- Redondo, Tomás
- Parejo-Pulido, Daniel
This package contains three Excel workbooks plus six accompanying R scripts documenting field and laboratory data and analyses on multicomponent nestling begging displays in spotless starlings.
The core raw datasets are (i) "Natural_broods.xlsx", a parental visit-level table from natural nests (4 188 rows) in which each row represents a nestling’s behaviour and feeding outcome during a parental visit; and (ii) "Laboratory_nestlings.xlsx", which includes a bout-level postural dataset (300 rows; one row per chick × deprivation treatment × stimulus) and a call-level acoustic dataset (1 904 rows; one row per call with within-bout sequencing and bout identifiers).The file "Data_figures.xlsx" provides processed, figure-ready tables used to plot the manuscript figures and is intended for direct reuse in visualisation or verification.
The six R scripts correspond to (1) the three analysis streams: (a) mixed-model effect sizes (R²-based)(scripts "R2_Models_/), (b) information-theoretic measures of coding effectiveness (scripts "Information_/"), and (c) receiver-based decoding via random-forest classification (script "Random_forest_laboratory_nestlings"; plus (2) One script for figure plotting.
All scripts assume that the Excel data files are available in the working directory (or that paths in the scripts are edited accordingly)., [Description of methods used for collection/generation of data] Data collected to quantify how multicomponent begging displays encode nestling hunger (food deprivation) in spotless starlings, using (i) observations of natural broods at nest boxes and (ii) a controlled laboratory “hunger experiment”.
1 Experimental efforts and data collection
1.1 Study system and timing
Fieldwork was conducted on a nest-box population of spotless starlings in central Spain. Begging behaviour was recorded when nestlings were 6 days old, a stage of rapid growth when nestlings obtain virtually all food through begging and physical interference among nestmates is still limited.
1.2 Observations at natural nests (field component)
Sample: 24 first broods containing 4–5 six-day-old chicks.
Recording context: early-morning feeding peak, using infrared video cameras.
Pre-recording standardisation: nestlings were fed to satiation with mealworms, weighed, and individually marked.
Behavioural scoring: a trained observer (blind to the study aims) reviewed the first three hours of footage after the first parental visit.
Key variables recorded in the field included: timing of parental visits, identity of fed chicks, maximum postural begging intensity, begging rank order, time since previous feeding (deprivation time), and the size of the last prey item (relative to bill length).
1.3 Laboratory hunger experiment (controlled component)
Sample: 25 first-brood nests; two intermediate-mass-ranked chicks per brood were transported to the laboratory the day before trials.
Housing: artificial nest cups under controlled temperature and humidity.
Conditioning: nestlings were habituated to a standard begging stimulus (playback of a parental feeding call from the same population) and fed on demand during multiple conditioning trials.
Feeding schedule before tests: four routine feeding trials at 45-min intervals (ration adjusted to nestling mass).
Hunger manipulation and recording: after the last meal, begging was recorded at 30, 60, and 90 min. Each trial comprised two playback stimuli, separated by sufficient time for begging to extinguish.
Recording equipment: synchronized video (posture) and close-range audio (vocalisations).
Post-trial handling: nestlings were fed ad libitum, blood-sampled for sexing, and returned to their original nests.
1.4 Behavioural and acoustic measurements
Postural begging was scored on an ordinal 0–4 scale:
0 = no begging
1 = gaping, neck flexed, body horizontal
2 = gaping, neck extended, body horizontal
3 = gaping, neck extended, body raised, back reclining
4 = fully stretched, vertical back (sometimes with wing flapping)
Laboratory videos were additionally used to quantify display fine structure (e.g., time spent at each postural level), from which total begging duration, mean postural intensity, maximum postural intensity, time at maximum intensity, and latency to beg were derived.
Begging vocalisations were analysed from spectrograms to obtain temporal measures (e.g., bout duration, number of calls, calling latency) and call-level spectral/entropy measures. Some spectral traits were summarised via PCA to produce single composite variables.
Files and variables
Data
Three Excel workbooks (.xlsx). Each workbook includes a sheet named Metadata containing the authoritative variable definitions (and, where relevant, abbreviations and units). Below, I summarise the contents of each file and list all variables used., [Methods for processing the data] Accompanying R code., This package contains three Excel workbooks plus six accompanying R scripts documenting field and laboratory data on multicomponent nestling begging displays in spotless starlings. The core raw datasets are (i) Natural_broods.xlsx, a visit-level table from natural nests (4 188 rows) in which each row represents a nestling’s behaviour and feeding outcome during a parental visit; and (ii) Laboratory_nestlings.xlsx, which includes a bout-level postural dataset (300 rows; one row per chick × deprivation treatment × stimulus) and a call-level acoustic dataset (1 904 rows; one row per call with within-bout sequencing and bout identifiers). Variables include identifiers (nest, chick, parent), deprivation (time since last feeding or controlled deprivation), ordinal maximum postural intensity (0–4), timing traits (latency, bout duration), feeding outcome (0/1), and multiple acoustic descriptors (temporal measures and frequency/entropy metrics, including PCA-derived composites). Data_figures.xlsx provides processed, figure-ready tables used to plot the manuscript figures and is intended for direct reuse in visualisation or verification. The R code for statistical analyses focuses on (a) mixed-model effect sizes (R²-based), (b) information-theoretic measures of coding effectiveness, and (c) receiver-based decoding via random-forest classification., Ministerio de Ciencia, Innovación y Universidades, PGC2018-099596-B-I00
Agencia Estatal de Investigación, MCIN/ AEI/10.13039/501100011033
Ministerio de Ciencia y Universidades, CGL2017-83843-C2-1-P
Ministerio de Ciencia y Universidades, PID2021-126673NB-I00, File List: #Data All data are provided as three Excel workbooks (.xlsx): "Natural_broods.xlsx" "Laboratory_nestlings.xlsx" "Data_figures.xlsx" #Code Code is provided as six R scripts (.R): "R2_Models_laboratory_nestlings.R" "R2_Models_natural_broods.R"" "Information_laboratory_nestlings.R" "Information_natural_broods.R" "Random_forest_laboratory_nestlings.R" "Figures.R", Peer reviewed