Divorce and extra-pair paternity in the Lundy house sparrows
Data files
Oct 24, 2023 version files 1.21 MB
Abstract
The question of why socially monogamous females engage in extra-pair behaviour is long-standing in evolutionary biology. Due to a lack of empirical support among passerine birds, recent work has moved away from the indirect-benefits hypothesis to explain extra-pair mating behaviour by females, instead favouring the hypothesis that this is the result of a pleiotropic effect. That is, a trait under strong positive selection in either or both sexes are genetically linked with another, potentially unrelated, trait. For example, genes beneficial to female fecundity (that promote within-pair solicitation of mating from a male partner) might also lead to extra-pair behaviour (by also promoting solicited copulations from extra-pair males). Here, we test two predictions from this hypothesis: We test the prediction that female divorce, measured as the number of social mates within a given year, is linked with (1) the number of extra-pair males engaged by the female and (2) the proportion of the female’s offspring that are extra-pair. Our results show that females who divorce their social partner are more likely to produce extra-pair offspring than those who maintain social monogamy, supporting the pleiotropy hypothesis. However, those females did not also have a higher proportion of extra-pair offspring. The number of broods initiated was also positively correlated with the number of extra-pair males that sired a female’s offspring, probably through increased opportunity for extra-pair males to sire offspring over a longer breeding season. Our results support the intrasexual pleiotropy hypothesis as a driver of female extra-pair behaviour.
README: Divorce_EPP_data
https://doi.org/10.5061/dryad.8pk0p2ntx
Data contains tables from the Lundy Sparrow Project long-term database which were used to answer questions relating to frequent mate-switching (divorce) and extra-pair paternity. I have also included permutation outputs (which were lengthy) and code files.
Description of the data and file structure
LH_rep is a table of sparrow life history and should be explanatory from the headings - includes information on codes and identifiers used to ID the sparrows and life history traits (when they hatched, and fledged).
Nest_Loc_Pedigree_rep is a table relating the the genetic pedigree of the sparrows used in our analysis and again has full column titles - Dam, Social farther and Cohort.
Suv just denotes if a bird had survived up to specific years. More information can be found in the manuscript
I have also attached two long simulation data files, and an R script showing how they were built.
Variables described in file associated files:
Dam - the ID of females used in the model,
Social farther - the unique ID of the social farther
Cohort - Cohort in which a bird hatched
Loc_code - an identifier for breeding location
-- Note that codes denote a unique nest site location (i.e. W12 is the 12th next box in the Workshop)\, but where nest locations are ephemeral wild nests\, the code just a unique identifier allocated by the fieldworker at the same\, but the first letter denotes the neighbourhood (i.e. W is Workshop). Some codes denote regions on Lundy without standing for anything explicitly. Nest location is not required to replicate models in this study.
-- Note that Sex is a binary measure (0 = female and 1= male)
ID - unique bird ID
BTO - A British Trust for Ornithology ring number
year - date to which a bird was confirmed alive.
EPP_sim - model outputs from MCMC for simulations. For details see MCMCglmm package notes.
NAs (or empty cells) in files denote a gap in information. Where they relate to identifiers (colour rings etc.) the bird was un-ringed/unmarked and in the life history table, they relate to missing information (i.e. bird was not recorded laying any eggs, or recruiting offspring - these may be inferred as zero in those cases, but with the caveat that some data is missing - birds not surviving or breeding in inaccessible places)
Code files are annotated and run in R we used the software packages Tidyverse and MCMCglmm (version 2023) for all data wrangling and analysis.
Sharing/Access information
The data here has taken along time to collect, please contact us if you require access to the expanded data (outside of what was used in this study)