cactus taxonomy, systematics only
cactus taxonomy, systematics only
I want to continue conversations about various studies related to systematics/cladistics/phylogeny of the cacti, and figured it might be good to just have a single thread, rather than constantly introducing the topic into other threads. I think taxonomy is tiresome and even upsetting to some and it's probably better to keep it separate, kind of like the agreement my family has to not talk about politics or religion at the dinner table.
Anyway, this article by Butterworth is a very interesting read:
http://www.usask.ca/biology/cota-sanche ... t_2002.pdf
and here's a fun lab assignment from Berkeley:
http://74.125.155.132/search?q=cache:uW ... =firefox-a
I still have to get to Crozier's article, thanks again for sending that Ian.
I do have a very simple, probably fairly stupid question: what goes into deciding which segment of DNA to compare for these studies?
peterb
here's a couple other articles available online that I included in other posts:
Relationships Between Sclerocactus and Toumeya based on Chloroplast trnL-trnF Sequences, Haseltonia No. 7, 2000, Porter, Kinney and Heil:
http://www.rsabg.org/research/Porter/pd ... al2000.pdf
A link on the page to the full article, Phylogenetic relationships in the cactus family (Cactaceae) based on evidence from trnK/ matK and trnL-trnF sequences, American Journal of Botany, 2002;89:312-326, Nyffeler
http://www.amjbot.org/cgi/content/abstract/89/2/312
Anyway, this article by Butterworth is a very interesting read:
http://www.usask.ca/biology/cota-sanche ... t_2002.pdf
and here's a fun lab assignment from Berkeley:
http://74.125.155.132/search?q=cache:uW ... =firefox-a
I still have to get to Crozier's article, thanks again for sending that Ian.
I do have a very simple, probably fairly stupid question: what goes into deciding which segment of DNA to compare for these studies?
peterb
here's a couple other articles available online that I included in other posts:
Relationships Between Sclerocactus and Toumeya based on Chloroplast trnL-trnF Sequences, Haseltonia No. 7, 2000, Porter, Kinney and Heil:
http://www.rsabg.org/research/Porter/pd ... al2000.pdf
A link on the page to the full article, Phylogenetic relationships in the cactus family (Cactaceae) based on evidence from trnK/ matK and trnL-trnF sequences, American Journal of Botany, 2002;89:312-326, Nyffeler
http://www.amjbot.org/cgi/content/abstract/89/2/312
Zone 9
It's a fairly steep learning curve for me!
Anyone have an idea of why certain segments of DNA are sampled rather than others?
Any pointers to hard science, non-speculative articles that raise objections to or even mild reservations about the assumptions, methods and/or conclusions of DNA-based cladistics/systematics are also appreciated.
I'm also particularly interested in a better explanation of parsimony as an underlying assumption.
peterb
Anyone have an idea of why certain segments of DNA are sampled rather than others?
Any pointers to hard science, non-speculative articles that raise objections to or even mild reservations about the assumptions, methods and/or conclusions of DNA-based cladistics/systematics are also appreciated.
I'm also particularly interested in a better explanation of parsimony as an underlying assumption.
peterb
Zone 9
Segments of DNA are chosen by the "suck it and see" methodology Not every one works for every plant. Not every one is even present in every plant. Segments need to be found that change quickly enough to present changes between species, or other taxa, but not so quickly that you just have chaos. Then you need to find some more because what you see from a single fragment of DNA may just be a coincidence. Fragments of DNA are often chosen from non-coding regions because these are thought to mutate more randomly and smoothly than coding regions. Chloroplast DNA has also been a common choice, partly because it is easy to work with and partly because it is thought to change more smoothly over time. Chloroplast DNA is almost entirely inherited in the maternal line which has obvious implications for phylogeny.
I'm not sure you're going to find a scientific paper discussing the downsides of DNA sequencing. On the one hand they're fairly widely known so not really discussed. On the other hand most of the arguments against it are philosophical rather than technical, hence many of the positions taken against the use of DNA in taxonomy are not very helpful. Many of the limitations, such as inheritance of some genes solely down the maternal or paternal line, can be used as tools in themselves.
