymphocyte (H/L) ratio of the bird community, which was assessed using a novel deep learning approach based on Convolutional Neural Networks applied to whole blood smear scans. Using phylogenetically controlled analyses across the bird species, we found higher H/L ratios in adult birds than in juveniles and observed higher stress levels in females than in males. While body condition had no effect on the H/L ratio, reproductive birds tended to have higher H/L ratios than non-reproductive birds, regardless of their sex. Furthermore, we found a robust phylogenetic signal of the H/L ratio in the studied bird community. Our results emphasize the importance of considering intrinsic factors when using stress physiology for assessing the condition of bird populations and communities.
https://doi.org/10.5061/dryad.0cfxpnw8d
Description of the data and file structure
This dataset contains empty cells and NA values, which may be present due to undetermined measurements in birds, missing samples, or low-quality data that prevented further analysis. Please take this into account when working with the dataset. The provided R script in this repository already accounts for these NA values.
Tabellen ID: Individual Identifier for each data row
Year, Month, Day, Date: The sampling day
Net, latitude, longitude: Location
capture-event: p = planned and scheduled capture event, ap = outside of the regular schedule
Success: Whether or not a bird was captured during the capture event
Recapture: Whether or not the bird was a first- or a recapture
Ring_ID: Ring number of the bird which serves as individual identifier
Species_lat, Species_engl: Latin and English species names
sex: The sex of the bird, either morphologically determined or genetically (P2/P8 Primers)
age: age of the bird: dj = hatched this year, ndj = not hatched this year, vj = hatched in the previous year, nvj = older than hatched in the previous year
ad_juv: ad = adult bird (ndj, vj, nvj), juv = juvenile bird (dj)
brood_patch = Whether or not the bird had a brood patch for incubating eggs
family = avian family that the bird belongs to
brutzeit = breeding season, indicates whether or not the capture date was within the breeding season of the bird species
brutgeschlecht = breeding sex, indicating which sex or sexes are incubating the eggs in the species, 1 = only females, 2= males and females
pot_brut = whether or not the bird was potentially breeding based on the species, sex, age, and date
migration = if and how the bird species migrates: SV (not migrating), ZV (migrating), TZ (partially migrating or short distance migrating)
Ticks = whether or not the bird was infested by ticks
wing_mm = wing length in mm
P8_mm = Length of the third outermost primary (P8) in mm
weight_g = weight in g
fat = fat score of the bird (0-8)
muscle = muscle score of the bird (0-3)
KI50_Er = Erythrocyte counts (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_H = Heterophil counts (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_L = Lymphocyte counts (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_E = Eosinophil counts (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_B = Basophil counts (not reliable) (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_M = Monocyte counts (not reliable) (Automatically obtained via deep learning; Vogelbacher et al. 2024)
KI50_P = Bloodparasite counts (not reliable) (Automatically obtained via deep learning; Vogelbacher et al. 2024)
infected = whether or not the bird was infected by blood parasites (Haemosporida)
category = Which blood parasites were found in this bird (genera)
Vogelbacher M, Strehmann F, Bellafkir H, Mühling M, Korfhage N, Schneider D, Rösner S, Schabo DG, Farwig N, Freisleben B. Identifying and Counting Avian Blood Cells in Whole Slide Images via Deep Learning. Birds. 2024; 5(1):48-66. https://doi.org/10.3390/birds5010004
Birds were captured during the breeding seasons between mid-March and August at ten sites in the forest interior. No lures were used to attract the birds to the nets to avoid artificially influencing their natural movement or stress levels. The mist nets were checked every 15 to 20 minutes. Upon capture, each bird was marked with a ring for future identification (e.g., re-captures), with the necessary permissions obtained from the Heligoland Bird Observatory (Institute of Avian Research "Vogelwarte Helgoland" (IAR), Germany). We measured morphological characteristics of each bird and assessed sex, age, and breeding status. Further, a blood sample was taken for identifying the sex if morphological identification was not possible, and a blood smear for leukocyte counts was taken. The blood smears were fixed in methanol and treated with Giemsa stain. To determine the H/L ratio, we digitized the blood smears by scanning them with a Leica Aperio AT2 Scanner and a 40X microscope objective lens and applied a novel Convolutional Neural Network for determining and counting the cells.