Peters has defended his amniote analysis as being superior to other published analyses, since it includes numerous genus-level OTUs. And in this he is correct, though there are other measures of analysis quality too. One obvious example is the number of characters. There's no hard and fast rule regarding how many characters are needed to analyze a certain number of taxa, but you should at least include all characters previously utilized for the clade you're examining, unless some are invalid for whatever reason. Peters' analysis has a large number of areas in which it differs not only from the consensus, but from any other proposed phylogeny. One of these is his Dinosauria, which instead of the standard Ornithischia vs. Saurischia dichotomy, with the latter including Theropoda and Sauropodomorpha, has...
--Theropoda including Marasuchus, Herrerasaurus and Trialestes
--- Panphagia+Pampadromaeus
---- a clade of Pisanosaurus, poposaurs and silesaurids
----- Sauropodomorpha including Saturnalia, Thecodontosaurus and Massospondylus
------ Daemonosaurus + standard Ornithischia
Thus it has an equivalent of Phytodinosauria and includes a host of non-dinosaurian taxa. Now, Saurischia is one of those clades (like Ornithothoraces) that while near-ubiquitous in published analyses, only takes a few extra steps to negate, so I'm not biased against it a priori. And I've said half-seriously before that ornithischian-like silesaurids and theropod-like shuvosaurines and hallopodines could threaten dinosaurian monophyly. Peters no doubt thinks his results are superior to the consensus, but did he actually test the consensus? Let's see which saurischian and dinosaurian synapomorphies (based on only the unambiguous synapomorphies of Nesbitt, 2011) Peters included in his analysis.
Saurischia
Yes- (1) Subnarial foramen between the premaxilla-maxilla where the border of the foramen is present on both the maxilla and the premaxilla (12-1)*.
(12) Metacarpal IV shaft width significantly narrower than that of metacarpals I–III (261-1).
(13) Distal end of pubis expanded relative to the shaft (5 pubis boot) (283-1).
No- (2) Lacrimal folds over (5 overhangs) the posterior/posterodorsal part of the antorbital fenestra (38-1)*.
(3) Ventral ramus of the opisthotic covered by the lateralmost edge of exoccipital in posterior view (111-1). (4) Foramen in the ventral part of the splenial (165-1)*.
(5) Epipophyses present in posterior cervical vertebrae (cervicals 6–9) (187-1).
(6) Hyposphene- hypantrum accessory intervertebral articulations in the posterior cervical and/or dorsal vertebrae (195-1)*.
(7) First primordial sacral rib C-shaped in lateral view (209-1)*.
(8) Manual length accounts for more than 0.3 but less than 0.4 of the total length of humerus plus radius (247-1).
(9) Distal carpal V absent (249-1)*.
(10) Manual digit I, first phalanx (1.1) is the longest nonungual phalanx of the manus (253-1)*.
(11) Lateral distal condyle strongly distally expanded relative to medial condyle of metacarpal I (254-1). (14) Cross section of the distal portion of the ischium subtriangular (293-2).
(15) Deep groove between the lateral condyle and crista tibiofibularis on the distal surface of the femur (322-1).
(16) Posterior prong of distal tarsal 4 pointed (350-1)*.
(17) Distal tarsal 4 with a distinct medial process present in the anteroposterior middle of the element (351-1)*.
(18) Astragalus, proximal surface possesses a marked rimmed and elliptical fossa posterior to the anterior ascending process (359-1)*.
Dinosauria
Yes- (7) Fourth trochanter a sharp flange (316-1)*.
No- (1) Exoccipitals do not meet along the midline on the floor of the endocranial cavity (115-1)*. Also in Crocodylomorpha and Effigia + Shuvosaurus.
(2) Supratemporal fossa present anterior to the supratemporal fenestra (144-1).
(3) Epipophyses present in postaxial anterior cervical vertebrae (186-0)*.
(4) Apex of deltopectoral crest situated at a point corresponding to more than 30% down the length of the
humerus (230-1)*.
(5) Radius shorter than 80% of humerus length (241-1).
(6) Proximal articular surfaces of the ischium with the ilium and the pubis separated by a large concave surface (297-2)*.
(8) Fourth trochanter asymmetrical, with distal margin forming a steeper angle to the shaft (317-1).
