Reptile & Amphibian Forums

A week ago Sunday, I spoke to the grad student that preformed a second and more comprehensive mtDNA study on the Rubber Boa. He indicated it would OK to reveal the major findings. I am throwing them out so perhaps others can hash over just what may have transpired with respect to the biogeographical history of the species and the taxonomic implications.

Along with incorporating the 38 samples used in the first mtDNA study published by Rodriguez-Robles, Stewart, and Papenfuss, Rick (who is Glenn Stewart's last grad student as Glenn is now retired) ran tests on 74 additional samples with much greater geographical representation that in the first study..

If I have my information correct, Rick ran the same type of mtDNA tests used by Javier Rodriguez-Robles. And as expected, he ended up finding two major clades, the Southern and Northern clades with the latter being composed of two major groupings, the Northwestern and Sierra Nevada subclades. However, there were three major new developments that emerged.

1) Beside occurring in the San Bernardino and San Jacinto Mts., a boa population of the Southern clade also occurs on the Southern Kern Plateau in southern Tulare County.
2) There is likely to be no break between the Sierra Nevada and Northwestern subclades as two specimens well south of the Mt. Lassen region from Butte County aligned with the Northwestern subclade.
3) Thirty-two samples with the same haplotype were dispersed from northern California to Oregon, Washington, B.C., Montana, Idaho, Utah, and I believe Nevada.

The ramifications of point #1 above are extensive and intriguing.

Richard F. Hoyer

I think result #3 is also very interesting. What about the populations between the San Bernardinos and the southern Kern Plateau? When are you going to include some nuclear markers?

DHL,
Another genetic study of the Rubber Boa has been proposed in which both molecular and nuclear tests would be incorporated. Whether or not that study gets beyond the talking stage is not known but at this point in time, Dr. Chris Feldman would likely be the lead investigator.

Despite new findings that emerged from the recent mtDNA study completed by grad student Richard Toshima, there remains a fair number of gaps in information relating to various populations of the species.

For instance you ask about the boa populations between the populations that occur in San Bernardino Mts. and the Southern Kern Plateau. At this point in time, there is only one such population that has gone untested and those are the boas that occur in the Scodie Mts. immediately south of the extreme southeastern part of the Kern Plateau. Although surrounded by Mojave Desert type habitat, the Scodie Mts. are connected to the Kern Plateau by Walker Pass at about 5300 ft. A reasoned guess is that those boas would also align with the Southern Clade.

I had once proposed to have that population tested it was bypassed. Then with the discovery that the Kern Plateau population in south of Tulare County aligned with the Southern Clade, Rick no longer had a lab in which he could test tissue from the Scodie specimen. All other known populations that occur between the San Bernardino Mts. and Kern Plateau
have been tested and align with the Northern Clade (Sierra Nevada subclade)

Those populations include boas that occur in the Piute Mts. about 12 miles west of the Scodie Mts. and again separated by the Mojave Desert. Only one sample from the Piute Mts. was tested by Rodriguez-Robles and repeated in the current study. I have another sample of tissue from a boa from the Piutes that wasn't tested. And just west of the Piute Mts. is Breckenridge Mt. and south of the Piutes are the Tehachapi Mts. The boas from Mt. Pinos boas to the southwest of the Tehachapi Mts. on the west side of the I-5 corridor and Frazier Park, Calif. were also tested. I have additional samples from all of those regions.

As for the 38 samples that had identical haplotypes, that certainly is of interest which from my limited understanding of mtDNA research would seem to suggesting a rapid dispersal by a successful general phenotype.

Richard F. Hoyer

Yes, so which clade do the Mt. Pinos and Frazier Peak group with? That's what I was asking about.

And yes, it does appear a recent and rapid range expansion through out much of the distribution of C. bottae. That's cool!

DHL,
The Mt. Pinos population (only one sample tested) aligned with the Northern Clade (Sierra Nevada subclade). No samples have been tested from nearby boa populations also known to occur west of I-5 on Mt. Abel, Alamo Mt., Sawmill Mt., and Frazier Mt.

The other intriguing aspect involved in the taxonomy of Charina bottae is the discovery of a reasonably distinctive dwarf form of the species that occurs throughout S. Calif. This is the project I have been working on for the past 10 years or thereabouts. Populations of the dwarf morph that belong to the Southern Clade occur in the San Bernardino and assumed
to occur in the San Jacinto Mts. although that population has never been studied. I now have reasonable evidence that the boa population on the southern Kern Plateau also are of the dwarf form which has now been identified as belonging to the Southern Clade.

The Northern Clade (Sierra Nevada subclade) contains boa populations that are of the dwarf morph and large morph. It is now my view that somewhere in Tulare and Inyo Counties, there likely are large morph boas that belong to the Southern Clade.

Richard F. Hoyer

So what do you think of gene flow between umbricata and bottae? Pretty minimal or none? Do you consider C. umbricata a valid taxon?

DHL,
Because the officially recognized SRB subspecies occurs as isolated populations only in the San Bernardino and San Jacinto Mts., it probably has been considerable time since any gene flow occurred between those populations and C. b. bottae.

However, if one includes the Southern Clade boa population that occurs in the southern Kern Plateau, then gene flow between the Southern and Northern Clades has likely been ongoing for eons somewhere in Tulare and Inyo counties .

Before I undertook the SRB study during 1993 - 1997, I had doubts about the SRB being recognized as a distinct subspecies. But during that study, the population in the San Bernardino Mts. did seem pretty unique at the time. The characteristic described for the subspecies by Klauber in 1943 from only two specimens, are rather consistent in the SRBs in the San Bernardino Mts. No one has undertaken a study of the SRB in the San Jacinto Mts. so nothing of much certainty can be said about that population.

Since that time, I have discovered that 1) there are other boa populations that similar to the SRBs, are also of the dwarf morph and 2) considerable overlap in defining traits occur between the SRB and other Northern Clade boa populations in S. Calif. So the subspecies designation comes into question again.

The mtDNA data however, is a powerful argument in favor of subspecies status. I know little about taxonomy per se and thus leave it up to those more versed to sort out the pros and cons. I have never considered umbratica as a separate species. Well before Javier's publication, I was aware that to much overlap in defining traits occurred between the SRB and boas from other populations including one population in SW Oregon. Then with my examination of other boa populations in S. Calif., and in particular the Mt. Pinos boa population, with the amount of overlap of key traits, those boas make a mockery of the subspecies designations----but only from a morphological perspective.

With the discovery of the Southern Clade occurring in the Kern Plateau, not only does that revelation likely sink the separate species scenario, it probably puts into question the subspecies designation as well. Perhaps it may come down to just how much divergence occurs between the San Bernardino and San Jacinto Mts. populations and the Southern Clade boas that occur on the Kern Plateau.

Richard F. Hoyer

>>DHL,
>>Because the officially recognized SRB subspecies occurs as isolated populations only in the San Bernardino and San Jacinto Mts., it probably has been considerable time since any gene flow occurred between those populations and C. b. bottae.>>

Yes, that is what the mtDNA data shows.

>>However, if one includes the Southern Clade boa population that occurs in the southern Kern Plateau, then gene flow between the Southern and Northern Clades has likely been ongoing for eons somewhere in Tulare and Inyo counties . >>

Not quite. The mtDNA tree shows an unresolved polytomy between most populations of the Sierra Nevada subclade and shallow mtDNA divergences among the mtDNA haplotypes of the Sierra Nevada Mountain populations. Together these two aspects of the mtDNA data suggests that the small morph rubber boas of the Kern County area, after becoming isolated from the southern California populations, gave rise to the Northwestern subclade, but subsequently the population contracted. The once continuous distribution of the Northwestern and Kern County boas was severed. In fact, the mtDNA haplotype of the Kern County boas suggest that they are derived from the Tulare County (#26) boa at about the same time that this same population gave rise to the Sierra Nevada Mountain populations, and more recently than the Northwestern subclade/Kern Plateau boa split. Hence the current Tular/Kern County animals have been isolated from both the Northwestern and Southern California boas for long periods of time without any gene flow among these three populations. That is why we don't see large morph phenotype animlas in the Tulare County and Kern County populations, at least this is the testable prediction of my theory of the evolutionary history of this species.

>>Before I undertook the SRB study during 1993 - 1997, I had doubts about the SRB being recognized as a distinct subspecies. But during that study, the population in the San Bernardino Mts. did seem pretty unique at the time. The characteristic described for the subspecies by Klauber in 1943 from only two specimens, are rather consistent in the SRBs in the San Bernardino Mts. No one has undertaken a study of the SRB in the San Jacinto Mts. so nothing of much certainty can be said about that population.
>>
>>Since that time, I have discovered that 1) there are other boa populations that similar to the SRBs, are also of the dwarf morph and 2) considerable overlap in defining traits occur between the SRB and other Northern Clade boa populations in S. Calif. So the subspecies designation comes into question again.
>>

As I suggested in my earlier posts, the SRB should be grouped in the same subspecies as the dwarf morph boas of Kern and Tulare County. These populations almost certainly represent the ancestral morphotype from which other populations have evolved.

>>The mtDNA data however, is a powerful argument in favor of subspecies status. I know little about taxonomy per se and thus leave it up to those more versed to sort out the pros and cons. >>

The subspecies category is an informal one, and there is no requirement that anyone recognize it. However, many biologists, as Ernst Mayr and others have pointed out, find the subspecies category useful and they therefore use it for pragmatic reasons. However, some taxonomists do object to the use of the subspecies category for ideological reasons, so they tend to either elevate subspecies to full species status (often without good reason) or else find an excuse not to recognize any subspecies at all. Unfortunately, Rodriguez-Robles et al. seem to fit the ideological category. That is why they found a reason (which you believe to be invalid) to elevate umbratica to full species. That is also why Rodriguez-Robles argued that no subspecies should be recognized in Lampropeltis zonata.

>>I have never considered umbratica as a separate species.>>

Good for you. And I agree with you. The supposed morphological differences between umbratica and the other populatioins have been falsified by your data. And even if they are valid, these differences can best be used to delineate subspecies, not species. Species ought to be delineated on the basis of reproductive isolation, not minor morphological differences.

>>Well before Javier's publication, I was aware that to much overlap in defining traits occurred between the SRB and boas from other populations including one population in SW Oregon. Then with my examination of other boa populations in S. Calif., and in particular the Mt. Pinos boa population, with the amount of overlap of key traits, those boas make a mockery of the subspecies designations----but only from a morphological perspective.
>>
>>With the discovery of the Southern Clade occurring in the Kern Plateau, not only does that revelation likely sink the separate species scenario, it probably puts into question the subspecies designation as well.>>

That would be true if we define umbratica as only those animals from San Bernardino and San Jacinto Mountains. A better definition of umbratica, in my view, is to include the dwarf morph Kern County and Tular County boas in umbratica. Since the Northwestern and Sierra Nevada Mountain rubber boas are both large morph, umbratica becomes a tenable designation for the dwarf morph populations. C. b. umbratica would be paraphyletic, but that is not inconsistent with any aspect of traditional taxonomy, although it may upset some cladists.

>>Perhaps it may come down to just how much divergence occurs between the San Bernardino and San Jacinto Mts. populations and the Southern Clade boas that occur on the Kern Plateau.
>>
>>Richard F. Hoyer

The "divergence" between these populations, in terms of mtDNA, is large, larger than that between the Kern Plateau populations and either the Sierra Nevada Mountain or Northwestern subclades. Morphologically, the divergence appears small, as they are both small morph snakes. If other characteristics that are traditionally used in snake taxonomy do not show a marked dichotomy, then these two populations should definitely be combined into the subspecies umbratica.

>>DHL,
>>The Mt. Pinos population (only one sample tested) aligned with the Northern Clade (Sierra Nevada subclade). No samples have been tested from nearby boa populations also known to occur west of I-5 on Mt. Abel, Alamo Mt., Sawmill Mt., and Frazier Mt.
>>

As my other post suggests, these populations are likely to be dwarf morph and descended from Tulare County dwarf morph population #26 in Rodriguez-Robles et al.'s sample, or alternatively, from the other Tulare County population with the umbratica haplotype.

