Female audience shapes the complexity and syntax of male courtship displays in a lek-mating bird
Data files
Dec 23, 2024 version files 1.31 GB
-
0_logistics.r
762 B
-
1_parse_data.r
12.93 KB
-
2_analyze.r
32.41 KB
-
3_plot.r
59.47 KB
-
data_banding.csv
2.80 KB
-
data_raw_2024-12-05.csv
3.70 MB
-
dictionary_behaviors.r
2.77 KB
-
Display_Footage_Examples.zip
1.31 GB
-
README.md
8.30 KB
Abstract
Courtship displays of lek-mating Golden-winged Manakins (Masius chrysopterus) are context-dependent. Presence or absence of a female audience, and female behavior, more than male identity, determines variation in the constituent elements (repertoire) and ordering (syntax) of displays. We analyzed 422 display videos in three contexts: displays without audiences (SOLO, n = 307), displays for female audiences that did not end in copulation (AUDI, n = 102), and displays ending with copulations (COP, n = 13). Using entropy and a metric we call compressibility (ratio of compressed to uncompressed display strings), we found that ordering of elements (syntax) decreased in complexity from SOLO to AUDI to COP displays. Jaro string distance, a record-linkage metric for assessing string similarity, showed that display string syntax corresponded more to audience context than to performer identity. COP displays of individual males differed more from their own AUDI or SOLO displays than from the COP displays of other males. Males responded to female behavior—her position upslope or downslope from him on the display log—with simple COP displays. Courtship displays of Golden-winged Manakins are dynamic interactions between female and male, depending more on male response to female audience behavior, than on traits intrinsic to particular males.
README: Female audience shapes the complexity and syntax of male courtship displays in a lek-mating bird
This README.txt file was updated on 16 December 2024 by L.U.T.
A. Paper associated with this archive
Female audience shapes the complexity and syntax of male courtship displays in a lek-mating bird
B. Originators
David McDonald, University of Wyoming
Liam Taylor, Bowdoin College
Nicholas Oakley, Missouri Department of Conservation
C. Contact information
Liam U. Taylor
Department of Biology, Bowdoin College, Brunswick, ME 04011, United States
l.taylor@bowdoin.edu
D. Dates of data collection
2015-2016 (December, June-August)
2017 (September-December)
E. Geographic Location(s) of data collection
Milpe Bird Sanctuary, Ecuador (~0°1’48” N, 78°57’12” W)
F. Funding Sources
National Science Foundation GRFP (#DGE1752134) for L.U.T.
ACCESS INFORMATION
1. Licenses/restrictions placed on the data or code
Dryad data and files: CC0 waiver..
Zenodo software files: MIT license.
2. Data derived from other sourcse
N/A
3. Recommended citation for this data/code archive
McDonald, David; Taylor, Liam; Oakley, Nicholas (Forthcoming 2024). Female audience shapes the complexity and syntax of male courtship displays in a lek-mating bird [Dataset]. Dryad. https://doi.org/10.5061/dryad.sn02v6xcj
DATA & CODE FILE OVERVIEW
This repository contains all data and code necessary to generate results and figures for the manuscript, "Female audience shapes the complexity and syntax of male courtship displays in a lek-mating bird." This study analyzes the repertoire and syntax of courtship displays in a tropical lekking bird, the Golden-winged Manakin (Masius chrysopterus). By representing displays as strings of individual behavioral elements, we quantify the complexity of these displays and directly compare the syntax of displays from the same males, or different males, across three display contexts: SOLO (male-only displays), AUDI (displays for a female audience), and COP (displays for a female audience that end in copulation).
Data files and variables
data_raw_2024-12-05.csv (Each row is an individual behavioral element, a component of an overall display, output from BORIS https://www.boris.unito.it/)
| Column | Datatype | Description |
|---|---|---|
| UID | (chr) | Unique ID for the display |
| Log | (chr) | The display log (territory) where the display took place |
| Observer | (chr) | Initials of the observer who transcribed display elements |
| Time | (dbl) | Video timestamp of the behavioral element (note decimal precision constrained by BORIS software) |
| Male1ID | (dbl) | Identification for primary displaying male bird |
| FemID | (dbl) | Identification for primary female audience (if present) |
| Bird2ID | (dbl) | Identification for an additional bird, male or female (if present) |
| FemOnOff | (chr) | Tag for female audience ON (="Fem On Log") or OFF (="Fem Off Log") log during display element |
| FemUpDown | (chr) | Tag for female UP (="FemUp") or down (="FemDown") on the log, if the female is ON the log during a display element |
| Behavior | (chr) | Category of behavioral element |
| MaleOtherBeh1 | (chr) | Additional specification for "Other" behaviors |
data_banding.csv (Each row is the band identification information for an individual bird)
| Column | Datatype | Description |
|---|---|---|
| Bands | (chr) | Field-readable color band combination |
| Alum# | (chr) | Suffix to unique aluminum band number |
| Date Banded | (chr) | The date the bird was banded (dd-mm-yy) |
| Sex | (chr) | Observed sex of the bird when banded, based on plumage, male (="M"), female (="F"), or unknown (="Unk") |
| Age | (chr) | Observed age of the bird when banded, based on plumage, green (="G") or definitive (="Def") |
dictionary_behaviors.r (R dictionaries for behavioral element abbreviations and filtering)
| Vector | Description |
|---|---|
| behaviors_cut_partial | elements to cut because they represent just part of a behavioral element |
| behaviors_cut_movement | elements to cut because they represent movement tracking data only |
| behaviors_cut_tracking | elements to cut because they are used for meta tracking only |
| behaviors_cut_other | elements to cut because they do not directly constitute dance display behaviors |
| behavior_short | abbreviated (3-4 chars) names for BORIS display elements |
| behavior_code | single-character codes for display elements |
Code scripts and workflow
Place the data in a Data/ directory. Place the four numbered .r scripts in a Scripts/ directory. Execute run.r in the main directory.
0_logistics.r (set runtime variables, file paths, load packages)
1_parse_data.r (prepare dataset from individual behavioral element records)
2_analyze.r (calculate display characteristics and write analysis report file)
3_plot.r (generate all figures and tables)
SOFTWARE VERSIONS
| Software | Version |
|---|---|
| R | 4.4.1 |
| tidyverse | 2.0.0 |
| lubridate | 1.9.3 |
| acss | 0.2.5 |
| brotli | 1.3.0 |
| stringdist | 0.9.12 |
| patchwork | 1.2.0 |
MOTION DETECTION SOFTWARE
Masius_Movement.zip contains the files for the video motion-detection program used to flag clips featuring bird displays from the full, day-long video recordings taken in the field. The unzipped folder contains its own README with execution instructions and file descriptions.
The motion detection program program is built to run in Python 2.7.12 using OpenCV 3.3.0-dev. Motion segments are flagged from the full video file. The user can then manually confirm the clips, downloading them for subsequent behavioral coding with BORIS software. The annotated, post-BORIS data is already provided in this repository -- the motion detection software is provided for reference, and is not needed to replicate the results of the manuscript.
These files are drawn from GitHub (https://github.com/ltaylor2/Masius_Movement).
EXAMPLE FOOTAGE
Display_Footage_Examples.zip contains a sample of 17 displays from our dataset, including displays performed alone (SOLO), displays for a female audience (AUDI), and displays for a female audience that end in a copulation (COP).
Filenames are given as "Display_
All footage in .mp4 video format.