Table 1.
Details regarding potential predictors of behavioural and reproductive problems in captive Psittacines: data sources, how quantified and the effects predicted by each hypothesis under test.
| hypothesis: behavioural and reproductive problems in captivity reflect… | predictor variable(s) | predicted relationship with behavioural and reproductive problems if hypothesis is correct |
|---|---|---|
| … constraints on natural social interaction | Data on social bond strength/number were unavailable, and so group size was used as a proxy (cf. [38]). Maximum feeding group size: maximum feeding group size during the non-breeding season, a measure of sociality while active [36,66]. | positive |
| Communal roostinga: whether or not a species roosts communally while sleeping [36]. | yes > no | |
| … restriction of natural foraging behaviours | Frustrating appetitive and consummatory aspects of foraging can give rise to stereotypic behaviour [67,68]. Quantitative wild time budget data were unavailable, so time spent on food-search and/or -handling [cf. 69] in the wild was inferred from diet-types reported in the EltonTraits avian foraging database [70] and other sources [65]. | |
| % natural diet requiring extensive locomotor/visual food search: reliance on food items that are naturally patchily distributed in space and time, and/or scarce or inconspicuous (based on [71–73] and adapted here in line with EltonTraits’ dietary categories), namely invertebrates, nectar and pollen, fruit, and tree seeds/nuts. | positive | |
| % natural diet requiring extensive food handling: reliance on food items requiring extensive oral manipulation (based on [71–73] and adapted here), namely invertebrates and tree seeds/nuts. (Note that Wilman et al. [70] pooled reliance on small/grass seeds and tree seeds/nuts into one category, “seeds”, but we made opposing predictions about each regarding welfare. Therefore, by referring to their literature sources [65], and following their methods, we split each species’ reliance on seeds/nuts proportionally between small/grass seeds and tree seeds/nutsb). | positive | |
| … being a habitat or dietary generalist | Diet breadth: total number (1-5) of main food types in the species-typical native adult diet [16,65], from: seeds/nuts, fruits/berries, pollen/nectar, other vegetative material, and animal material [58,59]. | Negative or positive depending on whether generalism is protective or a risk factor |
| Habitat breadth: total number (1-7) of main habitat types in the native range, from: mixed lowland forest, alpine scrub and forest, grassland and savannah, mixed scrub, marsh and wetland, cultivated and farmland, and urban environments [59]. | ||
| … being intelligent and innovative | Innovation frequency: total number of feeding innovations reported by regions [74–77], supplemented with unpublished data (Louis Lefebvre, pers. comm., 2013)c. Relative brain volume [encephalization] was used as a proxy for intelligence ([78–81]; see additional information in Discussion). A species’ average endocranial volume (ml) taken from skeletal specimens or converted from brain mass [82,83]d. | Negative or positive depending on whether intelligence is protect-ive or a risk factor |
| … being rare/threatened | IUCN Red List category: status in the International Union for Conservation of Nature Red List of Threatened Species [87] (ranked 1-5): Least Concern, Near Threatened, Vulnerable, Endangered, Critically Endangered. | Positive |
| Population sizes in US aviculture systems: number of pairs per species, taken from [9]. | Positive (N.B. only reproductive problems assessed here) |
aThe one categorical (rather than continuous) predictor.
bTwo species’ diets were recorded incorrectly in EltonTraits: the black-headed parrot Pionites melanocephalus, was coded as using 60% nectar, yet its source account [65] did not mention it using nectar but rather tree seeds; and the dusky parrot, Pionus fuscus, was coded as 100% fruit but according to the source [65] it uses seeds from two trees. On advice from EltonTraits’ authors (Y. Belmaker, pers. comm., 2020), we corrected these entries for our dataset.
cResearch effort (number of published papers on a given species) was included in all models to control for differential interest by birdwatchers [76], calculated from results of searches of species names (scientific and common) in the ‘Topic’ field of the Zoological Records web index (Thomson Reuters) between 1978-2004 [76,77].