![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
which_chickBear with me here, because I know that problems of taxonomy are not the sort of thing you probably want to read about. But this is kind of interesting and also it's the sort of thing I want to talk about. Again.
Back in February, I posted about how the "discussions" about what genera should exist and which cactus go in which genus was not all straightened out yet. In particular, I mentioned how Mammillaria was a bit of a mess, yes? It was one of the examples I offered up.
In fact, and I'm quoting here because you surely did not go back and read the linked post, absolutely nobody else is as fond of the ramble and murmur of my own prose as I am, I put it thusly: Lots of things that were not Mammillarias originally got lumped into Mammillaria and so here we are with that. Mammillaria is a mess, there are like maybe 500 species or something. Even in a perfect world, it would be... a lot. Lots of shit for the Splitters and Lumpers to play with in the Mammillaria world.
Come to find out here recently (I was doing some reading about what other Mammillarias I might like to try growing as, on average, they stay fairly small and fit under lights and have interesting spines and stuff) that in 2021, there was a fairly substantial upheaval in Mammilliara land. Of course it's paywalled. *sigh* That's how science works.
I'd be more enthused about paying to read it if, y'know, the guys who WROTE THE PAPER got the money, but I am almost 100% sure that the guys who wrote the paper do not get the money. I think the effing coyote (it's hosted at wiley.com) gets the money. To be fair, they have to do something to thin the utter hordes of people thrashing their servers in an attempt to read Molecular phylogeny of the Mammilloid clade (Cactaceae) resolves the monophyly of Mammillaria. I am certain that this is an area of extreme interest for everyone, right? Like, you were sipping your coffee and looking out the window thoughtfully this morning while wondering if they'd made any headway straightening out the Mammillaria problem or not. Sure, Jan. Sure.
But in the interests of science (and/or graphs and charts that I can use to inform my... selection of interesting cactus to try growing), I did indeed throw eighteen dollars at being able to read this article. That's like three lattes, so fuck it. I can afford it. Anyway, now you're stuck. For any academics out there (lol) who want to know for realsies what paper I am citing, in this here discussion, it's this one.
Fortunately, it's chock-full of delightfully amusing pull quotes (which I have to highlight and then retype by hand because paying eighteen dollars DOES NOT purchase the ability to copy text) like the following: "Currently accepted species numbers [for cactus as a whole] are between 800 and 1860, a range that shows enduring disagreement among authorities regarding taxon boundaries." That shit's droll as hell for a research paper. Two thumbs up. And we're on, like, the intro paragraphs. Anyway, the paper says that at the time of writing, there were something like 164-200 "accepted" species of Mammillaria. And that's not "like maybe 500 species or something" (because I am an idiot) but it's still a lot.
Assorted people have tried to sort out the Mammillaria problem since like the 1920's by splitting and lumping and there's a reasonable summary of those efforts in a non-paywalled article here that basically sums it up as "the phylogeny presented in this paper suggests that as currently circumscribed, the genus Mammillaria is likely polyphyletic on a number of levels." "Polyphyletic" means "derived from more than one common evolutionary ancestor or ancestral group and therefore not suitable for placing in the same taxon." So... in non-science English, they mean "The Mammillaria genus needs to be divided further because not all the stuff currently shoehorned into Mammillaria has the same ancestors."
So anyway, back to the paywalled paper. To do the study, they sampled cacti, both within the Mammillaria-like group and without, and focusing particularly on two broadly-different-looking sorts of Mammillaria, to wit: single-growing cacti with straight spines and clumpy-growing cacti with hooked spines. (They also made a concerted effort to get a right proper selection of the stuff growing on Baja California, probably with an eye toward investigating/discussing that segment of the cactus population in greater detail in another paper. This shit costs money, so best to work smarter, not harder.)
Why? If you think (as the authors did) that there are multiple different ancestors going on in the Mammillaria group and that it should be split up into smaller groups, and if you're gonna play Expensive Genetic Analysis to look into that, it would behoove you to make sure you have good representation of the different-looking cactus involved because (d'oh) different-looking cactus probably have different ancestors.
Also, morphology (physical features, "looking different", like hooked spines or shape of flowers or whatever) has been used to classify stuff into taxons forever, especially in plants. You can learn more (in kind of swear-y YouTubes featuring a dog and a guy looking at real plants in the real world) from youtuber Crime Pays But Botany Doesn't. He's fun and goes into stuff about plant characteristics and how that defines plant groups and shit. And botany has used morphology since forever because (a) it mostly works and (b) it's cheap and (c) you can do it in the field and (d) dna analysis is expensive, time consuming, and didn't exist back when people started looking at plants for science.
