Abstract

Captive parrots (Psittaciformes) commonly engage in 'feather-damaging behaviour' (FDB) that suggests compromised welfare. Susceptibilities to FDB have been suggested, but not empirically demonstrated, to vary across the more than 200 species kept in captivity. Other demographic risk factors have been proposed for particular species - but neither confirmed nor generalised across Psittaciformes. In this preliminary study, we analysed data from a previously-conducted survey of pet owners: among 538 companion parrots representing 10 non-domesticated, non-hybrid species (n≥17/species), FDB prevalence was 15.8% overall. We tested whether individual FDB status was predicted by four previously-suggested demographic risk factors: species, sex, age, or hatch origin. Available (limited) data on husbandry were assessed as potential confounding variables and controlled for as appropriate. Species identity was a predictor of FDB status (P=0.047), even after controlling for all other variables tested; however, in light of multiple statistical testing, this effect cannot be considered robust until it is replicated. The strongest predictors of FDB status were age (P=0.001; with odds of positive FDB status lower in juveniles versus adolescents or adults [P 0.036]), and sex (P 0.006; with odds of FDB lower in individuals of unknown, versus known, sex [P 0.037]). These findings need to be replicated with data that allow better statistical controls for systematic differences in housing. However, they do provide preliminary empirical evidence for within-species risk factors (suggesting new, testable hypotheses about the etiology of parrot FDB); and for intrinsic, cross-species differences in FDB susceptibility (providing a rationale for future study of the biological factors that might underpin any such taxonomic differences).