Parsimony is very simple. If only you knew what parsimony was, that is just the best pun ever! Parsimony is the term used to describe one of the statistical techniques applied to sorting all the gene fragments into relationships. It just means that the simplest possible explanation is assumed for the relationships. For example, where there is a single difference in the DNA between two samples, it is assumed that these are closely related, not that they have undergone multiple changes that happen to bring them back close together. Either is possible, but one is simpler and hence considered more likely. You may be able to visualise situations where this doesn't hold, if sufficient changes have occurred between two organisms that it becomes statistically likely for base pairs to match up again (remembering that there are only four base pairs).
I'm not sure you're going to find a scientific paper discussing the downsides of DNA sequencing. On the one hand they're fairly widely known so not really discussed. On the other hand most of the arguments against it are philosophical rather than technical, hence many of the positions taken against the use of DNA in taxonomy are not very helpful. Many of the limitations, such as inheritance of some genes solely down the maternal or paternal line, can be used as tools in themselves.
Parsimony is very simple. If only you knew what parsimony was, that is just the best pun ever! Parsimony is the term used to describe one of the statistical techniques applied to sorting all the gene fragments into relationships. It just means that the simplest possible explanation is assumed for the relationships. For example, where there is a single difference in the DNA between two samples, it is assumed that these are closely related, not that they have undergone multiple changes that happen to bring them back close together. Either is possible, but one is simpler and hence considered more likely. You may be able to visualise situations where this doesn't hold, if sufficient changes have occurred between two organisms that it becomes statistically likely for base pairs to match up again (remembering that there are only four base pairs).
--ian
ha, I did get your parsimony pun. Science jokes are rad.
I have found this a useful basic reference on cladistics:
http://www.palaeos.com/Systematics/Clad ... istics.htm
I do see how the fewest possible changes is an attractive idea. I'm not entirely convinced it's a necessary fact. Perhaps there are quite elaborate changes indeed, far more than what would constitute the fewest "needed" to get organisms C and B from common ancestor A. It seems in this messy business of life Occam's Razor might be too eagerly applied. I'm sure a consensus tree of greatest parsimony makes it easier to draw conclusions, but it also turns out there are several different possible consensus trees, so it seems we're still at the point of making our own selections based on some sort of subjective consideration. ?
Thanks for the explanation of why certain segments of DNA are selected and not others.
Not looking for a critique of DNA sequencing, but of the underlying assumptions behind either phylogenetics or cladistics. Not the philosophical arguments, but the scientific. Actually, the reference page I linked to above has a good paragraph citing some reservations.
peterb
I have found this a useful basic reference on cladistics:
http://www.palaeos.com/Systematics/Clad ... istics.htm
I do see how the fewest possible changes is an attractive idea. I'm not entirely convinced it's a necessary fact. Perhaps there are quite elaborate changes indeed, far more than what would constitute the fewest "needed" to get organisms C and B from common ancestor A. It seems in this messy business of life Occam's Razor might be too eagerly applied. I'm sure a consensus tree of greatest parsimony makes it easier to draw conclusions, but it also turns out there are several different possible consensus trees, so it seems we're still at the point of making our own selections based on some sort of subjective consideration. ?
Thanks for the explanation of why certain segments of DNA are selected and not others.
Not looking for a critique of DNA sequencing, but of the underlying assumptions behind either phylogenetics or cladistics. Not the philosophical arguments, but the scientific. Actually, the reference page I linked to above has a good paragraph citing some reservations.
peterb
Zone 9
I'm not sure how much mathematics you want to wade through, but you might read this link. It explains how maximum parsimony is used to infer a certain tree. It also closes with some pros and cons, although they are related strictly to this single method. It has a bit about equally parsimonious trees and consensus trees which should answer one of your questions.
--ian
Reading and trying to digest the Butterworth article I linked to yesterday. This caught my eye:
"Mammillaria glassii forms
a sister-group to all of the remaining members of the
Cacteae in the neighbor-joining tree. This incongruence
is caused by differences in sequence length of Mammillaria
glassii of only 650bp due to a large deletion
spanning the region with most informative characters."
Anyway, there is so much information packed into the cladogram and neighbor joining tree diagram that it's interesting to just sort of stare at them.
I'm more interested in the weaker clades. It seems there could be promising results in further investigating why they are weakly defined in this study.
I'm also interested in how close much of the molecular data is to gross morphological observation. The real power of this sort of DNA research will ultimately lie in completely unexpected results, probably, not results that confirm existing ideas of phylogeny.
