Re: New Taxonomy of the Elaphe obsoleta complex

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

Posted by Frank Burbrink on October 19, 2001 at 12:06:04:

In Reply to: Re: New Taxonomy of the Elaphe obsoleta complex posted by WW on October 18, 2001 at 06:09:24:

Wolfgang,

I appreciate your comments and the debate about modern taxonomy and systematics is very worthwhile. I still don't agree with all of your conclusions.

First off, I don't buy the BSC either. All species concepts are trying to get at the same thing...where are the evolutionary lineage break (species). However, all concepts use a different criteria. I happen to use the existence of evolutionary lineages as my criteria. You have to agree that at one time these three groups (the eastern, central and western clades) were separated. They could not have evolved the deep mtDNA splits that they did without lengthy separation (among other explanations, i.e., assortative mating). Currently, the barriers that keep them divided may not be working as well. This may not be fact either...you could still have had limited numbers crossing these barriers in the past. Anyhow, the barriers may be leaky. This does not say that the animals are hybridizing. No one has data that suggests this. So, if you notice in my molecular paper, there were no specimens found to have crossed the Apalachicola. That barrier may be leaky, but it hasn't been established. I believe you misread this. Also, the specimen from Craighead Co., Ark in the molecular central clade was found on the west side of the Miss. River. I recently discovered that the River historically used to flow to the west of this sample...thus putting it correctly in the central clade. Anyhow, this arm waving is not making my point. Essentially, you are trying to say that a certain amount of gene flow will discredit species recognition. We don't have data to support gene flow. Also, if there is gene flow at these barriers, why would it be wrong to call these different species with a hybrid zone? It is very arbitrary to say that enough gene flow has caused three historical lineages to cease to exist. Obviously gene flow isn't so rampant across these barriers that the mtDNA lineages are no longer geographically recognizable. If gene flow was so extensive, then you wouldn't have these lineages. You do have three lineages. Morevoer, the E. obsoleta complex is not an exclusive species group. The western lineage is actually more closely related to the established E. bairdi that it is to the eastern and central taxa.

The morphological data was never a test of species. It is presented to see which system it supports better (subspecies or clades). I agree that DFA does maximize separation. However, it cannot place individuals into groups if there isn't enough variation within/ between all groups.
The only risk is that the data set was over-paramterized and that any combination could exist. This did not happen, because the number of specimens used was much higher than the number of characters. There is good statistical separation between the central and western clades. This coincides with the depth of mtDNA splits. The northern morphs of the eastern and central clades are not as easily divided using morphology. This does not mean that they aren't species. You may not have had complete morphological separation between these newly formed species. However, they do show trends that follow these molecular lineages. Obviously when lineages form they may show no morphological difference. See some of Wake's Batrachoseps or Highton's Plethodon. So, just because you have examined PCA and DFA plots for many years (and I do respect your work with Naja etc.) doesn't mean that you can positively identify species with multivariate plots?particularly cryptic species. Many species within the above salamander genera could not be separated using multivariate analyses of morphological data.

I was simply examining trends. If there are nuclear data that shows these E. obsoleta lineages represent a thing of the past, then I would say sink the species. However, based on molecular data and morpholoigcla trends, I would say these are species. I am jut using the available data.which is a lot more than what we had two years ago. I don't mind sinking my own taxonomies if I can show convincing evidence that says these lineages are no longer valid. This is currently being examined by myself. Often people who keep these snakes as pets have strong priors on their taxonomy (possibly because they are used to these old names or they don't want to learn a new system). Unfortunately, they have very little to suggest a real system. I am trying to keep these taxomies in line with evolutionary history. It is simple biology that these lineages do not have great diagnosible morphological characters that can be used to distinguish these species. I am sorry that these lineages have not yet produced easily diagnosable morphological characters(i.e., black rat snake, yellow rat snake, gray rat snake, candy corn rat snake etc.).

I have already read Thorpe et al. paper. It is suspect that he excluded the samples the western mtDNA clades in the morphological analysis.

Thanks for the dialogue.

