Here's how the biggest titanosaurs rank out in first-last place:
1. Tie between Alamosaurus (referred Mexican fibula + Fowler & Sullivan's neck centrum), Puertasaurus (1 cervical, 1 dorsal, 2 unpublished caudals) and the "Chubut Monster" (majority of skeleton from at least six specimens). All of of these animals appear to top out around 120+ ft. long and probably 100 tons.
2. Tie between Argentinosaurus and the "MLP Monster" (briefly mentioned by GSP, 1988 with estimate measurements, and lost to history since). Both of these animals were probably pushing 110+ ft. long and 80-90 tons
3. Tie between Ruyangosaurus (cervical rib, anterior and posterior dorsals, additional unpublished dorsals, dorsal rib, upper femur, tibia), Notocolossus (dorsal and caudal vertebrae, foot, and limb elements) and "Argyrosaurus" sp. (the larger referred femur FMNH 13018) - probably between 75-90 tons. Ruyangosaurus may have gotten longer than 100 ft., Notocolossus and "Argyrosaurus" were probably not as long but still huge at 90+ ft.
4. Tie between Dreadnoughtus (majority of skeleton) and Paralititan (partial sacrals, 2 caudals, humerus, and shoulder and arm fragments) - both are around 80-90 ft. long and 65-70 tons (no, I do NOT buy the wimpy shrinkwrapped GSP estimates of 20 tons for Dreadnoughtus - he never accounted for the heavily crushed ribs and erosion, let alone a REALISTIC amount of soft tissue)
5. Tie between Futalognkosaurus (much of skeleton) and "Antarctosaurus" giganteus (2 femurs, pubis fragments, other random bits) - both at around 90ft. long and 65 tons. Either one of these animals could switch places with 4th place, it's a very close call. There was a bombed-to-dust centrum from "Aegyptosaurus sp." which may also have been a possible contender, or could be a Paralititan part.
6. Good luck figuring out who goes here! A lot of large titanosaurs could fill this slot of 70-80 ft. stompers. Traukutitan is definitely in the running, as are several undescribed femurs in various parts of Argentina.
As for the biggest non-titanosaur Somphospondyli:
1. The "French Monster" aka 'PEPE' (at least two specimens, including femurs, humerus, fibulae, caudals, toe bones and other assorted parts) - 110+ ft. and 90+ tons. Gracile Paluxysaurus-morph Chubutisaurid by the looks of things. Larger than Sauroposeidon, and there is a possibility with some of the bones that these estimates estimate are conservative, and this animal MAY turn out to be the biggest dinosaur yet found.
2. "Huanghetitan" ruyangensis (ribs, sacrum, caudals, unpublished cervicals) - 90+ ft. and 70-80 tons?
2. Yunmenglong ruyangensis (large part of the skeleton) - giant euhelopodid similar to Huabeisaurus in shape. 100+ ft. and 60 tons?
3. Daxiatitan binglingi (nearly complete cervical and dorsal series, sacrum and ilia, femur, scapula, isolated caudal, etc.) - giant euhelopodid, 95 ft. and 50-55 tons?
4. "Liaoningtitan" (unofficial name, not yet published. Partial skeleton, appears to missing the original arms and some cervicals and caudals at least). Giant euhelopodid. Dimensions unknown, but appears to be around 75-80 ft. long. Mass unknown but likely around 30+ tons?
5. Huanghetitan liujiaxiaensis (scapula, coracoid, sacrum, caudals etc. possibly some unpublished cervicals and dorsals) - stocky titanosauriform closely related to the euhelopodids. 60 ft. and 30 tons? (hard to tell since the museum mounts appear to be largely speculative, the true neck length is not certain.)
1. Tie between "Brachiosaurus" nougaredi and Breviparopus (if either of them are indeed brachiosaurs). Both around 120ft. long and 75+ tons?
2. Fusuisaurus zhaoi (fragments of ribs, hips, femur and first 3 caudals) - 100ft. long and 70 tons?
2. Brachiosaurus sp. (Potter Creek specimen) - 95ft. long and 60+ tons?
3. Abydosaurus macintoshi (referred adult ribs) - 95ft. long and 55+ tons?
3. Giraffatitan (HMN XV2) - 85-90ft. long and 50-55 tons?
4. Tie between "Ultrasauros" (BYU scapulacoracoid, referred partial scapula and dorsal), Brachiosaurus alithorax (holotype and referred specimens), and "The Archbishop" (majority of skeleton) - 80-85ft. long and 40 tons?
5. Lapparentosaurus (Lydekker's giant caudals) - 79ft. long and 35 tons?
6. Lusotitan (femur, tibia, fragments of hips, dorsals, ribs, caudals, fibula, humeri, scapula) - 72ft. long and 30+ tons?
1. "Mamenchisaurus" sinocanadorum (116ft. long and 70 tons?)
2. Chuangjiesaurus (90ft. long and 50 tons?)
3. "Mamenchisaurus" jingyanensis (85-90ft. long and 45 tons?)
4. Hudiesaurus sinojapanorum (80+ft. long and 35-40 tons?)
5. "Mamenchisaurus" anyuensis ????? (dimensions are a bit murky but allegedly huge, comparable to jingyanensis territory.)
6. "Omeisaurus" jiaoi (80ft. long, 20 tons?)
