Feather forensics: tracing the origins of parrots in wildlife trade with stable isotopes and citizen science

Abstract To supply the high demand for wildlife as exotic pets, animals may be illegally and unsustainably harvested from the wild and laundered as captive bred. Consequently, there is considerable interest in wildlife forensic tools that are capable of verifying captive origins. Stable isotope analysis is an emerging tool for verifying captive and wild origins by identifying key differences in dietary intake. While previous studies have effectively classified origins by differences in their stable isotope ratios, these studies are often limited to species with small population sizes and geographic ranges, masking potential variation caused by different environments and diets. We tested the accuracy of stable carbon (δ 13 C) and nitrogen (δ 15 N) isotope analyses to verify captive and wild origins using bird species that are common in pet trade, and have widespread distributions and generalist diets. Through a citizen science project in South Australia, we collected naturally dropped feathers from four native Australian cockatoo (Cacatuidae) species: Galahs ( Eolophus roseicapilla ); and three Cacatua species; sulphur‐crested cockatoos ( Cacatua galerita ), little corellas ( C. sanguinea ) and long‐billed Corellas ( C. tenuirostris ). We compared isotope ratios of captive and wild birds and calculated the classification accuracy of using stable isotopes to determine origin. Stable isotope values were significantly different between captive and wild adult birds, where captive birds had significantly higher δ 13 C and δ 15 N than wild birds. Captive and wild origins of individual Eolophus could be classified with relatively high accuracy (88%). However, Cacatua showed low repeatability and large overlaps between the origin groups, which reduced their classification accuracy (74%). Stable isotope analysis can be a potential classification tool in wildlife trade; however, before on‐ground implementation, we recommend that variation from different diets across a species' geographical range be more thoroughly investigated to better understand and explain the full range of possible δ 13 C and δ 15 N values.