(9) Cnemial crest arcs anterolaterally (328-2)*.
(10) Distinct proximodistally oriented ridge present on the posterior face of the distal end of the tibia (336-1)*.
(11) Proximal articular facet for fibula of the astragalus occupies less than 0.3 of the transverse width of the element (362-1).
(12) Concave articular surface for the fibula of the calcaneum (378-2).
Wow! Peters included 3 (17%) of saurischian characters and 1 (8%) of dinosaurian characters and didn't recover either clade? Who'd have thunk? In general, Peters' postcranial characters are few and usually reflect gross morphology. Wanna know the humeral characters included? Humerofemoral length and torsion. The tibial ones? Tibioilial length. Distal tarsal characters? None. Peters is fond of saying his sister taxa "look similar", and this is why- the characters reflect a general morphotype.
There's a reason making an amniote analysis is so challenging. It'd be easy enough to enter 300 taxa into Rieppel's or Mueller's analyses, but at that resolution, you need to add more characters to arrange taxa right within each subclade. Peters has not done this, but I'd bet doing so, and defining his states better, would result in a more standard topology, albeit one that may have genuine surprises.
Wednesday, June 13, 2012
Monday, June 4, 2012
Why David Peters' analysis sucks
For a while now, Peters has played the part of the outcast whose new ideas have been shunned by the outdated but stubborn establishment, locked in its consensus purely due to tradition and an unwillingness to include numerous genus-level OTUs in an analysis. And he actually has a point. Published large scale reptile analyses DO suffer from using suprageneric OTUS, are heavily miscoded, and include way too few taxa to be useful. However, Peters' analyses are even worse. One huge reason is that they include what are certainly wrong codings caused by Peters' method of running photos through his computer to "find" things people looking at the specimen missed, like the entire posterior skeleton of Longisquama, babies and crazy frills on it and various pterosaurs, teeth in toothless taxa, and anything else you can imagine. But even ignoring that, there's another level at which Peters' analysis fails, and that's when we look at the characters themselves. I asked Peters for his matrix today (and I will credit him for being so transparent as to send it to anyone who asks, despite his study not being published in peer reviewed literature), and here's what I found...
Let's just look at dental characters. Character 107 is palatal teeth present vs. absent, but character 96 covers vomer teeth being absent and character 101 covers pterygoid teeth being absent, so this is correlated right off the bat.
Character 108 is number of premaxillary teeth with the states four, more than four, one two or three, none, teeth fused and three or four plus medial. Ack, where to begin? First, no character is ordered, so while "four" should be between "1-3" and ">4", it's not. Second, multiple variables are involved here. Each character should only involve one variable. Having medial premaxillary teeth should be its own character. Having teeth fused should also be its own character. And since we don't know toothless premaxillae evolved via decreasing number of teeth, a toothless premaxilla should also be its own character. Any coding not falling within the states given should be inapplicable (e.g. Citipati for number of premaxillary teeth 1 vs. 2 vs. 3 vs. 4. vs. 5 vs. 6 vs. 7 or more). A good way to test for how well character states are defined is to see if any possible morphologies couldn't be coded. In this case, what if a taxon had six premaxillary teeth with medial teeth too? Or what if it had less than four fused teeth?
Character 109 is medial premaxillary teeth larger vs. smaller than lateral teeth. And yet all dinosaurs are coded 0, when none have medial premaxillary teeth. They should be coded inapplicable. This is the same for many derived reptiles, no doubt. I notice Thecodontosaurus is coded as derived though, when no premaxilla is known, and even Pantydraco only has lateral teeth.
Character 110 is premaxillary teeth robust vs. tiny to absent. Yet absent was already covered by character 108, and these states are never defined so that I could always know when a specimen has state 0 or 1.
Character 111 is procumbant premaxillary teeth, which is fine except needing an angle at which teeth are defined as procumbant.
No complaints about character 112- maxillary tooth length less than twice root depth (presubamly FABL).
Character 113 is canine maxillary teeth, but Peters codes taxa with toothless maxillae such as Hupehsuchus and Endennasaurus as lacking canine maxillary teeth when they should be inapplicable.
Character 114 is the position of the last maxillary tooth compared to the orbit, but is not ordered, so "mid orbit" is not between "anterior to orbit" and posterior to orbit". Furthermore, it includes a state for "toothless", which makes it correlated with the next character...