So, the paper is gonna try to do five things. They're listed and numbered in the original. The five things they're gonna try to do are as follows: 1. Investigate mammillaria using bigger hunks of chloroplast gene than before. 2. Clarify relationships within Mammillaria. 3. Are Coryphantha and Escobaria "real" genuses or not? 4. Is Cochemiea a "real" genus or not? 5. Is looking at morphology *actually* useful in trying to tease out relationships in this group of cactus? (While there are pluses to looking at morphology, see previous paragraph, there is also the problem that different cacti, even completely unrelated cacti, can sometimes evolve the same-looking morphology while not being related. Like... both birds and bats have wings but they're not the same group at all? Like that. Picking out "the right morphology" is not always easy. Sure, a thing is different but is it different enough to make a new genus? Does it define a group? And sometimes, cacti can lose "group morphology traits" used to identify a group but yet still be related to that group. So,morphology isn't 100% reliable.)
A digression: Why chloroplasts? Chloroplasts are used for studying plant genetics because they have a small, highly conserved genome.
Small: The entire gene sequence of a saguaro cactus is like 1.6 billion base pairs. The entire gene sequence of a land plant chloroplast is about 160,000 base pairs. So, yeah. SMALL. Like four zeroes less. That's a lot less.
Highly Conserved: This means a genome where large portions of DNA sequence remain very similar across different species, indicating that these sequences are essential for basic life functions and have changed very little over long periods of evolution. If it was changing all the time, it'd be a lot harder to see what was going on there. But if it's pretty consistent, the very slow changes in the chloroplast genome can be used to track speciation across different kinds of cacti because the speciation thing is happening FASTER than the drift or change in the chloroplast genome.
Fine, they're small and highly conserved. What is a chloroplast, anyway? They're small ... cell sub-assemblies that live outside the cell nucleus and are used to make energy from sunlight. Because they're "outside the cell nucleus" they do not do the sexual reproduction thing where they get genes from both parents. That's for the nucleus stuff. Instead, like mitochondria, chloroplasts generally come from just one parent. So, that helps simplify things, too. Less variability from generation to generation. Further, there's a lot of interest in using them for study, not just with cactus, but with all green plants. So, there are (relatively) a lot of tools and methodologies and shit right there, ready to use for your research project. So, people trying to work out plant dna relationships look at the chloroplast genome.
Right. Back to our regular programming.
So, they took like 80-odd samples from assorted cacti, most of them from Mammillaria and some from other non-Mammillaria groups that were deemed "closely related", including stuff like Coryphantha and Escobaria (see question 3, previous paragraph) and some not-at-all related ones too. Where'd they get their samples? Among other sources, Mesa Garden (large cactus and mesemb nursery, I've bought from them) and Desert Botanical Garden Living Collection. There's a list and stuff, but that's details we probably don't need.
They extracted the DNA and selected the hunks they wanted to study and used a variety of Science Math Stuff to sort the DNA hunks and subsequently generate a ... family tree of the cacti studied. I expect that this involved computers, a lot, and they did a couple of different methods of Science Math Stuff to generate the family tree thing to make sure that the resultant tree was reasonable and not... off the reservation. Also, to address question 5, they picked out some morphology things, like hooked spines vs. straight spines, flower diameter (large or small, basically), etc. to cross-reference with their DNA-based family tree after-the-fact to see if the morphology stuff was actually USEFUL for taxonomy in these groups of studied cacti. The morphology things they picked were based on traditionally-used cactus characteristics from people who described, for example, Coryphantha and stuff. There were probably some Science Math Things involved in correlating the morphology stuff with the DNA stuff to see if it was useful or not, but it'd still be an order of magnitude less complicated than the DNA sorting and family tree generation Science Math Things.
So, how'd it go? How'd they do on the five things they said they were gonna do?
1. Investigate Mammillaria using bigger hunks of chloroplast gene than before. They did this part. It's not so much a question to be answered as a statement of intent.
2. Clarify relationships within Mammillaria. They did this, there's a whole family tree graphic in the paper which you cannot see but I assure you it is there. I paid eighteen dollars to see the very nice family tree graphic. It's not 100% of the Mammillaria-related things, but it's a pretty good hunk of 'em.