>>The other intriguing aspect involved in the taxonomy of Charina bottae is the discovery of a reasonably distinctive dwarf form of the species that occurs throughout S. Calif. This is the project I have been working on for the past 10 years or thereabouts. Populations of the dwarf morph that belong to the Southern Clade occur in the San Bernardino and assumed to occur in the San Jacinto Mts. although that population has never been studied. >>

Yes, Richard, this is your invaluable contribution to the taxonomy of this species and genus. Without this information, Charina bottae taxonomy would be a big mess.

>>I now have reasonable evidence that the boa population on the southern Kern Plateau also are of the dwarf form which has now been identified as belonging to the Southern Clade. >>

Good to hear that.

>>The Northern Clade (Sierra Nevada subclade) contains boa populations that are of the dwarf morph and large morph.>>

As currently defined, yes. But I believe it would be a good idea to delineate species and subspecies on the basis of morphology rather than mtDNA haplotype. Hence it would be a good idea to align the dwarf morph boas of the Kern County/Tulare County area with umbratica, and classify the Sierra Nevada large morph populations as a separate subspecies, in effect, splitting the Sierra Nevada subclade into two.

>>It is now my view that somewhere in Tulare and Inyo Counties, there likely are large morph boas that belong to the Southern Clade.
>>
>>Richard F. Hoyer

That certainly is possible. Convergent evolution does occur. The Sierra Nevada large morph and the Northwestern subclade probably evolved their large morph phenotypes convergently and independently of each other. And the Rosy Boa is almost certainly a large morph descendant of a small morph rubber boa population that has taken refuge in northern Baja California during a period of severe climate that made most of southern California inhospitable to the small morph rubber boa. Not coincidentally, the unicolored form of the Rosy Boa occurs rather close geographically to the likely northern Baja California origin of the Rosy Boa, and the unicolored form of the Rosy Boa is also morphologically much closer to the Rubber Boa than the lined forms. Hence large morph snakes may have evolved at least 2-3 times from small morph ancestors in Charina bottae.

It appears to me that the boas occurring in Kern and Tulare Counties (within the area occupied by L. z. multifasciata) are of the dwarf morph variety, whereas the boas occurring within the range of L. z. multicincta in the Sierra Nevada to the north are large morph. Therefore we are seeing a L. z. multifasciata/small morph C. bottae and a L. z. multicincta/large morph C. bottae association. L. z. multicincta and L. z. zonata form a broad hybrid zone in the northern Sierra Nevada and the mountains to the north and west, but there appears to be little if any overlap between Northwestern subclade C. bottae and large morph Sierra Nevada C. bottae in the same area. It is intriguing how closely the distributions of these two species mirror each other in some instances, and how different they are in other instances.

>>All other known populations that occur between the San Bernardino Mts. and Kern Plateau have been tested and align with the Northern Clade (Sierra Nevada subclade)>>

That actually depends on how the Sierra Nevada subclade is defined. All those populations in the Sierra Nevada, north of Tulare County, are of the large morph variety, according to your data. The Kern County animals are all small morph. If you look at the tree in Fig. 3A of Rodriguez-Robles et al., you can see that sample #26, from Tulare County, is actually basal to (older than) a large group of populations which occur both north and south of Tulare County. Together with the new finding of umbratica in Tulare, it would appear to me then that the current Kern County animals are descended from a Tulare County population (#26). That suggests to me that the distributional gap between umbratica and the Kern County small morph animals was even larger in the past than it is today. Therefore the Scodie Mountain population may well have migrated to its current position from Tulare County, rather than being a bridge population between San Bernardino and Kern County populations.

>>Those populations include boas that occur in the Piute Mts. about 12 miles west of the Scodie Mts. and again separated by the Mojave Desert. Only one sample from the Piute Mts. was tested by Rodriguez-Robles and repeated in the current study. I have another sample of tissue from a boa from the Piutes that wasn't tested. And just west of the Piute Mts. is Breckenridge Mt. and south of the Piutes are the Tehachapi Mts. The boas from Mt. Pinos boas to the southwest of the Tehachapi Mts. on the west side of the I-5 corridor and Frazier Park, Calif. were also tested. I have additional samples from all of those regions.>>

Here is my prediction: all of these boas found between Tulare and San Bernardino Mountains are closer to the Sierra Nevada subclade than they are to umbratica. That means they migrated south from Tulare County rather than north from San Bernardino County.

>>As for the 38 samples that had identical haplotypes, that certainly is of interest which from my limited understanding of mtDNA research would seem to suggesting a rapid dispersal by a successful general phenotype.>>
>>Richard F. Hoyer

Agreed. These populations are all large morph, suggesting that the evolution of the larger morph has enabled it to radiate into new habitats. The small morph southern rubber boas are probably restricted in distribution to a few mountain refugia by its closest relative, the Rosy boa.

In case you haven't read some of my posts concerning rosy boa mtDNA, this newly published data on the Rosy boa suggests that it originated in northern Baja California. That means the rosy boa probably descended from an isolated population of the rubber boa in the vicinity of northern Baja, while all of the intermediate rubber boa populations in the lower elevations of southern California must have been extirpated, ostensibly by a drying of the climate, which also wiped clean of these mountains of the basal populations of L. zonata. L. z. pulchra and L. z. parvirubra are both recently derived from L. z. agalma according to mtDNA data, just as all other populations of Lichanura are derived from a northern Baja California population. There is amazing agreement between the mtDNA data of L. zonata and Charina/Lichanura, especially in what these data sets are informing us of past evolutionary events and likely geologic history. Fascinating stuff.

It looks like the Rosy Boa and the Northwestern subclade of the Rubber Boa may have converged on the same solution: a larger body size to cope with hotter and drier climates. The same solution has enabled both of these descendant lineages of the small morph rubber boa to exploit new habitats and expand their ranges.

CK,
Unfortunately, I do not know enough about mtDNA research in order to interpret those trees. But long ago I had noted that sample #26 from Tulare County seemed to stick out by itself but just didn't know how to interpret that particular aspect.

With the new results showing that specimens from southeastern Tulare county align with the Southern Clade and with #26 also from Tulare county but belonging to the Northern Clade (Sierra Nevada subclade), knowing the locality of #26 in Tulare Co. became important. Javier did not specify the locality information for that specimen, just the county. The specimen came from the Cal Poly, Pomona collection (#2223) but other than county, had no specific locality data on the document upon which other collection information was present.

Rick was producing a map of specimens tested in his study and he simply used the city of Tulare as the locality for specimen #26. Of course, that was not very satisfactory. I had examined all Cal Poly specimens and retained copies of the information sheets for each specimen. CP #2223 had the name of the collector whose last name was unusual and began with a 'Z'.. After contacting Glenn Stewart and getting his input, I was able to Google the name and found the gentleman, a dentist in Arizona. He recalled exactly where that specimen was found near the community of Camp Nelson in central-western Tulare County about 20 miles north of Johnsondale.

I have examined and taken tissue from a number of specimens south of Alta Sierra (Greenhorn Mts.) in northern Kern County and from 17 specimens north of Alta Sierra in southern Tulare County. I have also examined a number of voucher specimens from the Sequoia National Park region of northwestern Tulare County. The boa population south of Alta Sierra appear to be of the dwarf form and those in Sequoia National Park of the large morph. Some of the 17 boas examined from north of Alta Sierra in southwestern Tulare County appear to be intergrades as they have traits found in both the dwarf and large morph populations. As specimen #26 in Javier's study is a large morph female that probably was close to 24 inches when alive.

Rick tested several samples of the boas both south and north of Alta Sierra. All of those boas, along with Javier's #26, align with the Northern Clade (Sierra Nevada subclade).
Yet specimen #26 and the 17 specimens from southern Tulare County are only about 23 - 25 miles to the west of the Fish Creek drainage of the Kern Plateau where the dwarf morph boas of the Southern Clade occur. It can be noted that the N. Fork, Kern River gorge separates the western and eastern parts of Tulare County, the latter being the Kern Plateau where the southern clade boas occur.

You mention the following:
"Here is my prediction: all of these boas found between Tulare and San Bernardino Mountains are closer to the Sierra Nevada subclade than they are to umbratica,"

The populations Javier tested from Kern County were from the Piute Mts., Breckenridge Mt., Tehachapi Mts. and Mt. Pinos. He had all of those boas and specimen #26
belonging to the Sierra Nevada subclade.

You also mention: "That means they migrated south from Tulare County rather than north from San Bernardino County." You have previously mentioned that the dwarf form is ancestral to the large morph. However, specimens #26 is a large morph specimen.

Richard F. Hoyer

>>CK,
>>Unfortunately, I do not know enough about mtDNA research in order to interpret those trees. But long ago I had noted that sample #26 from Tulare County seemed to stick out by itself but just didn't know how to interpret that particular aspect.>>

It is not that difficult. Many trees are presented with the most basal nodes on the left. If we think of a tree as growing from the ground up, then it is perhaps not as easy to see which nodes are more basal when it is presented that way. If you want, you can print out the tree on a piece of paper, then cut it out, and flip it 90 degrees. When that is done, you can more easily see that the node, or point from which sample #26 from Tulare County branches off from the tree is lower than most of the other samples that comprise the Sierra Nevada subclade. That means sample #26 evolved earlier as a population, earlier for example, than #30, #31 and #32, since it originates at a point lower than these other samples. In mtDNA trees, that suggests #30-32 more likely than not evolved from #26 (or more correctly, the common ancestor of sample #26, 15, 17, 18, 19. Since the general direction of rubber boa migration is from south to north, and since samples #30-32 occur to the south of sample #26, this particular topology of the tree suggests that the rubber boa was restricted in distribution to the vicinity of Tulare County after it made its way out of southern California. At one time there were no boas in Kern County, south of sample #26. Later, probably when conditions are suitable for range expansion, Kern County was recolonized by rubber boas again, with seed stock coming from the area in Tulare County, close to #26. If that particular scenario did not happen, we would instead see a tree in which the Kern County boas occupy a more basal position of the tree than #26. I hope I have done an adequate job of explaining what I see from the tree.

In brief, since rubber boas migrated from south to north, one would expect Kern County boas to be older than Tulare County boas, unless there was a range contraction in the intermediate past which resulted in the Kern County area being wiped clean of boas (probably because of climatic changes). More recently, this area became suitable for boas, so it is recolonized by boas occurring to the north of Kern County.

>>With the new results showing that specimens from southeastern Tulare county align with the Southern Clade and with #26 also from Tulare county but belonging to the Northern Clade (Sierra Nevada subclade), knowing the locality of #26 in Tulare Co. became important. Javier did not specify the locality information for that specimen, just the county. The specimen came from the Cal Poly, Pomona collection (#2223) but other than county, had no specific locality data on the document upon which other collection information was present.>>

That lack of information is not as critical as one may imagine. If the county data is correct, it merely means that there is a mixture of umbratica haplotype and Sierra Nevada subclade haplotypes in Tulare County.

>>Rick was producing a map of specimens tested in his study and he simply used the city of Tulare as the locality for specimen #26. Of course, that was not very satisfactory. I had examined all Cal Poly specimens and retained copies of the information sheets for each specimen. CP #2223 had the name of the collector whose last name was unusual and began with a 'Z'.. After contacting Glenn Stewart and getting his input, I was able to Google the name and found the gentleman, a dentist in Arizona. He recalled exactly where that specimen was found near the community of Camp Nelson in central-western Tulare County about 20 miles north of Johnsondale.>>

Good for you. Nice detective work.

>>I have examined and taken tissue from a number of specimens south of Alta Sierra (Greenhorn Mts.) in northern Kern County and from 17 specimens north of Alta Sierra in southern Tulare County. I have also examined a number of voucher specimens from the Sequoia National Park region of northwestern Tulare County. The boa population south of Alta Sierra appear to be of the dwarf form and those in Sequoia National Park of the large morph. Some of the 17 boas examined from north of Alta Sierra in southwestern Tulare County appear to be intergrades as they have traits found in both the dwarf and large morph populations. As specimen #26 in Javier's study is a large morph female that probably was close to 24 inches when alive.>>

That is an unexpected result. If I were to guess, I would have guessed that #26 was a small morph snake. That means the Kern County boas are also likely to be large morph as well, since they shared a recent common ancestor with #26.

