RESUMO
The microscopic characters of the hair can be used to indirectly identify species that are either uncommon or are difficult to see, for instance small Neotropical felids of the southern Brazilian Atlantic Forests. This widely used technique, which has not yet been standardized, involves the identification of hairs collected from feces. We tested the hypothesis that this tool is effective to identify four species of small Neotropical felids: Leopardus guttulus, Puma yagouaroundi, L. wiedii and L. pardalis. To accomplish that, we used measurements of the length, width and area of the cuticular scales in the guard hairs and calculated the relationship between width and length (quotient) for each species. A multiple discriminant analysis was conducted on the measurements and the percentage of correct identification was obtained. We found a high overlap in the quotients of these species, which indicates that this technique does not identify Neotropical felids accurately. This inefficiency was also confirmed by the multiple discriminant analysis, where only 74% correct identifications were obtained. Therefore we recommend that hair analysis is used only in combination with other sources of evidence, for instance molecular tools.
Assuntos
Animais , Felidae/anatomia & histologia , Felidae/classificação , Pelo Animal/ultraestrutura , Especificidade da EspécieRESUMO
The microscopic characters of the hair can be used to indirectly identify species that are either uncommon or are difficult to see, for instance small Neotropical felids of the southern Brazilian Atlantic Forests. This widely used technique, which has not yet been standardized, involves the identification of hairs collected from feces. We tested the hypothesis that this tool is effective to identify four species of small Neotropical felids: Leopardus guttulus, Puma yagouaroundi, L. wiedii and L. pardalis. To accomplish that, we used measurements of the length, width and area of the cuticular scales in the guard hairs and calculated the relationship between width and length (quotient) for each species. A multiple discriminant analysis was conducted on the measurements and the percentage of correct identification was obtained. We found a high overlap in the quotients of these species, which indicates that this technique does not identify Neotropical felids accurately. This inefficiency was also confirmed by the multiple discriminant analysis, where only 74% correct identifications were obtained. Therefore we recommend that hair analysis is used only in combination with other sources of evidence, for instance molecular tools.(AU)