Intraspecific variation in group structure arises due to environmentally-mediated directional dispersal in a cooperative breeder
Data files
Aug 06, 2024 version files 4.91 MB
-
master_group_data_dryad.csv
-
pairwsie.fst.longform_dryad.csv
-
populations.snps_dryad.gen
-
README.md
-
shah_and_rubenstein_dryad_2024.07.01.R
-
starlings_inds_dryad.csv
Abstract
Many cooperatively breeding species live in groups with complex social structure—large group sizes, low but mixed kin structure, and multiple breeding pairs—that form as a result of immigration of unrelated individuals in addition to offspring delaying or foregoing dispersal. Differences in patterns of dispersal can therefore generate variation in social structure, even within the same species or population. Yet, the mechanisms underlying inter-individual variation in dispersal and intraspecific variation in social structure remain poorly studied. Here, we examine how environmentally-mediated dispersal patterns inferred from genetic data influence variation in social structure in the superb starling (Lamprotornis superbus), a complex cooperative breeder that inhabits a spatiotemporally variable savanna environment. Using genome-wide polymorphic loci and remotely-sensed, fine-scale ecological data, we find evidence of not only frequent and long-distance dispersal in both sexes (low isolation-by-distance and weak genetic structure), but also directional dispersal from small groups occupying low-quality territories to large groups occupying high-quality territories. Moreover, we find stronger genetic structure among groups in low-quality territories, but higher genetic diversity and lower overall relatedness of groups in high-quality territories. Long-distance, directional dispersal can thus lead to smaller and simpler kin-based social groups in low-quality territories, but larger and more complex mixed-kin groups in high-quality territories. Thus, individual-level dispersal decisions from lower to higher quality habitat that maximize fitness can generate intraspecific variation in social structure, even within the same population. Such within-population variation in social structure could have profound implications for the mechanisms underlying the formation of cooperative societies more broadly.
README
Intraspecific variation in group structure arises due to environmentally-mediated directional dispersal in a cooperative breeder
We have submitted raw data (starlings_inds_dryad.csv, pairwsie.fst.longform_dryad.csv, master_group_data_dryad.csv, and populations.snps_dryad.gen) and R script (shah_and_rubenstein_dryad_2024.07.01.R) belonging to Shah S. S. & Rubenstein D. R. (2024) Intraspecific variation in group structure arises due to environmentally-mediated directional dispersal in a cooperative breeder.
Description of the data and file structure
Metadata for "starling_inds_dryad"
Column | Entry | Value | Explanation |
---|---|---|---|
A | ID | unique ID of the individual | |
B | Group | Group IDs for the 22 superb starling groups part of the study | |
C | Latitude | Latitude | |
D | Longitude | Longitude | |
E | Sex | M/F | Sex of the individual |
F | Age | A, SA | categorical age (A: Adult, SA: Subadult) |
Metadata for "master group data_dryad"
Column | Entry | Value | Explanation |
---|---|---|---|
A | Site | group IDs for the 22 superb starling groups part of the study | |
B | Region | North, Central, South | geographical zone within Mpala Research Center where the group's territory is located |
C | Latitude | Latitude | |
D | Longitude | Longitude | |
E | r_group | -0.001 to 0.19 | average relatedness of the group; ranges from 0 or less = unrelated to 1=relatedness to self |
F | r_females | -0.03 to 0.33 | average relatedness of the females in the group; ranges from 0 or less = unrelated to 1=relatedness to self |
G | r_males | -0.01 to 0.23 | average relatedness of the males in the group; ranges from 0 or less =unrelated to 1=relatedness to self |
H | mean.NDVI2018 | 0.31 to 0.42 | mean normalized difference vegetation index for the year 2018; higher values indicate more green vegetation |
I | cv.NDVI2018 | 0.21 to 0.41 | coefficient of variation of normalized difference vegetation index for the year 2018; higher values indicate more variation in vegetation greenness over the year |
J | by.humans | yes or no | yes = territory overlaps with permanent human settlement, no = territory does not overlap with permanent human settlement |
K | Obs_Het | 0.24 to 0.27 | observed heterozygosity |
L | StdErr_Obs_Het | 0.003 to 0.004 | standard error of observed heterozygosity |
M | Pi | 0.21 to 0.25 | nucleotide diversity |
N | StdErr_Pi | 0.002 to 0.003 | standard error of nucleotide diversity |
O | Fis | -0.07 to 0.004 | inbreeding coefficient; values close to 0 indicate lack of inbreeding |
P | StdErr_Fis | 0.006 to 0.023 | standard error of inbreeding coefficient |
Q | Exp_Het | 0.196 to 0.238 | expected heterozygosity |
R | StdErr_Exp_Het | 0.002 to 0.003 | standard error of expected heterozygosity |
Metadata for "pairwsie.fst.longform_dryad"
Column | Entry | Value | Explanation |
---|---|---|---|
A | Group.1 | group IDs of the first group in the pair | |
B | Group.2 | group IDs of the second group in the pair | |
C | Pairwise.Fst | 0.02 to 0.09 | genetic differentiation between group pairs; ranges from 0 to 1 with higher values indicating more genetic differentiation |
D | Group.1.NDVI | mean NDVI of the first group in the pair | |
E | Group.2.NDVI | mean NDVI of the second group in the pair | |
F | Group.1.terr.qual | L or H | L = lower-than-average; H = higher-than-average |
G | Group.2.terr.qual | L or H | L = lower-than-average; H = higher-than-average |
H | Compairson.type | LL, HH, mix | LL = both groups have lower-than-average NDVI; HH = both groups have higher-than-average NDVI; mix = one group has lower- and the other higher-than-average NDVI |
Metadata for "populations.snps_dryad.gen"
This is a genind object with data on single nucleotide polymorphisms (4,081 loci; 8,162 alleles) for all 235 indvidiauls included in this study. The file can be loaded in R using the function adegenet::read.genepop.
The file structure is as follows:
@tab: 235 x 8162 matrix of allele counts
@loc.n.all: number of alleles per locus (range: 2-2)
@loc.fac: locus factor for the 8162 columns of @tab
@all.names: list of allele names for each locus
@ploidy: ploidy of each individual (range: 2-2)
@type: codom
@pop: population of each individual (group size range: 6-19)
Code/Software
R is required to run shah_and_rubenstein_dryad_2024.07.01.R. The script was created using version 3.6.3.