It's interesting too how the Ancistracanthae keep coming up in these studies of Mammillaria.
peterb
"Mammillaria glassii forms
a sister-group to all of the remaining members of the
Cacteae in the neighbor-joining tree. This incongruence
is caused by differences in sequence length of Mammillaria
glassii of only 650bp due to a large deletion
spanning the region with most informative characters."
Anyway, there is so much information packed into the cladogram and neighbor joining tree diagram that it's interesting to just sort of stare at them.
I'm more interested in the weaker clades. It seems there could be promising results in further investigating why they are weakly defined in this study.
I'm also interested in how close much of the molecular data is to gross morphological observation. The real power of this sort of DNA research will ultimately lie in completely unexpected results, probably, not results that confirm existing ideas of phylogeny.
It's interesting too how the Ancistracanthae keep coming up in these studies of Mammillaria.
peterb
Zone 9
Great stuff! Still reading!
Two questions to begin with...
First, according to Butterworth's cladogram, is Lophophora williamsii more related to Obegonia denegrii than to Lophophora diffusa, or am I reading this wrong?
Second, according to the same work, Echinocactus grusonii is more related to species of genus Ferocactus, than to members of the genus Echinocactus.
Now, I bear in mind what Ian said about this expected to happen because:
"Cladograms show with 100% certainty that Echinocactus grusonii is a Ferocactus, but we know with almost 100% certainty that this is simply because it results from hybridisation with a Ferocactus as the maternal parent", to preserve Ian's own words. (by the way this means that the E. grusonii sold worldwide is a Ferocactus hybrid?).
The problem is that grusonii isn't the only ...problem here.
Echinocactus texensis is called Homalocephala texensis instead ,by the author, and for a good reason!
It is related to the other members of the genus Echinocactus but to the same degree as Astrophytum myriostigma and Astrophytum capricorne.
The numbers representing the relationship between E. ingens and E. horizonthalonius are 76/1. They aren't very strong affiliated either, are they?
In Nyffeler's work we see E. platyacanthus related to Astrophytum myriostigma.
So, I wonder, is Echinocactus, numbering now only six species, a well established and solid genus or a leftover from the past, a summon of species that didn't fit elsewhere?
If the latter is true, is there any chance that some of its members will be under the same genus as Astrophytum myriostigma and capricorne and I don't know which else, in the future?
Two questions to begin with...
First, according to Butterworth's cladogram, is Lophophora williamsii more related to Obegonia denegrii than to Lophophora diffusa, or am I reading this wrong?
Second, according to the same work, Echinocactus grusonii is more related to species of genus Ferocactus, than to members of the genus Echinocactus.
Now, I bear in mind what Ian said about this expected to happen because:
"Cladograms show with 100% certainty that Echinocactus grusonii is a Ferocactus, but we know with almost 100% certainty that this is simply because it results from hybridisation with a Ferocactus as the maternal parent", to preserve Ian's own words. (by the way this means that the E. grusonii sold worldwide is a Ferocactus hybrid?).
The problem is that grusonii isn't the only ...problem here.
Echinocactus texensis is called Homalocephala texensis instead ,by the author, and for a good reason!
It is related to the other members of the genus Echinocactus but to the same degree as Astrophytum myriostigma and Astrophytum capricorne.
The numbers representing the relationship between E. ingens and E. horizonthalonius are 76/1. They aren't very strong affiliated either, are they?
In Nyffeler's work we see E. platyacanthus related to Astrophytum myriostigma.
So, I wonder, is Echinocactus, numbering now only six species, a well established and solid genus or a leftover from the past, a summon of species that didn't fit elsewhere?
If the latter is true, is there any chance that some of its members will be under the same genus as Astrophytum myriostigma and capricorne and I don't know which else, in the future?
Yes, Echinocactus is a "phenetic" genus, not a phylogenetic one, it seems. I expect it has only been the rules of botanical nomenclature that have prevented its complete dismantling. There is a real irony in this, as Echinocactus used to be a megagenus, with dozens and dozens of species.
peterb
peterb
Zone 9
Yes, I understand that, Süleyman. In fact, I'm 100% in favor of all the natural (not agricultural) hybrids, like Ferocactus californicus for example, gaining the status of species.süleyman wrote:grusonii is a stabilised ancient hybrid, so it is accepted a true species in the nomenclature laws.
So, grusonii is a true species, no doubt about that, but under which genus?
As a "splitter" my self I don't have any problem of keeping or creating more genera.
Nevertheless, "combiners" have the upper hand nowdays, and there were opinions( Taylor) that even the genus Stenocactus should lose its status as genus and be included to the Ferocacti (the DNA tests above suggest otherwise on this).
Now, should't they start with grusonii first before going to occlude something else?
We have seen sad things like Coryphantha odorata, Rebutia neocumingii, Sclerocactus uncinatus, Thelocactus setispinus, but the relic-genus Echinocactus wasn't touched!
Is Echinocactus grusonii a name written on a "libro d' oro" and therefor cannot be changed?
We have witnessed all the above miss-classifications, why would Ferocactus grusonii sound wrong?
Quite possibly. Other studies have also shown that Obregonia is simply a Lophophora with elongated tubercles. Note also the position of Acharagma. The exact relationships are not statistically certain in Butterworth's paper, but other work has shown that Acharagma is certainly closer to Lophophora than to anything else.according to Butterworth's cladogram, is Lophophora williamsii more related to Obegonia denegrii than to Lophophora diffusa, or am I reading this wrong?
Again the probabilities are a little low for this relationships. Other studies place it with the other Echinocactus and separate from Astrophytum. They're all close though. Crozier, after splitting Coryphantha and Mammillaria into a thousand pieces, decides that Homocephala should stay lumped in Echinocactus. The traditional differences include the flattened growth habit, many sharp ribs (but see E. polycephalus with sharp ribs), pubescent spines (but again see E. polycephalus), and bright red fleshy fruit protruding from the wool.Echinocactus texensis is called Homalocephala texensis instead ,by the author, and for a good reason!
E. grusonii is only related to Ferocactus on the maternal line. It is left in the genus Echinocactus because that's what the flowers, fruit, and seed look like and they are considered more taxonomically significant than shiny skin. A genus Ferocactus including E. grusonii would be a disaster because E. grusonii has none of the characters that set Ferocactus apart from the other Echinocactus species. You want Ferocactus grusonii, feel free to write it up and have it published, but I doubt it will catch on. I think you're cherry-picking data (ie. maternal line chloroplast DNA) to support a pet theory when far more data contradicts it.We have witnessed all the above miss-classifications, why would Ferocactus grusonii sound wrong?
These numbers don't mean that these species are not closely related, merely that this study does not strongly prove that relationship. Possibly they aren't close, possibly they are. Others studies show that they are certainlly related but not as close as, for example, E. parryi and E. polycephalus.The numbers representing the relationship between E. ingens and E. horizonthalonius are 76/1. They aren't very strong affiliated either, are they?
Where is this paper? I haven't been able to find it.In Nyffeler's work we see E. platyacanthus related to Astrophytum myriostigma.
It is certainly a diverse genus, very much a relic of the original very large genus. I don't think anyone would have created it as it stands, but you can't disband it without something else to put in it's place. Is it so hard to believe that Astrophytum capricorne would be an Echinocactus? Compare the flowers and fruits some time, they are remarkably similar, certainly more variable within each genus than they are different from eachother. The distinctive skin flecks of Astrophytum could easily have been ignored if the genus was being created today with all that we now know.is Echinocactus, numbering now only six species, a well established and solid genus or a leftover from the past
Stenocactus has long been considered closely related to Ferocactus, just as Ferocactus has long been considered a slightly dodgy genus, probably not monophyletic. The DNA supports both opinions, although most data suggests that Stenocactus can stand along as a monophyletic group. Look at the position of Ferocactus histrix in Butterworth's cladograms though, the relationships between these species are not completely clear yet. Again a hit-you-in-the-face feature like many very narrow ribs has created a genus when other features like flowers and fruit suggest they are very similar. Look at S. coptonogonus and its pretty hard to separate from a Ferocactus.even the genus Stenocactus should lose its status as genus and be included to the Ferocacti (the DNA tests above suggest otherwise
--ian
You can say that again!iann wrote: Look at S. coptonogonus and its pretty hard to separate from a Ferocactus.
I would go even further...
Put side by side a Ferocactus macrodiscus, a Stenocereus coptonogonus and an Echinocactus texensis and try to convince anyone, not so familiar with cactus taxonomy, that these three cacti not only represents different species but even more they belong to different genera!