3. Are Coryphantha and Escobaria "real" genuses or not? Yes, but maybe/likely they're only one thing, like, Coryphantha and Escobaria aren't very different and so it should be one thing called "Coryphantha" because... it was there first? (There are rules about which genus name takes precedence when stuff like this happens and I'm pretty sure the rule is seniority.) Anyway, maybe also M. sphacelata is in this newly arranged Coryphantha genus as Coryphantha sphacelata but Coryphantha/Cumarinia odorata IS NOT part of the Coryphantha group and is definitely its own thing, Cumarinia odorata.
4. Is Cochemiea a real genus or not? Wow, sure is, bucko, and it's way bigger than you think. If scientific journal articles were allowed to include the phrase booyah, it surely would have been included here. Also, a bunch of alleged Mammillaria-whatevers just turned out to all be Cochemiea-whatevers. There's a list and presumably cactus enthusiasts and growers and shit now all have to cross-reference and relabel their plants and seed catalogs so that people looking for Mammillaria grahamii will be able to find it now that it's been refiled under Cochemiea grahamii. Bet that's going to be fun for people, especially the nerdy sort of people who have scientific labels on their cactus plants. (Innocent look.)
5. Is looking at morphology *actually* useful in trying to tease out clades/relationships in this group of cactus? Yes, but you have to be looking at the RIGHT MORPHOLOGY. Here, they ran the DNA stuff and then looked at the DNA results to make the cactus groups and then they worked backward toward "what observable features, if any, out of the ten we identified up front do these DNA-defined groups share?". Once you already know what the "group" is, it's not so hard to look at them for "synapomorphy" (a characteristic present in an ancestral species and shared exclusively (in more or less modified form) by its evolutionary descendants) in that group. That's basically cheating. Also, it's patently obvious that looking at morphology is not 100% easy and reliable for TAXONOMIZING MAMMILLARIA-LIKE CACTUS because if it was 100% easy and reliable, we would not have ever had the Mammillaria problem in the first place. Anyhoo, the researchers started with 10 morphologies to look at and in the end wound up with 3 useful morphologies, one for each... group. Hooked spines with Cochemeia, grooved tubercules with Coryphantha (except one), lactiferous ducts with Mammillaria (mostly). There are, again, very nice graphics illustrating this. So, I give this one a "Yes, but..." answer.
Did we get eighteen dollars worth of entertainment out of this paper? Yes, yes we did. Fun times.
Back in February, I posted about how the "discussions" about what genera should exist and which cactus go in which genus was not all straightened out yet. In particular, I mentioned how Mammillaria was a bit of a mess, yes? It was one of the examples I offered up.
In fact, and I'm quoting here because you surely did not go back and read the linked post, absolutely nobody else is as fond of the ramble and murmur of my own prose as I am, I put it thusly: Lots of things that were not Mammillarias originally got lumped into Mammillaria and so here we are with that. Mammillaria is a mess, there are like maybe 500 species or something. Even in a perfect world, it would be... a lot. Lots of shit for the Splitters and Lumpers to play with in the Mammillaria world.
Come to find out here recently (I was doing some reading about what other Mammillarias I might like to try growing as, on average, they stay fairly small and fit under lights and have interesting spines and stuff) that in 2021, there was a fairly substantial upheaval in Mammilliara land. Of course it's paywalled. *sigh* That's how science works.
I'd be more enthused about paying to read it if, y'know, the guys who WROTE THE PAPER got the money, but I am almost 100% sure that the guys who wrote the paper do not get the money. I think the effing coyote (it's hosted at wiley.com) gets the money. To be fair, they have to do something to thin the utter hordes of people thrashing their servers in an attempt to read Molecular phylogeny of the Mammilloid clade (Cactaceae) resolves the monophyly of Mammillaria. I am certain that this is an area of extreme interest for everyone, right? Like, you were sipping your coffee and looking out the window thoughtfully this morning while wondering if they'd made any headway straightening out the Mammillaria problem or not. Sure, Jan. Sure.
But in the interests of science (and/or graphs and charts that I can use to inform my... selection of interesting cactus to try growing), I did indeed throw eighteen dollars at being able to read this article. That's like three lattes, so fuck it. I can afford it. Anyway, now you're stuck. For any academics out there (lol) who want to know for realsies what paper I am citing, in this here discussion, it's this one.
