DNA barcoding for species identification and discovery in diatoms

Diatoms are the largest group of microalgae, play an enormous role in the biosphere, and have major significance as bioindicators. Traditional identification requires inter alia long training, considerable microscopical skill, and use of a vast and scattered literature. During the life cycle, diatom cells change in size and pattern, often also shape, but the full cycle is known in <1% of described species. Recent evidence shows that there are many pseudocryptic and cryptic species of diatoms, requiring molecular methods for discovery and recognition. These and other factors argue that DNA barcoding would be highly beneficial. It could be 'strong', resolving nearly all species, or 'weak', resolving mostly species already recognized from light microscopy. Attempts have already been made to identify suitable genes and we evaluate these on the basis of universality and practicality, and ability to discriminate between species in the very few 'model' systems offering likely examples of sister-species-pairs. No candidate marker is ideal but LSU rDNA and rbcL may be acceptable, though their discriminatory power is lower than that of some other markers. We discuss the next steps in developing a full barcode system.