Character 115 is maxillary teeth sharp vs. blunt vs. multicusped vs. multicusped with constricted bases vs. absent vs. constricted bases. This makes 108 look formatted well. SO MANY variables are combined! Sharpness, number of cusps, constriction of base, presence. This needs to be four characters.
So yeah, that's just a sample. Can you all see why no one trusts Peters' analysis, even ignoring his coding accuracy in the first place? It's true some other analyses have some of the same issues, like Rieppel's (http://theropoddatabase.blogspot.com/2011/06/rieppels-reptile-matrix-and-turtle.html), but this doesn't make Peters' more likely to be correct. Honestly, Rieppel's, Mueller's and Peters' are all flawed. I don't trust the results of any of them. If someone wants to be the true shining light on the subpar tradition of reptile analyses, they need to understand how to construct characters well, and not put their faith in graphic artifacts created by Photoshop. Peters can accomplish the first step by posting this on his blog and adusting the characters as per my critique.
Let's just look at dental characters. Character 107 is palatal teeth present vs. absent, but character 96 covers vomer teeth being absent and character 101 covers pterygoid teeth being absent, so this is correlated right off the bat.
Character 108 is number of premaxillary teeth with the states four, more than four, one two or three, none, teeth fused and three or four plus medial. Ack, where to begin? First, no character is ordered, so while "four" should be between "1-3" and ">4", it's not. Second, multiple variables are involved here. Each character should only involve one variable. Having medial premaxillary teeth should be its own character. Having teeth fused should also be its own character. And since we don't know toothless premaxillae evolved via decreasing number of teeth, a toothless premaxilla should also be its own character. Any coding not falling within the states given should be inapplicable (e.g. Citipati for number of premaxillary teeth 1 vs. 2 vs. 3 vs. 4. vs. 5 vs. 6 vs. 7 or more). A good way to test for how well character states are defined is to see if any possible morphologies couldn't be coded. In this case, what if a taxon had six premaxillary teeth with medial teeth too? Or what if it had less than four fused teeth?
Character 109 is medial premaxillary teeth larger vs. smaller than lateral teeth. And yet all dinosaurs are coded 0, when none have medial premaxillary teeth. They should be coded inapplicable. This is the same for many derived reptiles, no doubt. I notice Thecodontosaurus is coded as derived though, when no premaxilla is known, and even Pantydraco only has lateral teeth.
Character 110 is premaxillary teeth robust vs. tiny to absent. Yet absent was already covered by character 108, and these states are never defined so that I could always know when a specimen has state 0 or 1.
Character 111 is procumbant premaxillary teeth, which is fine except needing an angle at which teeth are defined as procumbant.
No complaints about character 112- maxillary tooth length less than twice root depth (presubamly FABL).
Character 113 is canine maxillary teeth, but Peters codes taxa with toothless maxillae such as Hupehsuchus and Endennasaurus as lacking canine maxillary teeth when they should be inapplicable.
Character 114 is the position of the last maxillary tooth compared to the orbit, but is not ordered, so "mid orbit" is not between "anterior to orbit" and posterior to orbit". Furthermore, it includes a state for "toothless", which makes it correlated with the next character...
Character 115 is maxillary teeth sharp vs. blunt vs. multicusped vs. multicusped with constricted bases vs. absent vs. constricted bases. This makes 108 look formatted well. SO MANY variables are combined! Sharpness, number of cusps, constriction of base, presence. This needs to be four characters.
So yeah, that's just a sample. Can you all see why no one trusts Peters' analysis, even ignoring his coding accuracy in the first place? It's true some other analyses have some of the same issues, like Rieppel's (http://theropoddatabase.blogspot.com/2011/06/rieppels-reptile-matrix-and-turtle.html), but this doesn't make Peters' more likely to be correct. Honestly, Rieppel's, Mueller's and Peters' are all flawed. I don't trust the results of any of them. If someone wants to be the true shining light on the subpar tradition of reptile analyses, they need to understand how to construct characters well, and not put their faith in graphic artifacts created by Photoshop. Peters can accomplish the first step by posting this on his blog and adusting the characters as per my critique.
Subscribe to:
Posts (Atom)