Frank Burbrink

Follow Ups:

Can you put this into simpler terms? ... more ... Patricia Sherman 11/28/01 (0)

Post a Followup

Name:
E-Mail:

Subject:

Comments:
: Wolfgang, : I appreciate your comments and the debate about modern taxonomy and systematics is very worthwhile. I still don't agree with all of your conclusions. : First off, I don't buy the BSC either. All species concepts are trying to get at the same thing...where are the evolutionary lineage break (species). However, all concepts use a different criteria. I happen to use the existence of evolutionary lineages as my criteria. You have to agree that at one time these three groups (the eastern, central and western clades) were separated. They could not have evolved the deep mtDNA splits that they did without lengthy separation (among other explanations, i.e., assortative mating). Currently, the barriers that keep them divided may not be working as well. This may not be fact either...you could still have had limited numbers crossing these barriers in the past. Anyhow, the barriers may be leaky. This does not say that the animals are hybridizing. No one has data that suggests this. So, if you notice in my molecular paper, there were no specimens found to have crossed the Apalachicola. That barrier may be leaky, but it hasn't been established. I believe you misread this. Also, the specimen from Craighead Co., Ark in the molecular central clade was found on the west side of the Miss. River. I recently discovered that the River historically used to flow to the west of this sample...thus putting it correctly in the central clade. Anyhow, this arm waving is not making my point. Essentially, you are trying to say that a certain amount of gene flow will discredit species recognition. We don't have data to support gene flow. Also, if there is gene flow at these barriers, why would it be wrong to call these different species with a hybrid zone? It is very arbitrary to say that enough gene flow has caused three historical lineages to cease to exist. Obviously gene flow isn't so rampant across these barriers that the mtDNA lineages are no longer geographically recognizable. If gene flow was so extensive, then you wouldn't have these lineages. You do have three lineages. Morevoer, the E. obsoleta complex is not an exclusive species group. The western lineage is actually more closely related to the established E. bairdi that it is to the eastern and central taxa. : The morphological data was never a test of species. It is presented to see which system it supports better (subspecies or clades). I agree that DFA does maximize separation. However, it cannot place individuals into groups if there isn't enough variation within/ between all groups. : The only risk is that the data set was over-paramterized and that any combination could exist. This did not happen, because the number of specimens used was much higher than the number of characters. There is good statistical separation between the central and western clades. This coincides with the depth of mtDNA splits. The northern morphs of the eastern and central clades are not as easily divided using morphology. This does not mean that they aren't species. You may not have had complete morphological separation between these newly formed species. However, they do show trends that follow these molecular lineages. Obviously when lineages form they may show no morphological difference. See some of Wake's Batrachoseps or Highton's Plethodon. So, just because you have examined PCA and DFA plots for many years (and I do respect your work with Naja etc.) doesn't mean that you can positively identify species with multivariate plots?particularly cryptic species. Many species within the above salamander genera could not be separated using multivariate analyses of morphological data. : I was simply examining trends. If there are nuclear data that shows these E. obsoleta lineages represent a thing of the past, then I would say sink the species. However, based on molecular data and morpholoigcla trends, I would say these are species. I am jut using the available data.which is a lot more than what we had two years ago. I don't mind sinking my own taxonomies if I can show convincing evidence that says these lineages are no longer valid. This is currently being examined by myself. Often people who keep these snakes as pets have strong priors on their taxonomy (possibly because they are used to these old names or they don't want to learn a new system). Unfortunately, they have very little to suggest a real system. I am trying to keep these taxomies in line with evolutionary history. It is simple biology that these lineages do not have great diagnosible morphological characters that can be used to distinguish these species. I am sorry that these lineages have not yet produced easily diagnosable morphological characters(i.e., black rat snake, yellow rat snake, gray rat snake, candy corn rat snake etc.). : I have already read Thorpe et al. paper. It is suspect that he excluded the samples the western mtDNA clades in the morphological analysis. : Thanks for the dialogue. : Frank Burbrink

Optional Link URL:
Link Title:
Optional Image URL:

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