Michener’s group-size paradox in cooperatively breeding birds
Data files
Jul 30, 2024 version files 1.11 MB
-
Cooperative_bird_literature.pdf
621.54 KB
-
MP_Data_Extraction.txt
56.63 KB
-
MP_R_Code.R
30.08 KB
-
MP_Supp_Methods_Figures.pdf
350.67 KB
-
MP_Table_S1.csv
30.61 KB
-
MP_Table_S2.csv
16.56 KB
-
MP_Table_S3.csv
1.31 KB
-
README.md
5.03 KB
Abstract
According to Michener's Paradox, most altruistic groups in nature should be small, and large groups should not exist. This is because per capita productivity is thought to decrease as groups get larger, meaning that the share of indirect fitness available to each group member declines, which favours dispersal. The empirical evidence for a decrease in per capita productivity is contradictory, however, and limited to the social Hymenoptera. I report that per capita reproductive success decreased with increasing group size across 26 cooperatively breeding bird species. Small groups comprising two or three individuals were the most common (79% of 16,101 groups) and these had the highest per capita reproductive success. This close fit between per capita reproductive success and the distribution of group sizes in nature suggests that it may indeed be difficult for large groups to evolve through indirect fitness benefits alone.
Author: Philip A. Downing
Contact: philip.downing@oulu.fi
Number of supplementary items: seven
1. MP_R_Code.R
2. MP_Table_S1.csv
3. MP_Table_S2.csv
4. MP_Table_S3.csv
5. MP_Data_Extraction.txt
6. Cooperative_bird_literature.pdf
7. MP_Supp_Methods_Figures.pdf
File name: MP_R_Code.R
This R script contains all the code needed to replicate the analyses (including packages and functions).
Models cited in the manuscript are located in this code.
The lettering in the code matches the lettering used in the methods section of the manuscript.
- Data manipulation (lines 25 to 100)
- Part A. Per capita and total reproductive success across group sizes (lines 106 to 256)
- Part B. Per capita and total reproductive success and the frequency of group sizes (lines 262 to 434)
- Part C. Study effort (lines 440 to 627)
Note that the code includes two-stage models fit using restricted maximum likelihood in the metafor R package to investigate the sensitivity of the results to Bayesian methods (results reported in Table S3).
See Viechtbauer, W. (2010) https://www.metafor-project.org/doku.php/tips:two_stage_analysis
File name: MP_Table_S1.csv
This CSV document contains:
+ reproductive success estimates (annual number of fledglings) in different sizes groups for 26 species
+ the data are organised in a long format (137 rows)
+ read into R (some of the column headings will need changing to match the R code)
+ column descriptions:
A. common name = English name of each species
B. Latin binomial = scientific name of each species (matches the Jetz et al. nomenclature)
C. measure = the metric used to measure annual reproductive success in each study
D. mean = estimate of mean reproductive success
E. standard_dev = standard deviation of the mean reproductive success estimate
F. variance = variance of the mean reproductive success estimate
G. N = sample size used to estimate mean reproductive success
H. group_size = the group size pertaining to each mean reproductive success and % of groups estimate
I. source = where the mean reproductive success estimates were extracted from in the study
J. reference = study from which the mean reproductive success estimates were extracted
K. % of groups = the frequency of each group size in nature
L. N = sample size used to estimate % of groups
M. source = where the % of group estimates were extracted from in the study
N. reference = study from which % of groups estimates were extracted
File name: MP_Table_S2.csv
This CSV document contains:
+ the frequency of different group sizes in natural populations for 23 species
+ data are absent for three species in Table S1: chestnut-crowned babbler, toucan barbet, Pygmy Nuthatch
+ the data in columns D and E are entered in columns L and K in Table S1, but only for group size with reproductive success estimates
+ the data are organised in a long format (183 rows)
+ read into R (some of the column headings will need changing to match the R code)
+ column descriptions:
A. common name = English name of each species
B. Latin binomial = scientific name of each species (matches the Jetz et al. nomenclature)
C. group_size = the group size pertaining to each % of the group estimate
D. N = sample size used to estimate % of groups
E. % of groups = the frequency of each group size in nature
F. source = where the % of group estimates were extracted from in the study
G. reference = study from which % of groups estimates were extracted
File name: MP_Table_S3.csv
This CSV document contains:
+ parameter estimates from statistical models (see MP_R_Code.R for further information)
+ 34 rows = output from 8 statistical models
+ includes results from the two-stage models (see MP_Supp_Methods_Figures.pdf)
+ variable names:
beta = parameter estimate (slope) from the model
lwr CI = lower 95% credible interval from the posterior distribution of the model
upr CI = upper 95% credible interval from the posterior distribution of the model
File name: MP_Data_Extraction.txt
This plain text document contains details of the figures, tables, and text fragments from which data were obtained, and the calculations used to pool data, organised alphabetically by Latin name.
File name: Cooperative_bird_literature.pdf
A database of 879 studies on 182 species of cooperatively breeding birds, organised alphabetically by Latin name.
Comprises all known studies (published research, MSc and PhD theses, monographs, and edited volumes) with data on the breeding biology and fitness parameters of cooperative birds.
File name: MP_Supp_Methods_Figures.pdf
This PDF document contains further details on the methods used in the study (prior specification and model convergence, quadratic effects, and study effort) and supplementary figures S1 to S5.