This database presents parental care type for 5438 bird species (~ 50% of extant diversity). The data comes from a previously published study (Cockburn 2006) complemented with more recent published information to classify 351 species initially categorized as unknown in Cockburn (2006). Species were classified as presenting biparental care, female or male only care, cooperative care, or no care. The database includes the species name, the family and order to which it belongs, the parental care type, the source from which the parental care type was obtained, whether the species is placed on the Jetz et al. (2012) phylogeny based on molecular sequence information or not. The datased was used to analyze the most likely ancestral state for parental care type in extant bird species, as well as to describe the most likely transition rates among trait states. 

We used Cockburn’s (2006) review on parental care types across bird species as our primary data source, combined with search of more recent published information for species with missing data. Cockburn (2006) classified species into different parental care categories based primarily on information gathered in reviews and taxon-level monographs, as well as detailed literature searches, the classification is focused on care of eggs and nestlings. Cockburn’s review lacked parental care information for many species, which was inferred based on phylogenetic information i.e., the parental care type of a species’ nearest relative. Although related species tend to have similar parental care types, inferring care type may introduce biases in estimates of parental care because it assumes that the tendency for parental care type to be conserved applies consistently across all birds (Cockburn, 2006). To minimize biases, we did not include species for which parental care type was inferred from phylogenetic information. We cross-referenced species with unknown parental care types in Cockburn’s (2006) review with the Birds of the World Alive (BOW Alive) (Billerman et al. 2020) to check if more recent information on their parental care was available. BOW Alive provides comprehensive and updated information on the behavior and life histories of virtually all extant species of birds, enabling us to increase the number of species included in our analyses, as well as to update Cockburn’s original classification.

Our classification follows Cockburn (2006), who identifies six types of post-oviposition care based on number and sex of caregivers: geothermal heat (avoiding care via the use of geothermal heat to incubate eggs), brood parasite (avoiding care via brood parasitism), male-only care (care provided exclusively by the father), female-only care (care provided exclusively by the mother), biparental care (care provided by both parents), and cooperative care (care provided by more than two individuals). Because species using geothermal heat and those classified as brood parasites were very rare, to ensure an accurate estimation of transition rates, we did not attempt to estimate transition rates to and from these categories, rather we merged them into the “no parental care” category. Some species exhibit non-conventional parental care types, such as suspected or occasional cooperative breeders, as well as species that form family groups without cooperative breeding. However, information for these species is often limited, making it challenging to determine care types without detailed population studies or phenotypic differences between caregivers (Cockburn 2006). Although Cockburn’s classification follows a rigorous and conservative criterion for non-conventional species, we consulted BOW Alive to gather additional breeding information that could aid in their classification. Overall, the breeding information from BOW Alive was congruent with Cockburn’s (2006) classification and thus we followed it. The aforementioned verification for 327 species also gave us confidence regarding Cockburn’s (2006) classification as in all cases it was congruent with information from BOW Alive.

To classify species with unknown parental care types in Cockburn (2006), we used BOW Alive and performed the following steps: 1) extracted all species with unknown parental care and compared their scientific names in Cockburn’s database with those in Jetz et al.’s (2012) phylogeny, removing species that did not match; 2) using the R package rvest(Wickham 2022; https://CRAN.R-project.org/package=rvest) we performed an automated search of the following keywords on each species’ breeding section from BOW Alive: (fe)male, sexes, adult, parents, pair, cooperative, helper, communal, family group, relatives, geothermal, parasite. The list of keywords used for the automated search was based on a revision of parental care information for species for which information was available to identify the words that appeared in those descriptions;  3) we discarded species that did not mention any of the keywords in their breeding information, as these keywords are crucial to determine the number and sex of caregivers, which is essential for assigning a care type; 4) for species for which the breeding information mentioned at least one of the keywords, one of us (PLP) consulted the information to assign a parental care type, focusing on the contribution of each parent in incubation, brooding and/or food provisioning (either mate feeding while incubating and brooding or the direct feeding of young). We did not consider nest construction as parental care since it can serve as a mate attraction behavior. We also disregarded information from captivity or single observations as insufficient evidence for robust classification. Doubtful information (i.e., containing phrases like “suspected” or “not well documented”) was considered invalid unless additional information supported it. Species that breed in colonies, but for which the presence of helpers that contribute to care is unknown, were conservatively classified as biparental when information supported this classification, otherwise, they were discarded. Species exhibiting occasional cooperative breeding, that is species for which observations of cooperative breeding were sporadic or those for which cooperative breeding is described as being uncommon, were also classified as biparental. Examples of the aforementioned cases can be found in some tanagers of the genus Ramphocelus (e.g., Ramphocelus carbo and Ramphocelus dimidiatus), for which there are records of helpers at the nest, but it is an unusual behavior. In total, we assigned a parental care type to 351 species initially categorized as unknown in Cockburn (2006) (see Supplementary Table 14). While this number may appear small, it reflects the limited available information and our cautious approach aimed at minimizing errors.