Fortunately, it's chock-full of delightfully amusing pull quotes (which I have to highlight and then retype by hand because paying eighteen dollars DOES NOT purchase the ability to copy text) like the following: "Currently accepted species numbers [for cactus as a whole] are between 800 and 1860, a range that shows enduring disagreement among authorities regarding taxon boundaries." That shit's droll as hell for a research paper. Two thumbs up. And we're on, like, the intro paragraphs. Anyway, the paper says that at the time of writing, there were something like 164-200 "accepted" species of Mammillaria. And that's not "like maybe 500 species or something" (because I am an idiot) but it's still a lot.
Assorted people have tried to sort out the Mammillaria problem since like the 1920's by splitting and lumping and there's a reasonable summary of those efforts in a non-paywalled article here that basically sums it up as "the phylogeny presented in this paper suggests that as currently circumscribed, the genus Mammillaria is likely polyphyletic on a number of levels." "Polyphyletic" means "derived from more than one common evolutionary ancestor or ancestral group and therefore not suitable for placing in the same taxon." So... in non-science English, they mean "The Mammillaria genus needs to be divided further because not all the stuff currently shoehorned into Mammillaria has the same ancestors."
So anyway, back to the paywalled paper. To do the study, they sampled cacti, both within the Mammillaria-like group and without, and focusing particularly on two broadly-different-looking sorts of Mammillaria, to wit: single-growing cacti with straight spines and clumpy-growing cacti with hooked spines. (They also made a concerted effort to get a right proper selection of the stuff growing on Baja California, probably with an eye toward investigating/discussing that segment of the cactus population in greater detail in another paper. This shit costs money, so best to work smarter, not harder.)
Why? If you think (as the authors did) that there are multiple different ancestors going on in the Mammillaria group and that it should be split up into smaller groups, and if you're gonna play Expensive Genetic Analysis to look into that, it would behoove you to make sure you have good representation of the different-looking cactus involved because (d'oh) different-looking cactus probably have different ancestors.
Also, morphology (physical features, "looking different", like hooked spines or shape of flowers or whatever) has been used to classify stuff into taxons forever, especially in plants. You can learn more (in kind of swear-y YouTubes featuring a dog and a guy looking at real plants in the real world) from youtuber Crime Pays But Botany Doesn't. He's fun and goes into stuff about plant characteristics and how that defines plant groups and shit. And botany has used morphology since forever because (a) it mostly works and (b) it's cheap and (c) you can do it in the field and (d) dna analysis is expensive, time consuming, and didn't exist back when people started looking at plants for science.
So, the paper is gonna try to do five things. They're listed and numbered in the original. The five things they're gonna try to do are as follows: 1. Investigate mammillaria using bigger hunks of chloroplast gene than before. 2. Clarify relationships within Mammillaria. 3. Are Coryphantha and Escobaria "real" genuses or not? 4. Is Cochemiea a "real" genus or not? 5. Is looking at morphology *actually* useful in trying to tease out relationships in this group of cactus? (While there are pluses to looking at morphology, see previous paragraph, there is also the problem that different cacti, even completely unrelated cacti, can sometimes evolve the same-looking morphology while not being related. Like... both birds and bats have wings but they're not the same group at all? Like that. Picking out "the right morphology" is not always easy. Sure, a thing is different but is it different enough to make a new genus? Does it define a group? And sometimes, cacti can lose "group morphology traits" used to identify a group but yet still be related to that group. So,morphology isn't 100% reliable.)
A digression: Why chloroplasts? Chloroplasts are used for studying plant genetics because they have a small, highly conserved genome.
Small: The entire gene sequence of a saguaro cactus is like 1.6 billion base pairs. The entire gene sequence of a land plant chloroplast is about 160,000 base pairs. So, yeah. SMALL. Like four zeroes less. That's a lot less.
Highly Conserved: This means a genome where large portions of DNA sequence remain very similar across different species, indicating that these sequences are essential for basic life functions and have changed very little over long periods of evolution. If it was changing all the time, it'd be a lot harder to see what was going on there. But if it's pretty consistent, the very slow changes in the chloroplast genome can be used to track speciation across different kinds of cacti because the speciation thing is happening FASTER than the drift or change in the chloroplast genome.
Fine, they're small and highly conserved. What is a chloroplast, anyway? They're small ... cell sub-assemblies that live outside the cell nucleus and are used to make energy from sunlight. Because they're "outside the cell nucleus" they do not do the sexual reproduction thing where they get genes from both parents. That's for the nucleus stuff. Instead, like mitochondria, chloroplasts generally come from just one parent. So, that helps simplify things, too. Less variability from generation to generation. Further, there's a lot of interest in using them for study, not just with cactus, but with all green plants. So, there are (relatively) a lot of tools and methodologies and shit right there, ready to use for your research project. So, people trying to work out plant dna relationships look at the chloroplast genome.
