Why preen others? Predictors of allopreening in parrots and corvids and comparisons to grooming in great apes. (16th December 2019)

• Record Type: Journal Article
• Authors: Morales Picard, Alejandra; Mundry, Roger; Auersperg, Alice M.; Boeving, Emily R.; Boucherie, Palmyre H.; Bugnyar, Thomas; Dufour, Valérie; Emery, Nathan J.; Federspiel, Ira G.; Gajdon, Gyula K.; Guéry, Jean‐Pascal; Hegedič, Matjaž; Horn, Lisa; Kavanagh, Eithne; Lambert, Megan L.; Massen, Jorg J. M.; Rodrigues, Michelle A.; Schiestl, Martina; Schwing, Raoul; Szabo, Birgit; Taylor, Alex H.; van Horik, Jayden O.; von Bayern, Auguste M. P.; Seed, Amanda; Slocombe, Katie E.
• Other Names: Massen Jorg guestEditor.
• Is Part Of: Ethology. Volume 126:Number 2(2020)
• Page Start: 207
• Page End: 228
• Publication Date: 2019-12-16
• Subjects: affiliative relationships -- allogrooming -- allopreening -- corvids -- parrots -- primates -- social bonds; Animal behavior -- Periodicals; 591.5
• DOI: 10.1111/eth.12999
• Languages: English
• ISSNs: 0179-1613

Abstract

Abstract: Allogrooming in primates serves not only a hygienic function, but also plays a crucial role in maintaining strong affiliative bonds between group members, which in turn, underpin the emergence of cooperative behavior. In contrast, although allopreening occurs in many avian species, we know little about its social functions. Our study addresses this issue by investigating allopreening in a broad comparative data set including six corvid and nine parrot species. We assessed whether rates of allopreening initiations, proportion of time spent allopreening, and the number of grooming partners in captive group‐housed birds were comparable to patterns observed in captive chimpanzees and bonobos. While parrots and corvids were found to have similar rates of social grooming to bonobos and chimpanzees, Pan species dedicated significantly more time to social grooming. Animals in larger groups had more grooming partners, but when controlling for the number of potential partners, birds tended to have fewer grooming interaction partners than Pan species. We then investigated whether allopreening in parrots and corvids was predicted by behavioral markers of affiliative social bonds (close physical proximity, active feeding, and low levels of agonistic behavior). Results revealed that providing allopreening to a partner was significantly predicted by often being in close proximity, but not engagement in active feeding or agonistic behavior. We examined the region allopreened in a subset of species and found that preening a partner's head was predicted by both close physical proximity and active feeding, while body allopreening was only predicted by close physical proximity. Head preening may confer more hygienic benefits to recipients, and thus, may be more selectively provided to valued partners. Results support the hypothesis that allopreening in corvids and parrots helps maintain social bonds with an individual's most important social partners, showing some similarities to allogrooming in primates. Abstract : Parrots and corvids were more likely to allopreen partners that they chose to maintain close physical proximity with. When examining the region allopreened, we found that preening a partner's head was predicted by frequent close proximity and active feeding, indicating this might be a more valuable service reserved for highly valued social partners. Compared to Pan species, parrots and corvids initiated grooming bouts at similar rates, but invested less time in grooming and engaged with fewer partners. The figure illustrates the probability of dyadic allopreening being observed within a given dyad, as a function of their proximity score. Data illustrated derives from the model testing if occurrence of allopreening in a dyad could be predicted by proximity and active feeding ( N = 118 individuals from nine species). Each dot shows the average probability per bin of the proximity score, whereby the area of the dots depicts the number of dyads per dot ( N = 1 to 1, 005). The dashed and dotted lines depict the fitted model and its confidence interval (with all other predictors in the model being at their average and assuming an observation effort of 4 hr, which roughly equals the average observation effort). Allopreening was more likely to occur in dyads that spent a higher proportion of their time in close physical proximity.