>> Rick tested several samples of the boas both south and north of Alta Sierra. All of those boas, along with Javier's #26, align with the Northern Clade (Sierra Nevada subclade).
>>Yet specimen #26 and the 17 specimens from southern Tulare County are only about 23 - 25 miles to the west of the Fish Creek drainage of the Kern Plateau where the dwarf morph boas of the Southern Clade occur. It can be noted that the N. Fork, Kern River gorge separates the western and eastern parts of Tulare County, the latter being the Kern Plateau where the southern clade boas occur.>>

>>You mention the following:
>>"Here is my prediction: all of these boas found between Tulare and San Bernardino Mountains are closer to the Sierra Nevada subclade than they are to umbratica,"
>>
>>The populations Javier tested from Kern County were from the Piute Mts., Breckenridge Mt., Tehachapi Mts. and Mt. Pinos. He had all of those boas and specimen #26 belonging to the Sierra Nevada subclade. >>

Yes, and in fact as I pointed out above, the Kern County boas appeared to be more derived than sample #26, suggesting that they may have been derived from sample #26.

>>You also mention: "That means they migrated south from Tulare County rather than north from San Bernardino County." >>

Absolutely.

>> You have previously mentioned that the dwarf form is ancestral to the large morph.>>

Correct again. umbratica is basal to all other rubber boa populations, and umbratica is a dwarf morph.

>>However, specimens #26 is a large morph specimen.
>>
>>Richard F. Hoyer

That is indeed very interesting. As you mentioned, there is a mixture of dwarf and large morph snakes in Tulare County. So, the fact that #26 is a large morph snake would suggest that the Kern County boas would also be large morph as well. Is there data on these boas? There is also the possibility that intergradation between the large and small morph snakes have resulted in the snake in sample #26 to have a small morph haplotype but a large morph phenotype. That means #26 descended from a female with a small morph ancestor that was mated to a large morph male. If the Kern County boas are small morph, that would be the most likely explanation.

An addendum.

If the Kern County boas, samples #30-32, are small morph boas, then most likely #26 is an intergrade. mtDNA is inherted matrilineally. The situation would most closely resemble our President Elect Barrack Obama. Mr. Obama inherited his mtDNA from his mother, so if we look at his mtDNA haplotype alone, he would be considered a European. However, his phenotype is a mix of European and African because his father is Kenyan. His Y chromosome, because it is inherited entirely from his father, would be similar to other Kenyan males, with no trace of a European ancestry.

CK,
There are three 'trees' in Javier's paper. If I am following your process correctly, in Fig. 3A, specimen #26 would be the ancestral type for all boas in the Sierra Nevada subclade. In Fig. 3B, specimen #26 would be the ancestral type for all Sierra Nevada and Northwestern subclade boas. In Fig. 4, it would appear that specimens #15, 17, 18, and 19 are the ancestral types for all Sierra Nevada and Northwestern subclade boas. But of course, I may not be interpreting those figures in the manner you indicate.

But if I am, then it would appear that the Calabar and/or Rosy Boa are the ancestral types to the Rubber Boa.

Just because #26 branched off earlier and thus is older than the other specimens, I do not understand just why her lineage and not some other ancestral type, now extinct or still in existence but just haven't been found and tested, may have given rise to the other lineages in the tree.

A good number of years ago, I was surprised at the lack of representation in mtDNA studies. When I questioned that point, I was told that such representation was not necessary as new samples from the same region did not materially add to results. I am now of the opinion that I was correct all along, that inadequate representation can often lead to erroneous interpretations of the results obtained from such studies.

Interpretations are made on the basis that the samples tested are seemingly the last word and thus finite. Rick's study has demonstrate that is not the case and I venture to say that the current study also lacks adequate representation in a number of regions. Seemingly not taken into account is that there may be a good number of closely related or identical haplotypes of specimens #26 that may occur north, south, east, and west of where #26 originated near Camp Nelson. Without having much larger sample size and geographical representation, to interpret that this or that population migrated in any direction is premature to my way of thinking.

Concerning the distribution of the two size morphs, my sample size from the Tehachapi Mts. and Breckenridge Mt. are large enough to assert that like the San Bernardino boa population, those two populations are also of the dwarf morph. But besides having an adequate sample size, there are other clues that indicate whether the boas from any particular region belong to the dwarf or large morph phenotype.

The maximum and mean lengths of neonates of litters produced is one clue. The lengths of mature individuals is another. The lengths at which males and females attain maturity is another. And the rate of growth over time is another. For the populations in which I lack an adequate sample, some of the above clues indicate that the population of boas in the Mt. Pinos area, in the Scodie Mts., in the Greenhorn Mts. south of Alta Sierra all in Kern County, and on the Southern Kern Plateau in Tulare County are all of the dwarf form. Then by geographical proximity, one would expect the dwarf form to occur in the Piute Mts., Alamo Mt., Frazier Mt., Sawmill Mt. and Mt. Abel where the species has been documented in Kern County and the San Jacinto Mts. of Riverside County.

Until some nuclear markers are found, I believe it is not plausible to determine with any certainty just where a mixture of dwarf and large morph genotypes exists.

Richard F. Hoyer

>>CK,
>>There are three 'trees' in Javier's paper. If I am following your process correctly, in Fig. 3A, specimen #26 would be the ancestral type for all boas in the Sierra Nevada subclade. In Fig. 3B, specimen #26 would be the ancestral type for all Sierra Nevada and Northwestern subclade boas. In Fig. 4, it would appear that specimens #15, 17, 18, and 19 are the ancestral types for all Sierra Nevada and Northwestern subclade boas. But of course, I may not be interpreting those figures in the manner you indicate.>>

Yes you are correct. There are 3 trees. I was basing my conclusion on Fig. 3A. Fig. 3B is based on the assumption that the third position of the 3 nucluotide genetic code is more likely to undergo mutation since a change in this position does not change the amino acid that is coded in most cases. In many adaptive molecules, this would be a tenable hypothesis, but in the case of mtDNA, this is a rather dubious assumption since mtDNA has been shown to be a rather neutral molecule (at least the part of the genes used for genetic studies). Fig. 4 is the maximum likelihood tree and it shows the southern rubber boa being ancestral to both the Northwestern and Sierra Nevada subclades. It does show #15, 17, 18, and 19 being slightly more basal to the Sierra Nevada subclade than #26.

>>But if I am, then it would appear that the Calabar and/or Rosy Boa are the ancestral types to the Rubber Boa. >>

Not quite. The Calabar and Rosy Boa are chosen by Rodriguez-Robles et al. as the outgroups. This part is rather arbitrary and it is based on Kluge's hypothesis that the Calabar and Rosy boas are the closest relatives of the rubber boa. Since then, some studies have shown that the Exiliboa is actually the closest relative of the rubber boa and that the Rosy Boa may have descended from an isolated population of the rubber boa in northern Baja california. I have seen studies in which a member of the in group has been chosen as the outgroup. For example, there was one European researcher who picked Lampropeltis, a descendant of a species of Elaphe which migrated from the Old World to the New World, as an outgroup to the genus Elaphe. That researcher was implying that Lampropeltis is ancestral to Elaphe, which is of course absurd.

>>Just because #26 branched off earlier and thus is older than the other specimens, I do not understand just why her lineage and not some other ancestral type, now extinct or still in existence but just haven't been found and tested, may have given rise to the other lineages in the tree.>>

Of course it is certainly possible that some unknown lineage is the true sister group to the Sierra Nevada subclade. Most scientists point out that our closest relative is the chimpanzee, but of course there are a number of extinct hominids, such as Australopithecus and of course Homo erectus, which are actually more closely related to us than the chimpanzee.

>>A good number of years ago, I was surprised at the lack of representation in mtDNA studies. When I questioned that point, I was told that such representation was not necessary as new samples from the same region did not materially add to results. I am now of the opinion that I was correct all along, that inadequate representation can often lead to erroneous interpretations of the results obtained from such studies.>>

More extensive sampling is of course always better. However, no matter how extensive the sampling is, inadequate sampling can always be invoked. In many cases, most mtDNA haplotypes in any one locality tend to become extinct due to chance, so often a single sample from any one locality is good enough. However, in case of recent mixing of two or more populations in one area, a larger sample can be more informative.

>>Interpretations are made on the basis that the samples tested are seemingly the last word and thus finite. Rick's study has demonstrate that is not the case and I venture to say that the current study also lacks adequate representation in a number of regions. Seemingly not taken into account is that there may be a good number of closely related or identical haplotypes of specimens #26 that may occur north, south, east, and west of where #26 originated near Camp Nelson. Without having much larger sample size and geographical representation, to interpret that this or that population migrated in any direction is premature to my way of thinking.>>

One must realize that scientists must base their conclusions on the available facts. No study is the final word. Any new fact that comes into light may and often can falsify an existing theory. One needs to look no further than the case of Lord Kelvin, who miscalculated the age of the earth because he did not know anything about radioactivity.

>>Concerning the distribution of the two size morphs, my sample size from the Tehachapi Mts. and Breckenridge Mt. are large enough to assert that like the San Bernardino boa population, those two populations are also of the dwarf morph. But besides having an adequate sample size, there are other clues that indicate whether the boas from any particular region belong to the dwarf or large morph phenotype. >>

Thanks for the info. It appears then that the most likely explanation for the fact that #26 is a large morph is the intergradation hypothesis.

>>The maximum and mean lengths of neonates of litters produced is one clue. The lengths of mature individuals is another. The lengths at which males and females attain maturity is another. And the rate of growth over time is another. For the populations in which I lack an adequate sample, some of the above clues indicate that the population of boas in the Mt. Pinos area, in the Scodie Mts., in the Greenhorn Mts. south of Alta Sierra all in Kern County, and on the Southern Kern Plateau in Tulare County are all of the dwarf form. Then by geographical proximity, one would expect the dwarf form to occur in the Piute Mts., Alamo Mt., Frazier Mt., Sawmill Mt. and Mt. Abel where the species has been documented in Kern County and the San Jacinto Mts. of Riverside County.>>

If they are all small morph, then it would be interesting to see whether some of their mtDNA haplotypes are closer to those of the San Bernardino/San Jacinto Mt. areas. Presently, the mtDNA data suggests a contraction of the rubber boa range in the Kern and Tulare County area, with rubber boas being absent from Kern County. Subsequently and more recently, dwarf morph boas from Tulare Co. recolonized Kern County. At about the same time, the large morph evolved in Tulare County and migrated north into the Sierra Nevada Mountain, which had previously been closed to both L. zonata and C. bottae from both the north and the south.

>>Until some nuclear markers are found, I believe it is not plausible to determine with any certainty just where a mixture of dwarf and large morph genotypes exists.
>>
>>Richard F. Hoyer

There is no need to resort to nuclear markers, although they would be helpful. Morphology can tell us which morph a boa may belong to and mtDNA can tell us about ancestry. There is a recent paper on Taricha, which shows that the Sierra newt and the Southern California newt intergrade to a limited extent in the Tulare County area. This is quite unexpected because it had been thought that the Southern California newts were restricted to the coast. So, apparently, there is some exchange between the fauna of the southern Sierra Nevada and coastal Southern California.

I think the next researcher to investigate rubber boa phylogenetics should sample the Kern and Tulare County areas much more extensively than before as this area seems to offer the most clues as to how, when and where the migration of the Northwestern and Sierra Nevada subclades may have begun. Rodriguez-Robles et al. did not have the wisdom of hindsight or prior studies to guide their initial sampling, but they did a wonderful job nevertheless picking their samples. Future researches can certainly build on the findings of Rodriguez-Robles et al.

CK.,
Of interest is the notion you proposed that at some point in time, the Rubber Boa may have dispersed southward in part of its range. Going one step further, perhaps at one time, all of the mountainous regions in Kern, Ventura, Los Angeles, San Bernardino, and Riverside Counties (and south into Mexico) of S. Calif. may have been too warm and dry for the existence of C. bottae. Could it be possible that the only remaining refuge for the species was in the high elevations of the southern Sierras in Tulare and Inyo Counties (Mt. Whitney region) and it was from there that the species then expanded in all directions when suitable environmental conditions returned to the mountainous regions to the south.

That is, I now wonder if the species was once extinct in the San Bernardino and San Jacinto Mts. (and elsewhere in S. Calif.) and those areas were recolonized from the dwarf morph that occurs in the southern Kern Plateau in Tulare County.

Perhaps the newest results indicate the reverse, that the dwarf Southern Clade boas in the Kern Plateau arose from the population in the San Bernardino Mts. It seems you have mentioned that age and lineage can be inferred from the raw mtDNA results or am I wrong in that respect?