Right. Back to our regular programming.
So, they took like 80-odd samples from assorted cacti, most of them from Mammillaria and some from other non-Mammillaria groups that were deemed "closely related", including stuff like Coryphantha and Escobaria (see question 3, previous paragraph) and some not-at-all related ones too. Where'd they get their samples? Among other sources, Mesa Garden (large cactus and mesemb nursery, I've bought from them) and Desert Botanical Garden Living Collection. There's a list and stuff, but that's details we probably don't need.
They extracted the DNA and selected the hunks they wanted to study and used a variety of Science Math Stuff to sort the DNA hunks and subsequently generate a ... family tree of the cacti studied. I expect that this involved computers, a lot, and they did a couple of different methods of Science Math Stuff to generate the family tree thing to make sure that the resultant tree was reasonable and not... off the reservation. Also, to address question 5, they picked out some morphology things, like hooked spines vs. straight spines, flower diameter (large or small, basically), etc. to cross-reference with their DNA-based family tree after-the-fact to see if the morphology stuff was actually USEFUL for taxonomy in these groups of studied cacti. The morphology things they picked were based on traditionally-used cactus characteristics from people who described, for example, Coryphantha and stuff. There were probably some Science Math Things involved in correlating the morphology stuff with the DNA stuff to see if it was useful or not, but it'd still be an order of magnitude less complicated than the DNA sorting and family tree generation Science Math Things.
So, how'd it go? How'd they do on the five things they said they were gonna do?
1. Investigate Mammillaria using bigger hunks of chloroplast gene than before. They did this part. It's not so much a question to be answered as a statement of intent.
2. Clarify relationships within Mammillaria. They did this, there's a whole family tree graphic in the paper which you cannot see but I assure you it is there. I paid eighteen dollars to see the very nice family tree graphic. It's not 100% of the Mammillaria-related things, but it's a pretty good hunk of 'em.
3. Are Coryphantha and Escobaria "real" genuses or not? Yes, but maybe/likely they're only one thing, like, Coryphantha and Escobaria aren't very different and so it should be one thing called "Coryphantha" because... it was there first? (There are rules about which genus name takes precedence when stuff like this happens and I'm pretty sure the rule is seniority.) Anyway, maybe also M. sphacelata is in this newly arranged Coryphantha genus as Coryphantha sphacelata but Coryphantha/Cumarinia odorata IS NOT part of the Coryphantha group and is definitely its own thing, Cumarinia odorata.
4. Is Cochemiea a real genus or not? Wow, sure is, bucko, and it's way bigger than you think. If scientific journal articles were allowed to include the phrase booyah, it surely would have been included here. Also, a bunch of alleged Mammillaria-whatevers just turned out to all be Cochemiea-whatevers. There's a list and presumably cactus enthusiasts and growers and shit now all have to cross-reference and relabel their plants and seed catalogs so that people looking for Mammillaria grahamii will be able to find it now that it's been refiled under Cochemiea grahamii. Bet that's going to be fun for people, especially the nerdy sort of people who have scientific labels on their cactus plants. (Innocent look.)
5. Is looking at morphology *actually* useful in trying to tease out clades/relationships in this group of cactus? Yes, but you have to be looking at the RIGHT MORPHOLOGY. Here, they ran the DNA stuff and then looked at the DNA results to make the cactus groups and then they worked backward toward "what observable features, if any, out of the ten we identified up front do these DNA-defined groups share?". Once you already know what the "group" is, it's not so hard to look at them for "synapomorphy" (a characteristic present in an ancestral species and shared exclusively (in more or less modified form) by its evolutionary descendants) in that group. That's basically cheating. Also, it's patently obvious that looking at morphology is not 100% easy and reliable for TAXONOMIZING MAMMILLARIA-LIKE CACTUS because if it was 100% easy and reliable, we would not have ever had the Mammillaria problem in the first place. Anyhoo, the researchers started with 10 morphologies to look at and in the end wound up with 3 useful morphologies, one for each... group. Hooked spines with Cochemeia, grooved tubercules with Coryphantha (except one), lactiferous ducts with Mammillaria (mostly). There are, again, very nice graphics illustrating this. So, I give this one a "Yes, but..." answer.
Did we get eighteen dollars worth of entertainment out of this paper? Yes, yes we did. Fun times.
