Chad Arment posted this on the `Strange Ark` email group:
A couple of recent papers have relevance to the debate over the use of DNA to describe species:
Researchers register new species using DNA-based description
http://www.eurekalert.org/pub_releases/2011-01/uog-rrn012511.php
Myth of the molecule: DNA barcodes for species cannot replace morphology for identification and classification
http://onlinelibrary.wiley.com/doi/10.1111/j.1096-0031.2003.00008.x/full
I am a strong supporter of the idea that there needs to be a single well-defined degree of genetic difference that can determine just what the species are. It needs to be much better defined than the current models are, more univerally accepted and more universally applied. I have suggested a standard unit of one million year's worth of genetic divergence equals one species, but that is also arbitrary. It does have the advantage of being well-defined, easy to remember and probably acceptable to the majority of experts. I realise also that DNA is not always measured at the same rate of change over time. But what we need is the unit of measure established and recognised, whatever that unit should turn out to be.
ReplyDeleteI am always a little skeptical of MtDNA studies alone.
ReplyDeleteFor example, using MtDNA studies, researchers over-estimated when the two species of African elephant split:
http://www.nytimes.com/2010/12/28/science/28obelephant.html
But morphology can be just a screwy with results. Take those supposed golden jackals in Ethiopia. Now, the Egyptian variant has sometimes been classified as a wolf, but it was moved into the golden jackal species about ten years ago, using really limited samples of MtDNA and comparing them to just a few samples of golden jackal and wolf MtDNA. Now, we know that the Egyptian canid that has some wolf and jackal features is actually an ancient line of wolf. Not only is the Egyptian wolf-jackal a wolf, so are some supposed golden jackals in Ethiopia. Wolves have very, very, very variable phenotypes (even within the same population or even family group) and morphological analysis can lead to real errors.
Another error that was recently made was the morphological analysis of the American lion. Making skull measurements, Danish researchers found that the American lion was a jaguar. However, when the DNA was analyzed, it was found to be an ancient line of lion, closely related to the cave lion of Europe and northern Asia.
Every one of these techniques has pitfalls. I am much more inclined to trust nuclear DNA studies than MtDNA studies, and maybe I'm showing my age, but I'm more than a little skeptical of morphology. Morphology gets things wrong a lot. Look no further than the African butterfly fish to see what I mean:
http://www.nhm.ac.uk/about-us/news/2010/october/surprising-genetics-of-the-african-butterfly-fish83040.html