Re: Elaphe (ok Pantherophis) allegeheniensis Q

[ Follow Ups ] [ Post Followup ] [ The Taxonomy Forum ]

Posted by WW on January 31, 2003 at 04:19:55:

In Reply to: Elaphe (ok Pantherophis) allegeheniensis Q posted by troy h on January 30, 2003 at 16:06:26:

:My question is, has anyone read the paper that made these taxonomic recommendations? CNAH cites it as follows:

:"Burbrink (2001 Herpetological Monographs 15: 1-53), using external morphology, demonstrated that some eastern U.S. populations of this snake, including E. quadrivittata and E. rossalleni, actually consist of a distinct species, Elaphe alleghaniensis (Holbrook, 1836). The standard common name for Elaphe alleghaniensis is Eastern Rat Snake."

The Herp. Monogr. paper discusses various types of multivariate analyses carried out on an extensive morphological dataset. Basically, Burbrink shows that the pattern of geographic variation in the Pantherophis obsoletus complex corresponds better to the mtDNA phylogeny than to the conventional subspecies, and claims that it supports the idea of four independent evolutionary species.

I agree with the first part, namely that the analyses show that the morphol. pattern does not correspond to the conventional ssp., and corresponds better to the 4 mtDNA groups. To me, the conventional subspecies, as previously defined, are dead.

However, I do not agree that the morphological data show thatthese are separately evolving lineages. Many of the discriminant function analyses used simply pool all the samples of each individual "species" into one pooled sample (OTU - Operational Taxonomic Unit), and show the four taxa to be clearly separated. Unfortunately, this is a very common error in the use of DFAs (Reichling's elevation of the Louisiana pine snake to species status was based on the same misuse of DFA) - seeing separation between OTUs is unsurprising, as the method assumes homogeneity of OTUs and maximises separation between them. In other words, even if you have a perfectly smooth cline, if you section it into a few OTUs and run a DFA, you will get discrete taxa. Where Burbrink used methods that did not assume the homogeneity of the OTUs used (i.e., in the PCAs and in the DFAs using individual populations from the eastern anc central clades), the pattern is much less clear-cut, and there is considerable overlap among all the clades, taxa except for bairdi. In particular, a "zone of taxonomic uncertainty" (i.e., where the DFA can't tell them apart) seems to occupy ~ 50% of the range of "P. alleghaniensis". Although Burbrink acknowledges this, he still prefers to consider them as separate evolutionary lineages, based primarily on his mtDNA results.

This boils down to the usual old story that mtDNA can only tell a part of the story, namely that mtDNA, due to its clonal, matrilinear mode of inheritance, can only ever give you part of the story, namely the historical perspective. It cannot give yu a full picture of present-day patterns of genetic interchange. Brubrink suggests that the "zone of uncertainty" may be due to convergence on a "northern phenotype" due to natural selection, but also acknowledges of "leaks of gene flow across the Appalachain mountains". Those with more knowledge than myself on the altitudinal distribution of these critters in the Appalachians will be better placed to comment on that statement. He also acknowledges that nuclear markers would be required to resolve the question in a more definitive manner.

Bottom line is, I don't think the evidence presented justifies considering P. spiloides and P. alleghaniensis as different species - it's a possibility, but the present evidence does not justify it.

Moreover, Burbrink never presented an clear evidence of the relationship between his P. spiloides and P. obsoletus - are these as indistinguishable as P. alleghaniensis and P. spiloides? No evidence is given. Again, serious doubts remain in my mind.

:finally, CNAH lists the following as a "panel of systematists" that evaluated the validity of this name change. any names on the list suggest a "stacked deck"?

:"Collins & Taggart (2002 Standard Common and Current Scientific Names for North American Amphibians, Turtles, Reptiles, and Crocodilians. Fifth Edition) submitted for consideration the proposals by Burbrink, Lawson, & Slowinski (2000 op. cit.) and Burbrink (2001 op. cit.) to a snake systematist group composed of Frank T. Burbrink, Jeff Camper, Harry W. Greene, L. Lee Grismer, Robin Lawson, James R. McCranie, Andrew H. Price, Javier Rodriguez-Robles, and Samuel S. Sweet, and they agreed."

At the end of the day, taxonomy and species limits aren't decided by committees but by evidence. Globally, nobody can force you to use a taxonomic scheme you do not agree with (although some journal editors undoubtedly will), and any clasification scheme will stand or fall on the basis of future evidence. Anyone is free to start collecting that evidence (and a more detailed morphological analysis of specimens across the zone where these "species" meet is not exactly an unfeasible undertaking).

Cheers,
Wolfgang

Follow Ups: