tag:blogger.com,1999:blog-3248412803814730250.post8460158801161313357..comments2024-03-17T01:48:59.504-07:00Comments on The Theropod Database Blog: Peters' dinosaur phylogeny fails finale part 2 - The testMickey Mortimerhttp://www.blogger.com/profile/08831823442911513851noreply@blogger.comBlogger8125tag:blogger.com,1999:blog-3248412803814730250.post-22028103190946346422013-03-31T23:17:04.632-07:002013-03-31T23:17:04.632-07:00I didn't conjure any scores for Daemonosaurus....I didn't conjure any scores for Daemonosaurus. You merely miscoded as usual, and didn't include traits suggested by Sues et al.. An upcoming post will detail that exactly.<br /><br />Did you miss when adding fragmentary Chromogisaurus increased resolution at tree 12? Even if resolution is decreased, it's better to use their added information, then delete unstable taxa from your trees after you've run the analysis. This is standard practice, as described in such papers as Wiens (2006) and Wolsan and Sato (2009). Of course your analysis has so few characters that it's often impractical to add poorly known taxa, but you solve that by adding more characters, not adding less taxa.<br /><br />As for adding rauisuchians and crocs, I already tested your codings without them and found they still support your tree. So their absence isn't affecting things. I wrote it up there in the methods section- "If we run this reduced analysis of Avemetatarsalia plus two outgroups, we get the same result as running the entire analysis of 328 taxa, so we know any changes in topology aren't due to only analyzing this area of the tree."<br /><br />As for your summary, we don't agree on those clades. You include Marasuchus and SMNS 12352 in Theropoda, you have diphyletic sauropodomorphs, you have Pisanosaurus outside Ornithischia and Daemonosaurus in it, your poposaurs are mixed in with silesaurids. Tree 21 doesn't have "reduced resolution" between those. Poposaurs and crocs are basal, then silesaurids and Marasuchus, then ornithischians, then a Saurischia with theropods and sauropodomorphs. How is that reduced?Mickey Mortimerhttps://www.blogger.com/profile/08831823442911513851noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-61360606393580310302013-03-31T22:55:10.315-07:002013-03-31T22:55:10.315-07:00No, in my tree Silesaurus lacks a predentary becau...No, in my tree Silesaurus lacks a predentary because it doesn't actually have one. Sacisaurus also lacks a predentary. Note you code Silesaurus as having a "paired predentary", which is a fictional element not present in any taxon I know. There is no suture in either taxon, so the anterior dentaries are merely toothless and beak-like. This is like Asilisaurus, and if I would add things like the silesaurid characters of Nesbitt, I'd probably recover that as a synapomorphy inside Silesauridae as is standard. I'll also note in your tree since you didn't order any characters, your "paired predentary" state has nothing to do with your "single predentary" state. Even IF you ordered the characters and had correct codings, it would still be more parsimonious for "paired predentaries" to be convergent with real predentaries in your tree, since you have sauropodomorphs and Daemonosaurus separating them. So on multiple levels you didn't find the origin of the predentary.<br /><br />Lotosaurus clearly lacks a predentary (Holliday and Nesbitt, 2013- fig. 6B), and there is no predentary-dentary suture in Shuvosaurus where Peters draws it (Holliday and Nesbitt, 2013- fig. 6F). These are good examples of you coding taxa how you expect them to be instead of what the evidence shows. If you saw the Lotosaurus mandible and didn't know what it was, you wouldn't see a "paired predentary" any more than you would in an oviraptorosaur or ornithomimid. There are numerous vertical breaks in Effigia's mandibles, including two in the surangular which match your drawn surangular-dentary suture. The photos are not high enough resolution to confirm these are cracks instead of sutures, though Nesbitt clearly didn't think they were sutures. Yet with the other three taxa eliminated, there is no precedence for any amniote having a paired predentary, no phylogenetic expectation for Effigia to have any kind of predentary, and I'm unaware of any taxa that lack ventral exposure of the dentary or a dentary symphysis, as Effigia does in your reconstruction. The coronoid and splenial also extend medially to the anterior element, which is standard for dentaries, but unknown for predentaries. Thus even if your sutures are real, the anterior element would more parsimoniously be at least part of the dentary and have no obvious homology to predentaries.<br /><br />Using your characters and taxa, Herrerasaurus is also a derived dinosaur. It's only your miscodings that make it so close to Gracilisuchus and Turfanosuchus.<br /><br />As for Pisanosaurus, I said it only takes three more steps to make the standard Ornithischia topology. And indeed, it only takes one more step to make Pisanosaurus most basal. So tree 21 does not reject that, and even if it did, I haven't tried to add the characters suggested to make it most basal. Remember tree 21 isn't my tree, it's your tree with several characters and taxa added and characters corrected.Mickey Mortimerhttps://www.blogger.com/profile/08831823442911513851noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-5876013963248766272013-03-31T22:24:12.435-07:002013-03-31T22:24:12.435-07:00I don't "want to go with" any tree. ...I don't "want to go with" any tree. 21 is best of those I used here, but it's still bad because it doesn't have e.g. Nesbitt's silesaurid characters, Yates' sauropodomorph characters, Butler's ornithischian characters. The entire point of showing so many trees is that the topology depends on exactly which characters and taxa you include, especially when you have so few characters.<br /><br />The rest of the rauisuchians aren't necessary, since your data recovers your dinosaurian tree when only Arizonasaurus is present. Deleting it from tree 21 doesn't change much either- ornithischians have a slightly different topology, Sacisaurus and Panphagia+Chromogisaqurus are saurischians, and basal sauropodomorphs are ordered differently. <br /><br />Similarly, the rest of your crocs aren't necessary since your data finds your dinosaurian tree without them. I did those tests beforehand for a reason.<br /><br />Deleting Arizonasaurus before running tree 21 leaves the poposaur topology the same, so that instead of Lotosaurus being sister to Arizonasaurus then shuvosaurs, it's sister to shuvosaurs. And in actuality, your charactrers when corrected let Arizonasaurus be sister to Lotosaurus with only one more step. When uncorrected, it takes 14 more steps, so they're almost entirely so far apart due to your miscodings. As for neural spine difference, neither you or I included the dorsal sail as a character, so that's irrelevent.Mickey Mortimerhttps://www.blogger.com/profile/08831823442911513851noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-13649489659031591972013-03-31T21:31:07.412-07:002013-03-31T21:31:07.412-07:00Your Daemonosaurus, the only so-called 'therop...Your Daemonosaurus, the only so-called 'theropod' with a short, high, round rostrum, tiny antorbital fenestra and large postnarial processes on the premaxilla nests apart from Heterodontosaurus, which also shares these traits. These taxa should nest together. That's obvious if evolution indeed works in small steps, as I showed here: http://www.reptileevolution.com/daemonosaurus.htm. I can only wonder what scores you had to conjure to recover pointy-snouted Tawa as a sister to Daemonosaurus. There is no obvious sharing of traits. <br /><br />You've added several very partial skeletons. That will often lead to loss of resolution. I tried to avoid that. <br /><br />Your no. 21 tree does not resolve basal dinosaurs leaving five unresolved clades arising from a very partial Eucoelophysis. Get rid of the partial skeletons. That will help. Also add in several rauisuchians and basal crocs. This will help move the poposaurids back into the Dinosauria where they belong. Every taxon affects every other taxon. When you do these things, I think your tree will more closely resemble mine, a tree which is more logical in the order of the acquisition of traits and sisters greatly resemble one another (share more traits without untenable reversals). <br /><br />Summary: We agree on the various smaller clades (e.g. poposaurids (sans Arizonasaurus), theropods, sauropodomorphs, ornithischians. That's the majority of your tree. Where we differ is their interrelationships in which your best tree has reduced resolution. Your taxonomic exclusions in rauisuchians and crocs seem to be the reason why. Hope this helps. <br />D.P.https://www.blogger.com/profile/13009843520057633239noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-60343011338727375092013-03-31T21:30:41.736-07:002013-03-31T21:30:41.736-07:00So, in your tree Silesaurus has a predentary at th...So, in your tree Silesaurus has a predentary at the base of its own clade (unknown in other members). Sacisaurus has a predentary as a basal dinosaur. This bone is retained in derived ornithischians, and, according to your tree, lost in theropods and sauropodomorphs. Does that seem tenable? <br /><br />Then the question arises: Do Lotosaurus, Shuvosaurus and Effigia have a predentary? That's a matter of opinion at present because in two cases (Lotosaurus and Shuvosaurus) such a bone would have to be fused to the toothless dentary. Effigia gives some indication, but only if you consider the main mandible bone the dentary rather than the surangular. Your tree indicates the predentary preceded the Dinosauria in several clades. Is that the way you see it? <br /><br />Your Herrerasaurus is a derived dinosaur, far removed from Gracilisuchus and Trialestes, which it otherwise greatly resembles. Your tree found Sacisaurus to be closer to Gracilisuchus. That seems odd. My tree found the carnivorous Herrerasaurus close to the similar and equally carnivorous Trialestes and Gracilisuchus. My tree found one origin for the predentary bone. <br /><br />Your Pisanosaurus nests as a derived ornithischian, when in all other trees it nests as a basal taxon. That seems odd. <br />D.P.https://www.blogger.com/profile/13009843520057633239noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-86398020828576070802013-03-31T21:30:16.897-07:002013-03-31T21:30:16.897-07:00Hi Mickey, Dave here.
I'm going to assume your...Hi Mickey, Dave here.<br />I'm going to assume your last card played, no. 21, is the tree you want to go with. I see you worked hard to separate readily identified sisters and worked equally hard to unite several untenable pairings. You also recovered many of the same sisters I did. So, on to the differences and why you recovered differences. I think you'll find this instructive. <br /><br />First of all, where are the rest of the rauisuchians? Deleted? Why? They're necessary. Especially if you're going to include one rauisuchid, Arizonasaurus. <br /><br />Where are the rest of the basal crocs? Or at least some of them. They also are necessary for understanding basal archosaurs. Why? Because, like poposaurs, they also independently developed a large calcaneal tuber in derived taxa. That's key to their understanding. <br /><br />Arizonasaurus and Lotosaurus - the skulls aren't at all similar. That's obvious. If they're sisters, as your tree indicates, they should be similar. This is why I asked you to place images of your potential sisters next to one another. Such a final "zipper check" is designed to discover bad nestings quickly before embarrassment sets in. My tree nested the carnivorous Arizonasaurus with rauisuchids, which were deleted from your study. How can Arizonasaurus nest with rauisuchians if you don't include them? This echoes the same problem I noted earlier when pterosaurs are added to dinosaur studies while excluding their true sisters, Cosesaurus, etc. Furthermore, Parker and Nesbitt (2013) report, "Interestingly, the morphology of the neural spines that make up the sail in Lotosaurus adentus differs significantly from those of Arizonasaurus and Ctenosauriscus, supporting the idea that it was independently derived (Butler et al. 2011; Nesbitt 2011)." D.P.https://www.blogger.com/profile/13009843520057633239noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-34875588629770434332013-03-30T05:07:16.176-07:002013-03-30T05:07:16.176-07:00Wow - impressive stuff Mickey.Wow - impressive stuff Mickey.Mark Wildmanhttps://www.blogger.com/profile/17644970562043088893noreply@blogger.comtag:blogger.com,1999:blog-3248412803814730250.post-22208880024385623252013-03-30T01:06:00.115-07:002013-03-30T01:06:00.115-07:00Can't say I understand the details, but - wow....Can't say I understand the details, but - wow. Certainly starting to understand the process. Thanks!davidmaashttps://www.blogger.com/profile/16696298300141402317noreply@blogger.com