Arguing against the north to south dispersal scenario is that the Mt. Pinos population seems to be closer morphologically (scale counts and configurations) to the San Bernardino boa population than the dwarf boas from Breckenridge Mt. to the north. On the other hand, just this past year, I did find a few boas from Mt. Pinos with maximum dorsal scale row counts in the 44 - 46 range which overlap the lower range of the large morph boas from the Sierras. That scenario does suggest an influence from more northern populations.

The sample of dwarf boas I have examined from the Kern Plateau is only about 8 - 10 specimens. So at this point, it is not possible to make a reliable comparison of scale counts and head scale configurations between that population and the populations that occur in the San Bernardino Mts.and elsewhere. At any rate, up to now, I had only been thinking of how the Southern Clade dwarf boas dispersed northward from the San Bernardino Mts. to the Kern Plateau and not the reverse.

Your point about having greater representation in areas were clades or subclades were likely to meet was the same argument I presented to Rick and Glenn. And that is why they agreed to test additional samples from the area near Mt. Lassen, from the southern and northern part of the Greenhorn Mts., and from the S. Kern Plateau in which the dwarf vs. large morph scenario was also at play.

And another intriguing aspect of the species is what mechanism(s) are at play that has kept the two morphs reasonably distinct in the Sierra Nevada Mts.? That is, why hasn't the large morph over-run the dwarf form in the southern Greenhorn Mts. and on the southern Kern Plateau? Or conversely, is the dwarf form expanding its distribution into large morph territory?

Richard F. Hoyer

>>CK.,
>>Of interest is the notion you proposed that at some point in time, the Rubber Boa may have dispersed southward in part of its range. >>

Yes, the rubber boa probably dispersed southward to northern Baja California and probably occupied the mountains of San Diego County, Orange County, Coastal Los Angeles County (Santa Monica and San Gabriel Mountains) as well as Ventura County and Santa Barbara County in the past. The large morph Northwestern subclade no doubt migrated from southern Kern County to coastal San Luis Obispo and Monterey Counties via the Sierra Madre but currently there is a distributional gap between the Northwestern subclade and the Kern County boas.

The Rosy Boa originated in northern Baja California according to a recent mtDNA paper, and IMO, this population was once indistinguishable from the rubber boa. Climatic changes probably caused most populations of the rubber boa in Southern California to become extinct. Similarly, we see evidence of such extinction in L. zonata. There almost certainly were L. zonata in the mountains of San Diego Co., Los Angeles Co., San Bernardino County and Orange County at the time the rubber boa occupied this same area, but they all became extinct along with the rubber boa in these mountains.

Instead of seeing L. zonata lineages that rival those of L. z. agalma and L. z. multifasciata in lineage age, which is what is expected if there were no past extinction event, all of the Southern California populations assignable to L. z. parvirubra and L. z. pulchras are very recent lineages derived from L. z. agalma. In fact, the extinction of L. zonata in southern California was even more complete than Charina bottae.

C. bottae managed to hang on, but barely, in the San Bernardino or San Jacinto Mountains. All of the C. bottae populations of these two mountain areas have only recently diverged from a single common ancestor according to mtDNA, suggesting that C. bottae was reduced to a very small population in one of these mountains in the relatively recent past. If biologists think that the southern Rubber boa is endangered now, they might have thought that it was extinct if they can travel back in time when there was probably just a single small population hanging on for dear life.

>> Going one step further, perhaps at one time, all of the mountainous regions in Kern, Ventura, Los Angeles, San Bernardino, and Riverside Counties (and south into Mexico) of S. Calif. may have been too warm and dry for the existence of C. bottae. >>

That is probable. Not only C. bottae, but L. zonata, as well as the blotched form of Ensatina, show signs that they had at one time became extinct or have severely limited distributions in these areas.

>>Could it be possible that the only remaining refuge for the species was in the high elevations of the southern Sierras in Tulare and Inyo Counties (Mt. Whitney region) and it was from there that the species then expanded in all directions when suitable environmental conditions returned to the mountainous regions to the south.>>
>>That is, I now wonder if the species was once extinct in the San Bernardino and San Jacinto Mts. (and elsewhere in S. Calif.) and those areas were recolonized from the dwarf morph that occurs in the southern Kern Plateau in Tulare County. >>

It is certainly possible but not probable. The Tulare County boas are a relatively recent lineage (with the possible exception of the recently discovered umbratica haplotype). If C. bottae migrated from Tulare to San Bernardino County, then we should see small morph snakes with the umbratica haplotype in Southern Kern County. Instead, we see haplotypes that are more derived than #26 in these localities. The San Bernardino and San Jacinto Mountain populations are all descended from a common ancestor very recently, but this common ancestor is very different than sample #26 of Tulare County.

>>Perhaps the newest results indicate the reverse, that the dwarf Southern Clade boas in the Kern Plateau arose from the population in the San Bernardino Mts. It seems you have mentioned that age and lineage can be inferred from the raw mtDNA results or am I wrong in that respect?>>

I haven’t seen the new data. It will depend on how the Tulare Co. specimen with the umbratica haplotype is related to the other boas and to San Bernardino County boas. It is certainly possible that the San Bernardino Mt. Boas are derived from the Tulare County boa with the umbratica haplotype, but the haplotypes of the Kern County boas (samples #30-32) argue against this scenario.

>>Arguing against the north to south dispersal scenario is that the Mt. Pinos population seems to be closer morphologically (scale counts and configurations) to the San Bernardino boa population than the dwarf boas from Breckenridge Mt. to the north. On the other hand, just this past year, I did find a few boas from Mt. Pinos with maximum dorsal scale row counts in the 44 - 46 range which overlap the lower range of the large morph boas from the Sierras. That scenario does suggest an influence from more northern populations.
>>
>>The sample of dwarf boas I have examined from the Kern Plateau is only about 8 - 10 specimens. So at this point, it is not possible to make a reliable comparison of scale counts and head scale configurations between that population and the populations that occur in the San Bernardino Mts.and elsewhere. At any rate, up to now, I had only been thinking of how the Southern Clade dwarf boas dispersed northward from the San Bernardino Mts. to the Kern Plateau and not the reverse.
>>
>>Your point about having greater representation in areas were clades or subclades were likely to meet was the same argument I presented to Rick and Glenn. And that is why they agreed to test additional samples from the area near Mt. Lassen, from the southern and northern part of the Greenhorn Mts., and from the S. Kern Plateau in which the dwarf vs. large morph scenario was also at play.
>>
>>And another intriguing aspect of the species is what mechanism(s) are at play that has kept the two morphs reasonably distinct in the Sierra Nevada Mts.? That is, why hasn't the large morph over-run the dwarf form in the southern Greenhorn Mts. and on the southern Kern Plateau? Or conversely, is the dwarf form expanding its distribution into large morph territory?
>>
>>Richard F. Hoyer

According to the new mtDNA data you disclosed, it appeared that the Northwestern subclade has successfully invaded the northern Sierra Nevada Mountains, and it seems to be outcompeting the Sierra Mountain subclade because the Sierra Nevada Mountain subclade has not dispersed into Northwestern subclade territory. This pattern is the exact reversal of the L. zonata dispersal. In L. zonata, the Sierra Nevada subspecies (L. z. multicincta) has invaded territory used to be occupied by L. z. multifasciata, which shows one remnant population in the vicinity of Ashland, Jackson Co. Oregon. Evidence of L. z. multifasciata influence is the large intergrade zone between L. z. “zonata” and L. z. multicincta, based on morphological grounds. All of the Northern California specimens in the range of L. z. “zonata” have haplotypes belonging to the Sierra Mountain sublclade. That means L. z. “zonata” is actually either L. z. multicincta or L. z. multicincta x L. z. multifasciata.

The Northwestern subclade is probably geographically isolated from the Kern County boas. There appears to be a large gap between the southernmost Northwestern subclade specimen in San Luis Obispo County and the northernmost small morph Kern County speciments. That would probably account for the ability of the small morph to persist in this area. It is shielded from the Northwestern subclade, the most successful lineage within C. bottae. Although some of the Tulare County boas are “large morph,” this particular morph appears to have evolved independently of the Northwestern subclade, since it evolved after the Northwestern subclade last shared a common ancestor with sample #26 and all other Sierra Nevada subclade snakes. Perhaps the Sierra Nevada “large morph” is a more benign competitor and it is able to coexist with the small morph. The Northwestern subclade, for whatever reason, has managed to invade the Sierra Nevada Mountains while preventing the Sierra Nevada subclade from making excursions into its territory. To this date, there is no data to suggest that they can coexist at the same locality.

In the past, I have suggested that it is possible that the Northwestern subclade may have evolved into a different species. However, in view of the gap (now gone) in the distribution of the Sierra Nevada and Norhwestern subclades, there was no data to show whether they have evolved reproductive isolation. I now believe that this question may need to be re-examined in view of new data that shows the Northwestern subclade outcompeting the Sierra Nevada subclade at the intersection of their distribution. Is it possible that the Northwestern subclade individuals simply do not recognize the Sierra Nevada subclade as conspecific and vice versa, therefore there is no intergradation? Or is the sample too small to show evidence of intergradation? Or is it possible that the Northwestern subclade intergrades with the Sierra Nevada subclade but the hybrids are maladaptive? At the same time, the Northwestern subclade seems to be better competitors against the Sierra Nevada subclade. Could it be a combination of both factors which has allowed the Northwestern subclade to expand its range at the expense of the Sierra Nevada subclade.

Could it be body size that allowed the Northwestern subclade to dominate? The record length for L. zonata has been reported in a specimen from the Sierra Nevada Mountains. That means L. z. multicincta probably was able to evolve a larger body size than L. z. multifasciata. The larger body size may have allowed L. z. multicincta to outcompete L. z. multifasciata in Oregon and Northern California north of Sonoma County. Perhaps larger body size allowed the Northwestern subclade of the rubber boa to gain an upper hand on the Sierra Nevada subclade?

CK.,
You mention the large morph Northwestern Clade 'no doubt' migrated west to the San Luis Obispo area and northward to Monterey Co. and beyond. If that scenario were true, then a number of events needed to take place in order to explain the current situation in which the dwarf morph of the Sierra Nevada subclade currently occupies extreme southwestern Kern County.

Also, nothing is known about San Luis Obispo population --- whether it belongs to the Northwestern or Sierra Nevada subclade or if it is of the dwarf or large form. As a matter of fact, very little is known about the boas that occur in Monterey County.

My printout from CAS shows only 5 boas from that county, all collected from near Carmel in the early 1900s. Three boas collected in 1904 and 1905 were lost to the earthquake and fire in S. F. I have examined the other two specimens that were collected in 1908 and 1915.

Both boas were females of 466 mm and 425 mm stretched lengths respectively. The preservation process shrinks specimens to varying degrees with larger specimens incurring a greater percentage of shrinkage than smaller specimens. If those two specimen incurred 10% shrinkage, their stretched live lengths would still be below the lengths of the largest dwarf females estimated to be about 558 mm (22 inches). Not having them in hand now, there is no way of determining whether those females were subadults, adults, or older adults. The only clue is that both had scarred tail tips suggesting they were adults. But even some subadult boas incur scarring of their tail tips.

The only specimen in the MVZ collection from Monterey County was the DOR specimen found on Hwy. 1 in 1998 used in Javier's study. Depending on where specimens from Monterey County occur in other institutional collections (Smithsonian?), that particular boa my have established a new southerly distribution of the species in Monterey County. As best as I can determine from my de Lorme atlas software map, that specimens was found about 40 miles south of the CAS Carmel specimens and about 30 miles north of the San Luis Obispo County line.

I also have examined that specimen which measured 496 mm. It also was a female with a scarred tail tip. Factoring in a 10% shrinkage factor, the specimen's live stretched length would have been 551 mm and below the est.. maximum length of dwarf form females. So at this point, it is not known if either the San Luis Obispo or Monterey County boa populations are of the large or dwarf form. The only known is that the one Monterey Co. specimen tested aligned with the Northwestern subclade.

There is a distance of about 70 miles between the specimen Javier tested in Monterey County and the San Luis Obispo County boa population. A distance of about 90 miles exists between that population and the nearest known boa population to the southeast on Mt. Abel west of Mt. Pinos. Almost a tossup as to which boa population the San Luis Obispo boas are related although it would be my guess that the more likely scenario is they would align with the Northwestern subclade.

