So-called DNA barcodes have recently been proposed to answer the problem of specimen identification and to quantify global biodiversity. We show that this proposition is wanting in terms of rationale, methodology and interpretation of results. In addition to falling short of all its stated goals, the method abandons the benefits of morphological studies in favor of a limited molecular identification system that would ultimately impede our understanding of biodiversity. Ó The Willi Hennig Society 2004.''If the only tool you have is a hammer, you tend to see every problem as a nail.' Abraham Maslow DNA barcodes for species-level identification may, at first glance, seem to represent an appropriate use of new technology to solve an old problem-identifying and classifying the world's biodiversity. Working toward this goal of understanding biodiversity is commendable, but we see serious flaws in the rationale, methodology, and interpretation of results involved in abandoning morphological studies in favor of a narrow and wholly molecular identification system, as suggested by Hebert et al. (2003a). Recently the concept of DNA taxonomy, of which the DNA barcode is one instance, has been hotly debated (Tautz et al.

Approximately 600-bp sequences of mitochondrial DNA (mtDNA) have been designated as "DNA barcodes" and have become one of the most contentious and animated issues in the application of genetic information to global biodiversity assessment and species identification. Advocates of DNA barcodes have received extensive attention and promotion in many popular and refereed scientific publications. However, we suggest that the utility of barcodes is suspect and vulnerable to technical challenges that are particularly pertinent to mtDNA. We review the natural history of mtDNA and discuss problems for barcoding which are particularly associated with mtDNA and inheritance, including reduced effective population size, maternal inheritance, recombination, inconsistent mutation rate, heteroplasmy, and compounding evolutionary processes. The aforementioned could significantly limit the application and utility of mtDNA barcoding efforts. Furthermore, global use of barcodes will require application and acceptance of a barcode-based species concept that has not been evaluated in the context of the extensive literature concerning species designation. Implementation of mtDNA barcodes in spite of technical and practical shortcomings we discuss may degrade the longstanding synthesis of genetic and organism-based research and will not advance studies ranging from genomic evolution to biodiversity assessment.

No abstract