The discussion over the past day or so on ID-frontiers has centered on "what is a subspecies?" This is an interesting topic because it really sits at the edge of the divide in the world view on how we define biological units. Birders who worry about their lifelists fall rather cleanly into two camps on the issue, those who hate all the lumping and splitting preferring listing stability, and those who look forward to the next AOU checklist, because it could mean a "free" new species.
But I find the whole discussion more interesting at the ornithological level, because it sorts out the "classical" biologists from the "quantum" biologists. Students of 20th century physicists will recognize the distinctions. Classical models tend to use standard mathematical tools and produce discrete, well defined results. Quantum models depend on randomness and produce statistical likelihoods. Einstein's theory of relativity is a classical model. The standard model for subatomic particles is a quantum mechanical model.
In biological taxonomy the analogs are the biological species concept (the classical model) and phylogenetic species concept (the quantum mechanical model). First some definitions:
Biological Species Concept (BSC) - defines species based on reproductive isolation. Species are more or less reproductively discrete units. Reproductive isolation may be mechanical, geographic or temporal. In general, any two individuals that are capable of reproducing under natural conditions to produce viable offspring are considered species even if they come from different looking populations.So, the definition of a species is fuzzy at the start. The biological species concept is pretty much useless for species capable of asexual reproductions and is inconsistently applied in sexual species (Fox Sparrow types have a zone of hybridization and can't be split, but Glaucous-winged/ Western Gulls, which also have a significant hybrid zone, can). The phylogenetic species concept says diagnosable characters make the species, but without clearly defined boundaries on what counts and how many characters are needed, could conceivably lead to every local population being arguably diagnosable as a species (my White-crowned Sparrows sing differently than yours).Phylogenetic Species Concept (PSC) - defines species based on diagnosable forms. Species are groups of individuals with a unique evolutionary history that can be represented by morphological, ecological, behavioral and genetic characters. Present day reproductive isolation is not a requirement, but historical isolation of some sort is generally presumed.
This makes subspecies definitions even more nebulous. From a BSC viewpoint, subspecies definitions are all about diagnosable characters to define regional populations. The PSC concept, because it favors calling diagnosable units species resists the term subspecies except as an investigative waystation on the road to research into possible full species designation and looks for other terms to describe intermediate forms (intergrades, clines, ring species, etc). As a consequence, most discussions about subspecies bring us back to the original divide. Are you BSC or PSC?
Now I have to admit I'm a PSC kind of guy. Understanding the evolutionary history of a species (and the ecological framework that drives that history) is far more useful than knowing what name to put on the toetag. One can make the argument that subspecific designations within a BSC model can get you to the same evolutionary history, but that would only be genuinely true if one adopted a consistent application of PSC-like methodologies and rules for identifying subspecies. There really has been little or no effort to connect the dots between diagnosable subspecies within a BSC species framework, most are still just descriptive entities without genetic connection. And as often as not, when the dots are connected the species gets split even from a BSC viewpoint....
My favorite case in point- During the great AOU lump of 1973 Bullock's Oriole and Baltimore Oriole were combined into a single species called Northern Oriole. They are clearly diagnosable forms, but interbreed in the midwest. By BSC standards they should be a single species, but when the evolutionary history of these species was charted using genetic markers, it turned out that they were not closest relatives. There were other oriole species more closely related to Bullock's Oriole. Evolutionary history did not support the BSC view of reproductive isolation as the determining factor for species designation and Bullock's and Baltimore Oriole have been split again.
Ultimately, it's all about people. Any system says more about the human need to tidy up the universe and try to make some kind of sense of things than it does about the real nature of things. The big Canada Geese were just as big and the small Canada Geese just as small before the AOU split them. The only thing that's really changed is what we choose to call them and how we model their geneology.
As seems the case with so many human issues, Shakespeare had prescient insight into things taxonomic when he wrote, "a rose by any other name would smell as sweet."
Posted by mbalame at June 10, 2005 7:56 PM