Richard F. Hoyer

>>CK.,
>>You mention the large morph Northwestern Clade 'no doubt' migrated west to the San Luis Obispo area and northward to Monterey Co. and beyond. If that scenario were true, then a number of events needed to take place in order to explain the current situation in which the dwarf morph of the Sierra Nevada subclade currently occupies extreme southwestern Kern County.>>

Yes, there is no doubt about it because the Central Valley is not the route of this migration, and because the Sierra Nevada appears to be closed to the migration of both L. zonata and C. bottae until after the coastal populations of both of these species have migrated north. The Sierra Nevada populations of both C. bottae and L. zonata have very little differences in their mtDNA over the entire range of this mountain from north to south, suggesting a recent divergence from their respective common ancestors and a rapid expansion of their ranges for both species.

After L. z. multifasciata and Northwestern subclade C. bottae made their ways north along the coast, the area around Kern, Ventura, Los Angeles, Orange, San Diego, and Santa Barbara Counties probably experienced climatic changes that made them uninhabitable for both C. bottae and L. zonata. The absence of C. bottae from Ventura and Santa Barbara Counties would support this scenario. There is also geological evidence that the San Francisco Peninsula moved north from southern California due to plate tectonics. Perhaps this movement isolated the coastal populations of both species from their southern Sierra Nevada relatives.

As I said before, the current Kern County Populations of C. bottae (samples 30-32) are rather recent lineages, derived from the ancestor of sample #26, which is located in Tulare County. This is evidence of a past contraction of the range of C. bottae in this area. Otherwise, we would see the exact reverse: the Tulare County boas should have been derived from the Kern County boas instead.

>>Also, nothing is known about San Luis Obispo population --- whether it belongs to the Northwestern or Sierra Nevada subclade or if it is of the dwarf or large form. As a matter of fact, very little is known about the boas that occur in Monterey County. >>
>>My printout from CAS shows only 5 boas from that county, all collected from near Carmel in the early 1900s. Three boas collected in 1904 and 1905 were lost to the earthquake and fire in S. F. I have examined the other two specimens that were collected in 1908 and 1915. >>

The Monterey area is probably little different from the nearby Santa Cruz County. Both C. bottae and L. zonata are known from this area. I am surprised so few specimens come from this county. Where L. zonata has been recorded in this area, C. bottae should also be present.

>>Both boas were females of 466 mm and 425 mm stretched lengths respectively. The preservation process shrinks specimens to varying degrees with larger specimens incurring a greater percentage of shrinkage than smaller specimens. If those two specimen incurred 10% shrinkage, their stretched live lengths would still be below the lengths of the largest dwarf females estimated to be about 558 mm (22 inches). Not having them in hand now, there is no way of determining whether those females were subadults, adults, or older adults. The only clue is that both had scarred tail tips suggesting they were adults. But even some subadult boas incur scarring of their tail tips.>>

If you like you can probably make a trip next year to the area and find your own specimens to determine their morph.

>>The only specimen in the MVZ collection from Monterey County was the DOR specimen found on Hwy. 1 in 1998 used in Javier's study. Depending on where specimens from Monterey County occur in other institutional collections (Smithsonian?), that particular boa my have established a new southerly distribution of the species in Monterey County. As best as I can determine from my de Lorme atlas software map, that specimens was found about 40 miles south of the CAS Carmel specimens and about 30 miles north of the San Luis Obispo County line. >>

C. bottae has also been reported in the Landels-Hill Big Creek Preserve, operated by UC. If you contact the biology department of UC Santa Cruz, perhaps they may have some specimens from this preserve, whether they are live or pickled. Students from UC Santa Cruz make frequent trips to this Preserve to conduct field research.

>>I also have examined that specimen which measured 496 mm. It also was a female with a scarred tail tip. Factoring in a 10% shrinkage factor, the specimen's live stretched length would have been 551 mm and below the est.. maximum length of dwarf form females. So at this point, it is not known if either the San Luis Obispo or Monterey County boa populations are of the large or dwarf form. The only known is that the one Monterey Co. specimen tested aligned with the Northwestern subclade.>>

Perhaps a bigger sample will show that they belong to the large morph.

>>There is a distance of about 70 miles between the specimen Javier tested in Monterey County and the San Luis Obispo County boa population. A distance of about 90 miles exists between that population and the nearest known boa population to the southeast on Mt. Abel west of Mt. Pinos. Almost a tossup as to which boa population the San Luis Obispo boas are related although it would be my guess that the more likely scenario is they would align with the Northwestern subclade.
>>
>>Richard F. Hoyer

Even though the distances are about the same, one must not forget that the Kern County boas have only recently (geologically speaking) arrived at their current positions from their ancestors' at Tulare County. At one time the gap between the Tulare County dwarf morph boas and any Northwestern subclade boa wer probably much greater. Further, the lack of boas between San Luis Obispo and southern Monterey Counties may be due to patchy nature of suitable habitat, difficult terrain (for human collectors), lack of roads in suitable areas and the remoteness of the area.

>>I also have examined that specimen which measured 496 mm. It also was a female with a scarred tail tip. Factoring in a 10% shrinkage factor, the specimen's live stretched length would have been 551 mm and below the est.. maximum length of dwarf form females. So at this point, it is not known if either the San Luis Obispo or Monterey County boa populations are of the large or dwarf form. The only known is that the one Monterey Co. specimen tested aligned with the Northwestern subclade.>>

On second thought, Richard, there is of course the possibility that some of the basal lineages of the Northwestern clade has retained the small morph genotype/phenotype of umbratica.

If that is the case, then the large morph may have evolved much farther north than San Luis Obispo or Monterey County. If that is the case, it would be one more piece of evidence that the small morph is the ancestral condition in C. bottae. This would not be a surprise since the basal most members are small morph, and the likely ancestor of C. bottae, namely Exiliboa, is a small snake. If it is indeed the case, it also further supports the theory that the large morph genotype and phenotype evolved at least twice within C. bottae, once in the Northwestern subclade somewhere north of San Luis Obispo County and a second time in the area of Tulare County.

Therefore I shouldn't have taken for granted that the Monterey and/or San Luis Obispo County specimens are large morph boas simply because they belong to the Northwestern subclade. Thanks for the heads up.

CK,
Just to add a bit more to our 'brainstorming'.session.
The sample of vouchers I examined from the Santa Cruz Mts. indicates that population belongs to the large morph. So due to geographical proximity, up to now I too considered it likely that the boa population in Monterey County (and San Luis Obispo Co.) would belong to the large form as well. Yet I always try to remain open to alternative explanations.

When I went through my folders on the vouchers I had examined from the Smithsonian, CAS, and MVZ collections, I found just the three Monterey Co. females all of which were somewhat small. Not much can be gleaned from such a small sample so here we have another region that needs to be explored.

Seems to me that at one time, the mountainous areas of Santa Cruz and Monterey Counties were separate by intrusion of the sea into the central valley. So I wonder if that barrier may have any implication on the nature of those two populations.

And like yourself, I originally 'assumed' that the dwarf morph was ancestral. After confirming that the San Bernardino boa population was indeed a dwarf form of the species, in about 1996, I gave considerable thought to the issue large and dwarf morphs. I devised a hypothesis to explain the likely selection factors whereby the large morph evolved from the dwarf form. At the time, I passed those thoughts on to Glenn Stewart and Dr. Robert Mason here at Oregon State U. but haven't concerned myself with that aspect since.

But unless the raw mtDNA data somehow can ascertain the relative ages of populations, then I have to concede that perhaps my original views were in error and there exists the possibility that the dwarf morph evolved from the large morph.

I know so little about mtDNA research but I can envision one potential piece of evidence that may indicate the relative age of populations. You can give me your thoughts about the following: If a large number of mtDNA haplotypes exist in a relatively small geographical region, wouldn't that indicate that over time, a large number of mutational events had occurred? In contrast, if there are much fewer mtDNA haplotypes in a geographical region, would that possibly indicate a much younger age?

Should there be a large number of different mtDNA haplotypes and few identical haplotypes in the San Bernardino and San Jacinto Mts. in relation to other regions, then perhaps relative age can be inferred. Perhaps the ratio of unique to identical haplotypes per given region is such a clue. The existence of 32 identical haplotypes from N. Calif. to B.C. to Montana, Utah and Nevada may suggest such a scenario.

If the frequency of different and identical haplotypes per geographical area can be interpreted as an indication of age, then here again is another argument for testing much larger samples per any geographical area.

As a side note, as I was looking up the information on the specimens I have examined from Monterey County, I ran across three more old vouchers that originated within the Mt. Lassen area. One specimen from the Smithsonian was collected along Mill Creek in northeastern Tehama County due south of Lassen Peak. Then the CAS collection has two specimens collected in 1926 and 1949 with the former coming from the Lake Almanor peninsula and the other from near Westwood. This latter specimens is just about half way between the Northwestern subclade Eagle Lake specimen and Sierra Nevada subclade Quincy specimen in Javier's study.

It would be my guess that Javier and others entertained the notion of a break in the distribution of the species based on the distance of 120 km they thought occurred between the two subclades, the specimen from Eagle Lake in Lassen County and specimen from Nevada County. Had they realized that their specimen from near Quincy in Plumas County
cut that distance about in half, I don't think they would have indicated a break occurred in the distribution of the species.

I get a sense they may have rushed the publication a bit. And it still strikes me odd that none of the reviewers caught that error. I suspect the reviewers were versed in aspects of molecular research and not in natural history.

Richard F. Hoyer

>>CK,
>>Just to add a bit more to our 'brainstorming'.session.
>>The sample of vouchers I examined from the Santa Cruz Mts. indicates that population belongs to the large morph. So due to geographical proximity, up to now I too considered it likely that the boa population in Monterey County (and San Luis Obispo Co.) would belong to the large form as well. Yet I always try to remain open to alternative explanations.
>>
>>When I went through my folders on the vouchers I had examined from the Smithsonian, CAS, and MVZ collections, I found just the three Monterey Co. females all of which were somewhat small. Not much can be gleaned from such a small sample so here we have another region that needs to be explored.
>>
>>Seems to me that at one time, the mountainous areas of Santa Cruz and Monterey Counties were separate by intrusion of the sea into the central valley. So I wonder if that barrier may have any implication on the nature of those two populations.>>

The mountains of Monterey and Santa Cruz Counties are separated by the Salina Valley, which is probably an effective barrier to the dispersion of some species. For example, the northern and southern newts currently assigned to Taricha torosa are allopatric, separated by the Salinas Valley, with one population to the north and the other to the south, without any evidence of mixing. On the other hand, this area appears to be ineffective in isolating L. zonata populations from one another.

>>And like yourself, I originally 'assumed' that the dwarf morph was ancestral. After confirming that the San Bernardino boa population was indeed a dwarf form of the species, in about 1996, I gave considerable thought to the issue large and dwarf morphs. I devised a hypothesis to explain the likely selection factors whereby the large morph evolved from the dwarf form. At the time, I passed those thoughts on to Glenn Stewart and Dr. Robert Mason here at Oregon State U. but haven't concerned myself with that aspect since.>>
>>But unless the raw mtDNA data somehow can ascertain the relative ages of populations, then I have to concede that perhaps my original views were in error and there exists the possibility that the dwarf morph evolved from the large morph.>>

It is highly unlikely, but certainly possible, that the large morph is ancestral. MtDNA data can in fact ascertain the relative ages of populations. It is a rather neutral molecule, which means nucleotide substitution is due almost entirely to chance, much like radioactive decay. The molecular clock is of course not as accurate as the atomic clock, and of course the molecular clock must be calibrated with the fossil record, which is highly fragmentary. What we see in the rubber boa is that the umbratica lineage is the oldest. How do we know? Basically we know because we see a large number of differences in the mtDNA compared to other rubber boa populations. The greater the number of these mutations, the greater the amount of time has elapsed since two populations last shared a common ancestor. All of the populations within umbratica, OTOH, share very few differences. That means all umbratica populations share a very recent common ancestor. That is how I came to the conclusion that umbratica almost became extinct in the recent past.

