Back

Allonursing (Co-BreeD): Its prevalence, functions and hypotheses across mammals and human cultures

Ben Mocha, Y.; Woith, M.; Scemama de Gialluly, S.; Bruscagnin, L.; Pipper, L.; Kestel, N.; Agustin Bechayda, S.; Chaudhary, N.; Clarke, C. M. K.; Garber, P. A.; Gettler, L. T.; Pilastro, A.; Quintana, R.; Radford, A. N.; Richner, H.; Rosenbaum, S.; Santos, E. S. A.; Michael Griesser, M.

2026-01-12 animal behavior and cognition
10.64898/2026.01.10.698751 bioRxiv
Show abstract

Allonursing, where dependent young are nursed by females other than their mothers, has been observed across diverse mammalian species and human cultures. Yet, the proximate functions underlying this seemingly altruistic behaviour, and its benefits for the offspring, the mother, and the nursing allomother, remain poorly understood. To facilitate testing hypotheses concerning these questions, we compiled the to date largest allonursing dataset from wild populations (N = 101 species, including 133 populations and 5 human cultures). Using this dataset, we (i) map the taxonomic distribution of allonursing, thereby expanding the list of mammal species qualified as cooperative breeders by 17 additional species, (ii) differentiate cases of non-voluntary from voluntary allonursing and consequently reject the hypothesis that milk stealing explains regular allonursing in most species, (iii) quantify the within-population extent of allonursing to assess its significance across species and evaluate the energy/time saved for mothers, (iv) use phylogenetically controlled analyses to revisit the association between allonursing and polytocy (i.e., litter size >1), and provide an alternative explanation for this association. We conclude by proposing that allonursing has multiple functions, with the most empirically supported ones being provisioning and insurance against maternal loss. Finally, the presented AlloNursing dataset complements the Cooperative-Breeding Database (Co-BreeD) to further expand this integrative resource for cooperative breeding research.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.