Reanalysis of 73 cooperative breeders finds assumed female skew was irrelevant, while co-breeding occurs when multiple mature females share a group

Reproductive skew among females is a ratio. It describes how breeding is distributed among the sexually mature females sharing a group, so it needs at least two of them to mean anything. Western bluebirds and superb fairywrens had been scored in earlier comparative work as extreme cases of female skew, and those scores were used to explain why subordinate females give up breeding to help a dominant female ^[5]. "So, by definition there cannot be female reproductive skew in these species, and there is also no female alloparental care," Ben Mocha told Phys.org ^[7]. Those facts, he said, make the species "irrelevant for answering questions about female reproductive cooperation" ^[8].

What the groups actually contained is what bears on the theory. "In most cases, allomothers were sexually immature females or there were only male helpers, so no female really forwent self-breeding," Ben Mocha said ^[9]. A yearling feeding her mother's chicks has not declined a breeding attempt she was able to make. A group with only male helpers poses a question about males.

Where adult females did share a group, the pattern ran the other way. "We found that when multiple sexually mature females lived in the same group, they more often co-bred and helped to raise each other's offspring, rather than only suppressed each other's breeding," Ben Mocha said ^[10]. The comparison there is between co-breeding and suppression among adult females who actually co-reside, and it does not establish that competition is absent. The Phys.org report does not say how many populations lie behind that pattern ^[18].

The authors put the methodological claim first. Ben Mocha said the work began with "finding more and more comparative analyses that classified 'species' in a superficial way and without paying attention to important nuances in the social system of the species (including between-population differences)" ^[12]. Co-BreeD is built to match: "Co-BreeD provides data that is specific for a species, living in a specific location (i.e., population) in a specific time period," he said, and "In this way, we avoid generalization to the 'species' level and allow accounting for within-species variation" ^[13][14]. The error demonstrated in this account, though, sits inside groups: which classes of animal were present, and whether any adult female could have forgone breeding at all ^[9][6]. On between-population variation, this account carries the database design and the authors' stated rationale. A paired comparison of two populations of one species would have to come from the paper itself.

In my view the methodological objection is the durable result. The team checked 73 cooperatively breeding bird and mammal species against the reported composition of their groups ^[4], and the cases that failed the check were the ones supporting theoretical claims about female altruism ^[5]. Ben Mocha framed that as the second goal of the paper: "to demonstrate the misleading effect of ignoring the social composition of animal groups when 'classifying' species into social categories and using this classification to test theoretical models" ^[11]. Co-BreeD entries are drawn mainly from peer-reviewed studies of wild populations and are reviewed by the project curators and by the original authors of each study ^[16]. That check runs on the primary data, so coverage still depends on which populations someone went out and studied. The database was started by Ben Mocha and Michael Griesser as a standardized, open-access resource for comparative work on cooperative breeding ^[15].