Similarly, if you take two boas from, say, the Sierra Nevada subclade, you will see that they are also very similar to each other in their mtDNA, suggesting again that these populations share a relatively recent ancestor. But if you compare the mtDNA of the Sierra Nevada subclade and Northwestern subclade, you will see a greater number of differences. The greatest number of differences in mtDNA can be found if you compare the mtDNA of an umbratica with the mtDNA of a Northwestern subclade boa, or if you compare umbratica with the Sierra Nevada subclade. It shows that the umbratica lineage has been isolated from other boas forthe longest period of time. This lineage has accumulated a large number of mutations in its mtDNA that are unique to itself and not shared with any other lineage. It is this large number of nucleotide substitutions that has enabled Rodriguez-Robles to estimate the age of the rubber boa as a species. It takes time to accumulate all of these mutations. Therefore umbratica appears the oldest lineage. Other lineages, such as the Sierra Nevada subclade, are younger, because they share a number of mutations with the Northwestern subclade.

>>I know so little about mtDNA research but I can envision one potential piece of evidence that may indicate the relative age of populations. You can give me your thoughts about the following: If a large number of mtDNA haplotypes exist in a relatively small geographical region, wouldn't that indicate that over time, a large number of mutational events had occurred? In contrast, if there are much fewer mtDNA haplotypes in a geographical region, would that possibly indicate a much younger age? >>

See above for how lineage age is determined. If you see a large number of haplotypes in an area, it may simply mean that the populations within a particular area are isolated from each other, so that they are unable to mix freely. For example, there are several distinct mtDNA haplotypes within Ambystoma tigrinum californiense within its relatively limited range. That only indicates that A. t. californiense from different parts of its range has been prevented from interbreeding freely probably because of geographical barriers. To measure lineage age, one must examine the number of nucleotide substitutions that have occurred, not the number of haplotypes that may be present.

>>Should there be a large number of different mtDNA haplotypes and few identical haplotypes in the San Bernardino and San Jacinto Mts. in relation to other regions, then perhaps relative age can be inferred. Perhaps the ratio of unique to identical haplotypes per given region is such a clue. The existence of 32 identical haplotypes from N. Calif. to B.C. to Montana, Utah and Nevada may suggest such a scenario. >>

The haplotypes of the Northwestern subclade are not identical. The differences in mtDNA sequence between populations within this subclade are relatively large. If you look at fig. 4 of Rodriguez-Robles, you will see that the branch lengths between different populations are much longer within this subclade than in either the Sierra Nevada subclade or in umbratica. That means there are many more mtDNA sequence differences between a boa from Santa Cruz County and Berkeley than there are differences between a boa from El Dorado County and Madera County. All of the boas of the Northwestern subclade, however, do share a more recent common ancestor with each other they they do with either the Sierra Nevada subclade or umbratica.

>>If the frequency of different and identical haplotypes per geographical area can be interpreted as an indication of age, then here again is another argument for testing much larger samples per any geographical area.>>

Unfortunately it cannot. If you go to New York City, you will see a lot of different mtDNA haplotypes among the human beings living there. It does not mean that New York City has one of the oldest human populations in the world. If you go to Africa, however, you will see that the mtDNA differences between two Africans from different parts of Africa are often much greater than the mtDNA differences between two non Africans anywhere else in the world. That means all non-Africans in the human species share a more recent common ancestor with each other than many Africans do with each other. Normally, we would expect to see the same in umbratica. We should see lots of old lineages within umbratica, comparable to what we see among Africans. But instead we only see a single ancient umbratica lineage with lots of newly evolved branches at the top. That is indicative of a recent bottleneck, or population crash within umbratica. Without the presence of the umbratica lineage, scientists may have concluded that C. bottae is a much younger species than it really is.

>>As a side note, as I was looking up the information on the specimens I have examined from Monterey County, I ran across three more old vouchers that originated within the Mt. Lassen area. One specimen from the Smithsonian was collected along Mill Creek in northeastern Tehama County due south of Lassen Peak. Then the CAS collection has two specimens collected in 1926 and 1949 with the former coming from the Lake Almanor peninsula and the other from near Westwood. This latter specimens is just about half way between the Northwestern subclade Eagle Lake specimen and Sierra Nevada subclade Quincy specimen in Javier's study.
>>
>>It would be my guess that Javier and others entertained the notion of a break in the distribution of the species based on the distance of 120 km they thought occurred between the two subclades, the specimen from Eagle Lake in Lassen County and specimen from Nevada County. Had they realized that their specimen from near Quincy in Plumas County
>>cut that distance about in half, I don't think they would have indicated a break occurred in the distribution of the species.
>>
>>I get a sense they may have rushed the publication a bit. And it still strikes me odd that none of the reviewers caught that error. I suspect the reviewers were versed in aspects of molecular research and not in natural history.
>>
>>Richard F. Hoyer

Well, biology is a highly specialized discipline nowadays. It is often difficult for one researcher working on one species to judge the work of another researcher working on another species. Besides, professional biologists are often very busy, teaching, doing research, and so on, so reviewing papers often does not get the amount of time or attention that it deserves.

>>CK,

>>But unless the raw mtDNA data somehow can ascertain the relative ages of populations, then I have to concede that perhaps my original views were in error and there exists the possibility that the dwarf morph evolved from the large morph.
>>
>>I know so little about mtDNA research but I can envision one potential piece of evidence that may indicate the relative age of populations. You can give me your thoughts about the following: If a large number of mtDNA haplotypes exist in a relatively small geographical region, wouldn't that indicate that over time, a large number of mutational events had occurred? In contrast, if there are much fewer mtDNA haplotypes in a geographical region, would that possibly indicate a much younger age?>>

Hi, Richard, I gave a pretty long winded and convoluted answer in my original answer to your post. So I will try to be more coherent, if not more concise, this time. Since life began some 3.5 billion years or so ago, it has evolved into the present bewildering variety. Since all life can be traced back to a single ancestor, all organisms have the same absolute age. When we talk about the age of a population, we are really talking about how long ago a population has diverged from a common ancestor, either with another population of the same species or a different species. To measure how long ago two organisms last shared a common ancestor, biologists turn to something called genetic distance. Genetic distances are best measured using molecules because the rates of morphological evolution vary greatly. Some organisms change very little morphologically over hundreds of millions of years, and living individuals look almost identical to fossils that are hundreds of millions of years old. The coelacanth is a good example. Some organisms can change drastically over a short period of time. So it is not possible to know how long since two organisms last shared a common ancestor by measuring morphological disparity.

Molecules, however, evolve more or less in a clock like manner, because chance mutations occur randomly and the probability of a mutation during DNA/RNA replication is quite similar for a vast majority of organisms. Of course many molecules change very little over the eons because changes can make them non-functional. The chlorophyll molecule is a very good example of an adaptive and slowly evolving molecule that resists changes. Other molecules are less vital and they can change more rapidly. The late Allan Wilson of UC Berkeley, a pioneer in molecular systematics, realized this. He and his students contributed a great deal to the understanding of the relationships and relative ages of divergence among different living organism by measuring molecular distance using more or less neutral molecules. To do this, he searched for molecules that are not vital to an organism's survival. One of those molecules he found was serum albumin, a protein with no known function, and which can mutate rather freely. Another molecule he found is mtDNA.

By measuring or estimating the genetic distance (i.e. the number of differences in their nucleotide sequences) using mtDNA or other molecules, the amount of time that has elapsed since two organisms last shared a common ancestor can be estimated. The theory is that when two organisms share the same parent, their mtDNA molecules are identical. As these organisms become separated geographically and evolved, the number of nucleotide differences would increase through time. Of course, the molecular clock must be calibrated using fossils, so there are disagreements among biologists as to how fast a molecule may have evolved over time, if their calibrations are different.

The genetic distance between umbratica and the northern subclades (Northwestern plus Sierra Nevada subclades) is the greatest within C. bottae. The distance between the Northwestern subclade and the Sierra Nevada subclade is much smaller. And of course the genetic distance between any two populations within the Sierra Nevada subclade is among the smallest seen within this species. The distances between any 2 population of umbratica is also very short, suggesting a relatively recent common ancestor.

The last common ancestor of umbratica and the 2 northern subclades is the ancestor to all of the C. bottae populations existing today. Was this common ancestor found in southern California or the southern Sierra Nevada? We may never know the answer to that question because we cannot observe history but we can only infer it. Chances are, however, that this ancestor lived in southern California because the likely ancestor of both C. bottae and Lichanura trivirgata is the Mexican dwarf boa. Therefore it is a simpler (but not necessarily correct) and more likely explanation is that Exiliboa gave rise to C. bottae in Southern California before retreating south into Mexico due to climatic and geological changes occurring in the area of northwestern Mexico.

If umbratica is ancestral to the 2 northern subclades, then the ancestral form of C. bottae is more likely to be the dwarf form. Again, this would be the simpler (but again not necessarily correct) and more likely explanation, since Exiliboa is a small snake that thrives in mesic environments. If the large morph is ancestral, then we would need Exiliboa to give rise to a large morph snake that is perhaps better adapted to xeric environments (like those boas found in eastern Oregon for example), and from this large morph animal, a small morph snake adapted to a more mesic environment re-evolved. Evolutionary biologists call this an evolutionary reversal. Reversals are of course possible and in some instances they are even probable. In most other cases, they are less likely because for biological organisms, it is often not possible to "go home again." For an organism like a whale, an evolutionary reversal back to life on land is probably not possible because whales have become so well adapted to living in the sea and because they have lost almost all traces of their hindlimbs.

Therefore, on the basis of genetic distance, biogeography and the fact that reversal is less likely, I believe that the dwarf morph is the ancestral form of C. bottae.

CK
Too busy to comment at this time. Will get back to this thread later.

RF Hoyer

>>CK
>>Too busy to comment at this time. Will get back to this thread later.
>>
>>RF Hoyer

Happy Holidays, Richard.

>>CK,
>> But of course, I may not be interpreting those figures in the manner you indicate.
>>
>>But if I am, then it would appear that the Calabar and/or Rosy Boa are the ancestral types to the Rubber Boa.
>>

I would like to elaborate a little bit on this point. The Calabar and Rosy Boas are chosen by Rodriguez-Robles et al. as the outgroup for the rubber boa, ostensibly because they relied on Kluge's morphological analysis, which has since been proven to be inaccurate. That means these two species were thought to be close relatives or possible ancestors, but not descendants, of the Rubber Boa, but they may not. The choice of outgroup is important because it can potentially mean success or failure of the analysis. An outgroup is chosen to "anchor the tree." There are of course phylogenetic studies in which the tree is not anchored. These trees are called "unrooted." It is difficult to tell from an unrooted tree which particular species or population is the most basal. Hence most researchers choose to root their trees with a carefully chosen outgroup.

An outgroup serves more than anchoring a tree. They provide a set of characters (either molecular or morphological) that are likely to be ancestral within the group being analyzed. For example, an analysis of the relationships among terrestrial vertebrates would often utilize a fish of some sort as an outgroup. An analysis of relationships among reptiles may utilize an amphibian as an outgroup. Characters that are shared between the outgroup and the ingroup would most likely be ancestral characters. For example, the vertebral column is shared between fish, amphibian and reptile, hence this character is regarded as an ancestral character. If an amphibian or reptile have evolved without a vertebral column, then most likely this is a new character state (although it must be pointed out that this hypothetical situation has not evolved).

Fish do not have fingers and toes, but most terrestrial vertebrates do. This means fingers and toes can be used to identify the basal most group of tetrapods. Some, more derived groups such as the caecilians also lack fingers and toes, but since they share other features with tetrapods, their lack of fingers and toes identify them as a more derived group than the basal tetrapods. If, however, the caecilians were chosen erroneously as the "outgroup" in an analysis of the tetrapods, then one may mistake other features found in the caecilians as basal to all tetrapods. For example, the features the caecilians share with salamanders, frogs and toads may result in the Lissamphibians being classified as among the most basal tetrapods, whereas the earliest known amphibians would be misclassified as much more derived. Hence it is very important not to choose a member of the ingroup as the outgroup.

When molecular characters are used to analyze relationships, it is often difficult to tell whether a particular nucleotide substitution is ancestral or not, unlike well studied morphological characters like the vertebral column. If a particular molecular character is shared by both the outgroup and most (but not necessarily all) of the members of the ingroup, then systematists can be reasonably sure that this character is an ancestral one. When the ancestral characters have been identified, systematists (or their computer programs) will start looking for more recently evolved characters that can identify the next branching points.

For example, if character a (analogous to fingers and toes in the example above) is not found in the outgroup but is shared by almost all members of the ingroup, then this character most likely evolved after the ingroup has become isolated from the outgroup. Fingers and toes are examples of the so called "synapomorphs" or shared derived characters.

