In many birds males are presumed to protect their paternity by closely guarding their mate or copulating frequently with her. Both these costly behaviours are assumed to reduce the risk and/or intensity of sperm competition. However, despite many studies on avian extra-pair paternity, it remains unclear how strongly these behaviours are related to fitness and other key life-history traits. Here we conduct meta-analyses to address two questions. First, are mate guarding and/or frequent copulation positively correlated with a male's share of paternity at his nest? We find a significant positive correlation between both presumed paternity protection behaviours and paternity share. The relationship is, however, weak (r = 0.08-0.23). This is perhaps unsurprising if the risk of partner infidelity, hence the need to protect paternity, varies among males. For example, more attractive males might have less need to protect their paternity. Second, do males with higher indices of so-called male ‘quality’ (phenotypic measures, usually subjectively defined by researchers as predictors of male attractiveness) exhibit lower levels of paternity protection behaviour? We find a negative correlation between male quality and paternity protection. This finding might partly explain the weak relationship between paternity protection and paternity, although we discuss other, non-mutually exclusive possibilities.
Q1All
Standardized effect sizes (Fisher's Z) for the relationship between paternity protection behaviours and realized paternity, across all categories of paternity protection. Multiple effects from the same study are combined as a weighted mean effect so that each study contributes a single effect to this dataset. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q1Cat
Standardized effect sizes (Fisher's Z) for the relationships between each category of paternity protection behaviour and realized paternity. Multiple effects from the same study are combined as a weighted mean effect so that each study contributes a single effect per paternity protection category to this dataset. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q1phylo
Standardized effect sizes (Fisher's Z) for the relationships between paternity protection behaviours and realized paternity. For use in multi-level & phylogenetic meta-analyses: a single study could contribute multiple (non-independent) effects for the same relationship, and study ID and species are coded as moderator variables. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q1 R code
R script for standard and multilevel (phylogenetic) meta analyses of the relationship between paternity protection behaviours and realized paternity.
Q1 standard and multilevel.R
Q2All
Standardized effect sizes (Fisher's Z) for the relationship between paternity protection behaviours and measures of male quality, across all categories of paternity protection and quality. Multiple effects from the same study are combined as a weighted mean effect so that each study contributes a single effect to this dataset. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q2Cat
Standardized effect sizes (Fisher's Z) for the relationships between categories of paternity protection behaviours and different measures of male quality. Multiple effects from the same study are combined as a weighted mean effect so that each study contributes a single effect per combination of categories to this dataset. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q2phylo
Standardized effect sizes (Fisher's Z) for the relationships between paternity protection behaviours and measures of male quality. For use in multi-level & phylogenetic meta-analyses: a single study could contribute multiple (non-independent) effects for the same relationship, and study ID and species are coded as moderator variables. Note that all raw effects reported in the included studies are available online in the supplement to the paper.
Q2 R code
R script for standard and multilevel (phylogenetic) meta analyses of the relationship between paternity protection behaviours and measures of male quality.
Q2 standard and multilevel.R
Ericson tree
Phylogeny of the species present in the included studies, following the 'Ericsson tree' (see Jetz et al. 2012; birdtree.org)
Ericson.tre
Hackett tree
Phylogeny of the species present in the included studies, following the 'Hackett tree' (see Jetz et al. 2012; birdtree.org)
Hackett.tre