What are we dealing with? An explicit test reveals different levels of taxonomical diagnosability in the Sylvia cantillans species complex

2010 
Diagnosability is the ability to discriminate between similar taxa, including sibling or cryptic taxa. We have developed an explicit test of diagnosability, using the Sylvia cantillans species complex as a model, which compares an identification based on phenotype with that based on genotype. Individual warblers sampled during their migration in central Italy were first identified to the (sub)species level using putatively diagnostic plumage traits. Nucleotide sequences of a (598-bp) fragment of the mitochondrial (mt)DNA cytochrome b were then used to assign each individual to distinct phylogenetic clades, as determined by reference haplotypes that had been sequenced in breeding individuals. This resulted in the construction of clearly distinct clades corresponding to known taxa of the complex. The new haplotypes were assigned to one of the previously identified groups (corresponding to three different taxa); no sample was assigned outside of them. In contrast, when plumage traits were used to assign the birds into distinct phylogenetic clades, 11 of 58 birds were classified as ‘uncertain/intermediate’ among two taxa, while five were classified differently with the two methods. A perfect agreement between the two methods was found for only for one taxon (Sylvia subalpina, syn. S. moltonii). For the other two taxa of the complex, diagnosability is therefore not guaranteed, and their field identification by hand should be carefully addressed. We provide here an example of an explicit test for establishing the diagnosability of taxa in which two or more ‘markers’ can be used for determining discordant identification and/or unambiguous diagnosability. Our results outline the importance of considering different features for taxa diagnosis and illustrate the weakness of visual appearance-based identification (currently widely used for taxa determination) in our study complex.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    23
    References
    5
    Citations
    NaN
    KQI
    []