If another character, b, for example, is shared by all boas north of the San Bernardino mountains, but not found in either the outgroup or umbratica, then this character probably evolved in the last common ancestor of the northern subclade (Northwestern Sierra Nevada). The search continues to identify characters that are shared by a progressively less inclusive groups. That is how systematists are able to identify branching points and subgroups within the tree. Of course, convergence can occur and sometimes a section of a gene can be lost due to mutation. So it is never as easy as one may imagine and the outcome may not be as perfectly reliable as a textbook perfect example may be. Of course if the outgroup is chosen incorrectly, the tree can be downright anomalous.

If the Rosy Boa is in fact a descendant of the Rubber Boa, as a recent mtDNA study of the Rosy Boas suggests to ME personally, then its choice as an outgroup in Rodriguez-Robles et al.'s analysis of the rubber boa may have caused some of the result to be anomalous. I said "may" because the Rosy Boa may have been sufficiently far removed phylogenetically from most members of the Northern subclades to have misled the analysis.

Nevertheless, I think it wise to exclude the Rosy Boa when the next researcher analyzes rubber boa taxonomy. A more suitable choice as outgroup, which is suggested by some recent independent mtDNA analyses, is Exiliboa, the Mexican dwarf Boa. Studies have shown that this species is probably ancestral to both the Rubber and Rosy Boas. Hence there is little chance that Exiliboa may have descended from Charina bottae. If Exiliboa is used as the outgroup for a combined analysis of Charina bottae and Lichanura trivirgata with a large number of samples of each species, then perhaps there is clear evidence of how the Rubber and Rosy Boas are related to each other.

>>1) Beside occurring in the San Bernardino and San Jacinto Mts., a boa population of the Southern clade also occurs on the Southern Kern Plateau in southern Tulare County. >>

Hi, Richard. Thanks for the data. As I have pointed out before, the dwarf morphotype of the southern subclade (umbratica) is most likely the ancestral form of this species. This same morphotype is, according to your findings, found in the Kern County populations. That is evidence of stasis. The Kern County specimens have become isolated from umbratica for a long period of time without undergoing any evolutionary changes. Their mtDNA of course will continue to evolve, since these nucleotide substitutions are mostly neutral genetic changes that occur in a more or less clock like manner. The latest finding of an umbratica haplotype found in Tulare county is clear evidence that these two populations were once connected geographically. Of course they have to be because their mtDNA suggests that the Kern County animals are derived from the southern subclade. As I said before, the Kern County dwarf morphs should be included in umbratica, and umbratica should be treated as a subspecies of bottae, not a full species.

>>2) There is likely to be no break between the Sierra Nevada and Northwestern subclades as two specimens well south of the Mt. Lassen region from Butte County aligned with the Northwestern subclade.>>

That is an interesting result. It clearly is in agreement with the mtDNA findings of Rodriguez-Robles et al., which shows that the northwestern subclade migrated along the coast and at an earlier date that the Sierra Nevada subclade. It seems, then, that the Northwestern subclade has been able to move close to the northern limits of the Sierra Nevada subclade. It would be intresting to see if there is any overlap between the two. Until there is evidence of overlap, it remains an unanswered question.

>>3) Thirty-two samples with the same haplotype were dispersed from northern California to Oregon, Washington, B.C., Montana, Idaho, Utah, and I believe Nevada.>>

That is interesting but not unexpected. It certainly looks like the Northwestern subclade has been a very successful lineage and it again confirms Rodriguez-Robles data, which shows the Northwestern subclade is the first lineage to move into Northern California and beyond. If the Nevada specimens belong to the Northwestern subclade, then there was probably no back door migration of the dwarf morph Kern County animals in a northeasterly direction into Nevada.

>>The ramifications of point #1 above are extensive and intriguing.
>>
>>Richard F. Hoyer

Finding #1 is interesting indeed. It remains to be seen whether this population migrated to Tulare recently, or whether it is a remnant of an old lineage. It is also interesting how much difference there is between the mtDNA of this population and that of the SRB. As I pointed out before, all of the populations of the SRB have similar mtDNA, suggesting a recent divergence and expansion from a small population, probably within a small refuge. My guess is that the Tulare County population is a much older remnant of the ancestral bottae population that made its way a long long time ago from the south. Again, very exciting new findings that fill that gaps of Rodriguez-Robles' data. It shows that Rodriguez-Robles et al. did a great job.

CK,
Concerning your point #2: You may recall that Javier suggested that a break occurred in the distribution of the boa in the area of Mt. Lassen National Park and thus indicated the two subclades had an allopatric distribution. New information suggests that no such break in the species' distribution occurs in the region of Mt. Lassen for the following reasons:

1) Two samples of the Northwestern subclade occur well south of Mt. Lassen in northern Butte County and almost due west of where a specimen tested by Javier near Quincy aligned with Sierra Nevada subclade. Butte County adjoins Plumas County to the southwest. Three other samples tested from southeastern Butte County align with the Sierra Nevada subclade. That is, both subclades occur in Butte County with continuous suitable habitat connecting the two regions.

2) As mentioned above, in Javier's study, the one sample he tested (and overlooked in his treatment of the subclades) from just east of Quincy in central Plumas County aligned with the Sierra Nevada subclade. An additional sample in the new study just west of Quincy also aligns with the Sierra Nevada subclade.
However, two samples in extreme northwestern Plumas county just northwest of Chester align with the Northwestern subclade. Here again, we have both subclades occurring in Plumas County with suitable boa habitat occurring throughout that region.

3) The two samples from northwestern Plumas County plus two specimens found and tested from extreme northeastern Tehama County are in the area where a break in the species' distribution was support to occur. The two Tehama County specimes also align with the Northwestern subclade.

4) This summer, a biologist with the CDFG observed a specimen on a road located at the east central side of Lake Almanor. The locality of that specimen cuts the distance about in half between the specimens near Quincy and the specimens northwest of Chester thus adding support to the likelihood that the two subclades are at least parapatric and more than likely sympatric. It makes no difference to which subclade that specimen would align.

But I agree that in spite of a few glitches, the paper by Javier was an exceptional contribution to our understanding of the species.

Richard F. Hoyer

>>CK,
>>Concerning your point #2: You may recall that Javier suggested that a break occurred in the distribution of the boa in the area of Mt. Lassen National Park and thus indicated the two subclades had an allopatric distribution. New information suggests that no such break in the species' distribution occurs in the region of Mt. Lassen for the following reasons: >>

Their conclusion was scientific because it was based on the available evidence at the time. If and when the available evidence changes, then the theory must also change. I have no problem with the provisional nature of all scientific theories.

>>1) Two samples of the Northwestern subclade occur well south of Mt. Lassen in northern Butte County and almost due west of where a specimen tested by Javier near Quincy aligned with Sierra Nevada subclade. Butte County adjoins Plumas County to the southwest. Three other samples tested from southeastern Butte County align with the Sierra Nevada subclade. That is, both subclades occur in Butte County with continuous suitable habitat connecting the two regions. >>

Yes, and according to what you said about the mtDNA of these samples, they belong to the Northwestern subclade, not the Sierra Nevada subclade. That is almost the reverse of the situation seen in Lampropeltis zonata. In L. zonata, the mtDNA haplotype that are found in this region happened to be L. z. multicincta, the Sierra Nevada Mountain haplotype, not L. z. multifasciata. In fact, L. z. multicincta was so successful that it swarmed the previously arrived L. z. multifasciata in Oregon and Northern California (north of the S.F. Bay). There is but one single sample with the L. z. multifasciata haplotype found in Rodriguez-Robles' paper, and it was sample 31, found in the area of Ashland, Jackson Co, Oregon.

That shows that both L. zonata multifasciata and Northwstern clade C. bottae, despite migration along the coast earlier their the Sierra Nevada conspecifics, were both unable to invade the Sierra Nevada Mountains, leaving this mountain range free for the invasion by L. zonata and C. bottae from Tulare County. The question then becomes: what barrier was presented by nature that had successfully prevented both L. z. multifasciata and C. bottae (Northwestern subclade) from entering the Northern Sierra Nevada? This question requires further investigation.

It seems that whatever barrier may have existed, it is now gone, because L. z. multicincta was able to traverse it and migrate north and west beyond the Sierra Nevada Mountains, to reach not only Washington but coastal Oregon, and Northern California from Sonoma County north to the Oregon border. And this migration was so successful that it swamped the pre-exisiting L. z. multifasciata haplotypes, leaving L. z. multifasciata represented by a single locality in the sample. Phenotypically, the snakes in this broad contact zone still exhibits some traits of L. z. multifasciata (e.g. the wider red bands). L. z. zonata is apparently little more than L. z. multifasciata x L. z. multicincta minus the red snout of multifasciata.

Turning our attention back to C. bottae, we don't see any evidence of the Sierra Nevada subclade in Oregon, Washington or in coastal California. The success of L. z. multicincta has not been duplicated by Sierra Nevada subclade C. bottae. Instead, the latest data shows that Northwestern subclade C. bottae made it all the way to the northern edge of the Sierra Nevada Mountains. Why the difference? Is it possible that the Sierra Nevada boas are poorly equipped to compete against the Northwestern subclade except in the higher elevations of the Sierra Nevada Mountains? What then accounts for the tremendous success of L. z. multicincta? These are fascinating questions that invite further investigation.

>>2) As mentioned above, in Javier's study, the one sample he tested (and overlooked in his treatment of the subclades) from just east of Quincy in central Plumas County aligned with the Sierra Nevada subclade. An additional sample in the new study just west of Quincy also aligns with the Sierra Nevada subclade.
>>However, two samples in extreme northwestern Plumas county just northwest of Chester align with the Northwestern subclade. Here again, we have both subclades occurring in Plumas County with suitable boa habitat occurring throughout that region.>>

Fascinating. It appears then that the Northwestern subclade has indeed invaded the Sierra Nevada Mountains. But again, why? Why hasn't the Sierra Nevada subclade successfully migrated into Northwestern subclade territory? It appears then, from analyzing the mtDNA haplotypes of both L. zonata and C. bottae that there is no current or recent barriers to migration of either species between the Mt. Lassen region and the Sierra Nevada Mountains. Therefore one must look for biological reasons for the differences in the success of the Sierra Nevada populations of C. bottae and L. zonata.

>>3) The two samples from northwestern Plumas County plus two specimens found and tested from extreme northeastern Tehama County are in the area where a break in the species' distribution was support to occur. The two Tehama County specimes also align with the Northwestern subclade.
>>
>>4) This summer, a biologist with the CDFG observed a specimen on a road located at the east central side of Lake Almanor. The locality of that specimen cuts the distance about in half between the specimens near Quincy and the specimens northwest of Chester thus adding support to the likelihood that the two subclades are at least parapatric and more than likely sympatric. It makes no difference to which subclade that specimen would align.>>

It is now clear to me that the Northwestern subclade has successfully invaded the Sierra Nevada Mountains, and judging from the lack of Sierra Nevada subclade specimens in the Mt. Lassen region, the Sierra Nevada subclade is being driven south by the Northwestern subclade. Give it some time, perhaps a few tens of thousands of years, or perhaps a couple of million years, and we may see no trace at all of the Sierra Nevada subclade haplotype within the Sierra Nevada Mountains. A conservation biologist of the future may need to designate the Sierra Nevada subclade as a threatened or an endangered species.

>>But I agree that in spite of a few glitches, the paper by Javier was an exceptional contribution to our understanding of the species.
>>
>>Richard F. Hoyer

I agree. Rodriguez-Robles' mtNDA studies are consistent high in quality. The same cannot be said of some of the molecular studies I have seen from other researchers.

CK,
You mention: "Their conclusion was scientific because it was based on the available evidence at the time."

We have discussed this point before. I contend that there was no evidence and thus no scientific basis for making the assertion that a break occurred in the distribution of the Rubber Boa in the vicinity of Mt. Lassen Nat. Park. When I read the draft of the paper, that very point stood out like a sore thumb. But at the time, I wasn't able to convince Glenn of my reasoning and it seems that I have not been convincing with you as well.
.
Having tissue taken and tested from two specimens from distant localities does not represents scientific evidence of a break in distribution for any species. Had the two nearest specimens tested from the two subclades been from near Portland, Oregon and Yosemite Park, would that indicate a break in the species distribution somewhere between those two localities?

Where samples of tissue originate has absolutely no bearing whatsoever on the distribution of a species whether such samples were taken from specimens 500 meters or 500 km apart.. And I found no attempt at any scientific inquiry with respect to whether or not a break occurs in the boas distribution in that region.

Had the authors conducted some survey for the species in that region, had they made inquiries with rangers, biologists, ranchers, and other residents in the region, had they undertaken an analysis of habitat and elevation components of the species distribution and accepted the concept of habitat association, had they conferred with experts in the history of climate and geology in the region, or had they simply consulted institutional records for vouchers in that region, then I would agree that they had some scientific basis to assert that a break occurs in the species distribution.

The first error was indicating a distance of 120 km between the two nearest specimens of each subclade, one at Eagle Lake (Northwestern Subclade) and one in Nevada County (Sierra Nevada subclade) by overlooking a much closer Sierra Nevada subclade specimen that occurred just east of Quincy in central Plumas County. And had the authors simple examined the list of vouchers in the MVZ collection at Berkeley, they would have noted two specimens closer to Mt. Lassen then either the Eagle Lake or Quincy specimens.

The MVZ printout (I obtained in about 2000) has a specimen found at Silver Lake in Lassen Co. which is about 15 mile due east of the summit of Mt. Lassen and about 25 -30 miles southeast of the Eagle Lake specimen. Also, there is a specimen from Morgan Springs (Ranch) about 10 -12 miles due south of the Mt. Lassen summit in northeastern Tehama Co. and about 12 - 15 miles west of Chester. Either of those two specimens dispel any notion of a break occurring in the species' distribution in the Mt. Lassen region.

Richard F. Hoyer

>>CK,
>>You mention: "Their conclusion was scientific because it was based on the available evidence at the time."
>>
>>We have discussed this point before. I contend that there was no evidence and thus no scientific basis for making the assertion that a break occurred in the distribution of the Rubber Boa in the vicinity of Mt. Lassen Nat. Park. When I read the draft of the paper, that very point stood out like a sore thumb. But at the time, I wasn't able to convince Glenn of my reasoning and it seems that I have not been convincing with you as well.>>.

Feeling or intuition is very useful and it has often guided scientists to great discoveries and it has also helped overturn existing orthodoxy. In this case, your "feeling" has proven correct. You are justified to criticize Rodriguez-Robles' conclusion. They appeared to have treated the lack of evidence of the presence of rubber boas in the Mt. Lassen region as evidence of absence. They perhaps should have concluded that this region may have been poorly sampled and concluded that it is unknown whether the Northwestern subclade and Sierra Nevada subclade may overlap in distribution. However, Rodriguez-Robles' conclusion is also valid based on the available evidence at the time, as they probably have no way of knowing just how poorly sampled the Mt. Lassen area may have been.

>>Having tissue taken and tested from two specimens from distant localities does not represents scientific evidence of a break in distribution for any species. Had the two nearest specimens tested from the two subclades been from near Portland, Oregon and Yosemite Park, would that indicate a break in the species distribution somewhere between those two localities?
>>
>>Where samples of tissue originate has absolutely no bearing whatsoever on the distribution of a species whether such samples were taken from specimens 500 meters or 500 km apart.. And I found no attempt at any scientific inquiry with respect to whether or not a break occurs in the boas distribution in that region.>>

These criticisms are justified and valid scientific criticisms. Of course, anyone is free to disagree with Rodriguez-Robles' theory and disprove them with new facts. And you and your collaborators have done that by providing new specimens from this area. I congratulate you on your efforts and persistence.

>>Had the authors conducted some survey for the species in that region, had they made inquiries with rangers, biologists, ranchers, and other residents in the region, had they undertaken an analysis of habitat and elevation components of the species distribution and accepted the concept of habitat association, had they conferred with experts in the history of climate and geology in the region, or had they simply consulted institutional records for vouchers in that region, then I would agree that they had some scientific basis to assert that a break occurs in the species distribution. >>

New distributional records for many species are published every month within the pages of Herpetological Review. There is certainly a lot that we do not know about the distribution of many species, even the most well known ones. Snakes are secretive by nature, and their true abundance and distribution have often been shrouded in mystery, and of course the Rubber Boa is a well known example. It was once thought to be very rare, but your research helped shatter that misconception. Perhaps Rodriguez-Robles et al. were too hasty in concluding that there was a break in distribution. However, the data shows no overlap between the two subclades, and even with additional samples there is still no evidence that they occur in the same locality. Their data may have misled them into thinking that there was a break. Rodriguez-Robles is also the senior author of the mtDNA paper on Lampropeltis zonata. He may not have seen the similarities in distributional history between L. zonata and C. bottae. If he had, he might have wondered why L. zonata multicincta was able to move past the Sierra Nevada mountains and made it all the way to Oregon and Northern California and therefore there is little reason for Charina bottae not to be able to do the same. But then again, are scientists correct that there is a break in the distribution of L. zonata in Oregon? Perhaps the range of L. zonata is continuous from Oregon to Washington? At some point, scientists have to decide whether the evidence avaialable is reasonably complete or not. In hindsight, Rodriguez-Robles made the incorrect decision. But that is life, you win some and you lose some.

>>The first error was indicating a distance of 120 km between the two nearest specimens of each subclade, one at Eagle Lake (Northwestern Subclade) and one in Nevada County (Sierra Nevada subclade) by overlooking a much closer Sierra Nevada subclade specimen that occurred just east of Quincy in central Plumas County. And had the authors simple examined the list of vouchers in the MVZ collection at Berkeley, they would have noted two specimens closer to Mt. Lassen then either the Eagle Lake or Quincy specimens.
>>
>>The MVZ printout (I obtained in about 2000) has a specimen found at Silver Lake in Lassen Co. which is about 15 mile due east of the summit of Mt. Lassen and about 25 -30 miles southeast of the Eagle Lake specimen. Also, there is a specimen from Morgan Springs (Ranch) about 10 -12 miles due south of the Mt. Lassen summit in northeastern Tehama Co. and about 12 - 15 miles west of Chester. Either of those two specimens dispel any notion of a break occurring in the species' distribution in the Mt. Lassen region.
>>
>>Richard F. Hoyer

I don't know if their printout would have showed the same results or not, because their paper was finished prior to AD 2000. But, again, if scientists base their conclusions on the available evidence, then they have a scientific theory supported by evidence. Of course, as the late S.J. Gould pointed out, scientists are wrong most of the time, because new facts turn up constantly. Unless one has the ability to see these new facts before they are discovered, nothing much will change, and scientists will make wrong conclusions well into the future. However, there is no harm when scientific theories are falsified, although there often is harm to scientific careers. Hence, even though scientific theories are meant to be tested and falsified by new, undiscovered facts, some scientists often act as though they are religious zealots and treat their own theories as indisputable, unfalsifiable objective truth.

CK,
Actually, intuition was not involved with the manner in which I viewed the claim that a break occurred in the boa's distribution in the Mt. Lassen region. Instead, I relied on existing evidence, application of basic biological concepts, and rational thought processes to support my position. So where we disagree is that I cannot find any evidence, of a scientific nature, that would support the claim of a break in the species' distribution.

I do not consider that randomly collected vouchers which by chance, had tissue taken and tested, as representing evidence for claiming a break in any species' distribution let alone that of the Rubber Boa. But I do understand how the authors were misled by their data in reaching such a conclusion. After all, their data do show the two subclades as having quite separate geographical distributions.

How easy it is to criticize as it is free. But then one has to be aware of 'he who dare casts the first stone be without sin---or similar sage advice. Hah.

When I contacted Glenn about this point, I wasn't as 'forceful' as I am here as I thought that surely that glitch would be caught during the peer review process. But such glitches aside, I agree that the results obtained by Javier's mtDNA research was an important step forward. In addition, at least for the time being, it also put to rest the notion of the Great Basis subspecies (C. b. utahensis).

I certainly agree with your last paragraph. I do not know exactly what Gould mentioned but I agree in the broad context. Too many individuals in and out of science accept scientific findings as if etched in stone.

Also, I have read and reviewed a number of scientific publications in which the experimental design and results obtained were quite nice. The mistakes I see being made are in the interpretation of those results. This is particularly true in the field of Conservation Biology where individuals have pre-existing biases that interfere with analyzing their results in an impartial and objective manner.

Richard F. Hoyer

>>CK,
>>Actually, intuition was not involved with the manner in which I viewed the claim that a break occurred in the boa's distribution in the Mt. Lassen region. Instead, I relied on existing evidence, application of basic biological concepts, and rational thought processes to support my position. So where we disagree is that I cannot find any evidence, of a scientific nature, that would support the claim of a break in the species' distribution. >>

You relied on inference because at the time you made your conclusion, there was no known collection records that could invalidate Rodriguez-Robles' theory of a break in the distribution of this species. Sometimes inference can be correct, but sometimes it may lead to the wrong conclusion.

>>I do not consider that randomly collected vouchers which by chance, had tissue taken and tested, as representing evidence for claiming a break in any species' distribution let alone that of the Rubber Boa.>>

Certainly, the absence of evidence, in most cases, should not be considered evidence of absence. That appears to be Rodriguez-Robles' mistake.

>>But I do understand how the authors were misled by their data in reaching such a conclusion. After all, their data do show the two subclades as having quite separate geographical distributions. >>

Absolutely. That is why I said that their conclusion is scientific, since it is based on the available evidence.

>>How easy it is to criticize as it is free. But then one has to be aware of 'he who dare casts the first stone be without sin---or similar sage advice. Hah.
>>
>>When I contacted Glenn about this point, I wasn't as 'forceful' as I am here as I thought that surely that glitch would be caught during the peer review process.>>

The peer review process is often overrated. Sometimes a paper based on bad data is approved, but sometimes a good paper is rejected because of personal animosity, fashion or partisan politics.

>>But such glitches aside, I agree that the results obtained by Javier's mtDNA research was an important step forward. In addition, at least for the time being, it also put to rest the notion of the Great Basis subspecies (C. b. utahensis).>>

>>I certainly agree with your last paragraph. I do not know exactly what Gould mentioned but I agree in the broad context. Too many individuals in and out of science accept scientific findings as if etched in stone. >>
>>Also, I have read and reviewed a number of scientific publications in which the experimental design and results obtained were quite nice. The mistakes I see being made are in the interpretation of those results. This is particularly true in the field of Conservation Biology where individuals have pre-existing biases that interfere with analyzing their results in an impartial and objective manner.
>>
>>Richard F. Hoyer

Looks like we agree on many things. BTW, you mentioned in the past that you believe the ancestor of the rubber boa should be found in the mountains of northwestern Mexico. The most likely ancestor of the rubber boa, however, is Exiliboa, and it is found a lot farther south than you had suggested. It turns out there was a lot of climatic changes and other geologic disturbances in Northwestern Mexico over the past several million years. It is likely that Exiliboa once extended much farther north than its present day range but it has retreated to its present day location. Often we forget that the distribution of organisms do change over time, and that the ranges of some animals have expanded and contracted multiple times during their evolutionary history. That is why L. z. multifasciata is much more closely related to L. z. agalma than multifasciata is to either L. z. parvirubra or L. z. pulchra, even though multifasciata is much closer geographically to pulchra and parvirubra than to agalma. Similarly, the closest relatives of the Kern County boas are those from Tulare County, instead of those from southern California.

CK,
Actually, the two specimens I cited from the MVZ collection are old. The specimen from Silver Lake in Lassen county is MVZ 18187 collected in 1935 and the Morgan Springs (Ranch) specimen from Tehama County is MVZ #14946 collected before I was born in 1932.

Richard F. Hoyer

>>CK,
>>Actually, the two specimens I cited from the MVZ collection are old. The specimen from Silver Lake in Lassen county is MVZ 18187 collected in 1935 and the Morgan Springs (Ranch) specimen from Tehama County is MVZ #14946 collected before I was born in 1932.
>>
>>Richard F. Hoyer

Then obviously they overlooked the evidence. I don't know when the MVZ computerized its records, but back in the mid 1990's computer literacy was not very widespread, and the Internet was not available to the general public until the mid-1990's. Back then most people were surfing the Net with 14.4K modems.

In Lord Kelvin's case radioactivity was not discovered until half a century or so after he made his famous calculation of the age of the earth. Of course he also ignored all the geological processes which must have taken much longer than his calculations would suggest.