1. "Amphicoelias" fragillimus (as per Zach Armstrong's latest reconstruction, most likely a basal rebbachisaur-like diplodocoid at 95ft. long and 55+ tons - a far cry from previous estimates using Diplodocus-like proportions) - may be similar in size to Parabrontopodus distercii.
2. Tie between Supersaurus (majority of skeleton from multiple specimens) - 120ft. long and 50 tons, Dinheirosaurus (similar size to Supersaurus), and Apatosaurus ajax (Oklahoma specimens) - at least 95ft. long and 45 tons
3. Probably some sort of African barosaurine, Tornieria sp. or other.... who knows lol.
And then of course with the really basal sauropods you have Turiasaurus, which some reports overhyped as "the biggest dinosaur" and for some reason people still like to believe it was some kind of supergiant - yet it lags far behind most of the animals on this list, at 70-80ft. and MAYBE 30 tons (like most basal sauropods, it's relatively lightly built and most of its length is tail). Its size has been greatly exaggerated and basically it just looks like an oversized Camarasaurus or Jobaria with a crazy-long tail attached. Turiasaurus may be the biggest turiasaurid (a family that has only a handful of known species anyway), but that's the only place where it holds any kind of record. All of the major neosauropod groups (see above) have species that easily outclass Turiasaurus.
And questions or suggestions, feel free to give input.
2. Wikipedia sucks. To the extent that they get things right, it's only because of the edits of people (most of whom I know) that have read the current research papers and are sharp enough to question their conclusions when they are flawed. However Wikipedia is always being changed and usually their edits get reversed because Wikipedia has far more idiots than genuine paleo-aware people (i.e. lots of Aweseombro's that insist Seismosaurus must be 150ft. long, or that Argentinosaurus or the Chubut Monster must be over 200 tons or some other stupidly exaggerated claims).
If you want the truth, don't rely on Wikipedia. Rely on people (preferably artists and art-appreciative scientists with a sense of the true scale of these giants) who have actually read and understood the papers and the problems of scaling from scarce remains, and thus only have a few disagreements about likely sizes (nearly all of which are from very incomplete fossils).
Those people are (off the top of my head): Myself, Bricksmashtv, Paleop, Franoys, Brolyepyfusion, ChrisMasna (to the extent that he does sauropods at all, which is mostly due to my megalomaniac influence!) and especially the great folks at SV-POW, all of whom have PhDs and have seen a lot more sauropod specimens than anyone else you will talk to (and have their museum codes memorized too!)
1. Some of the bones are sculpted placeholders rather than casts of the actual bones, and could be oversized. There was a lot of material but no one complete specimen from what I know.
2. There may be too many sculpted vertebrae, or just casts of repeat positions from overlapping skeletals.
3. It may really be 120 feet long. The skeleton is at best a composite cast, based on different-sized individuals, perhaps one of them was 120ft. long, i.e. Puertasaurus range. I'd estimate most of the individuals are the site were closer to 100ft. though. The huge femur on the pallet may belong to an animal 120ft. long.
It's hard to tell until we have scaled photos of all the specimens, so far the Chubut Monster's actual proportions are a bit murky (even a few percent difference on one body part can have huge effects on overall size and porportions, this is coming from someone who has scaled and rescaled lognkosaurs over and over again as new photos and data become available, due to even published measurements being inconsistent).
That's probably what happened to the AMNH titanosaur too.
The Argentinosaurus cast is SO badly done it's not even funny. Even if you ignore the awful neck posture and the cheesy steel-wire teeth. The worst part is that most of the "replicas" are NOT based on related dinosaurs - they are 100% fake imaginary sculpts - and not even good ones. The cervical ribs aren't even aligned together, they're all spiked out downwards as if bursting out of the neck (no reason for doing this in a titanosaur whatsoever, especially as the cervical ribs are all FAKE in this mount anyway). The cervical vertebrae are almost all copies of each other (again all fake, with obviously fake sharp edges on the spines and postzyg laminae, but at least they could have done done a few length and shape variations!) and don't even get me started on the crazy-oversized lower limbs and that ridiculously over-long tail with a diplodocus-whip on the end. There were some titanosaurs with whip-end tails, but the whip was thicker and a LOT shorter, did these guys even read Powell (2003)? It appears the ilia are also oversized, and the scapulae definitely are (not to mention their angle is far too low). Oddly the dorsal ribs are undersized (at least the middle ones) and the rib cage too shallow. But there are somehow 13 dorsals instead of 10 or 11!? This is a titanosar, not a turiasaur!
I don't see either of the two Argentinosaurus specimens being much more than 110 ft. long. The Fernbank mount oversized everything, even the fake feet were oversized relative to the already oversized fake tibiae, and whoever designed the thing also invented some extra foot/ankle bones that didn't exist! Same color as the other bones! In place of where the cartilage would have been....
By contrast the only major biffs with the AMNH's Chubut Monster mount appear to be some odd scaling in the torso (versus the neck and hips) and perhaps too many tail vertebrae. Oh yeah, and the head. There's no way a main-line lognkosaur had that sort of head. They made it look like a nemegtosaurid head, only even MORE squashed. It's even more flattened than the ROM's Futa head (which was a scaled up Rapetosaurus skull clone), and that was something I advised against from the beginning!
Titanosaurs were a diverse group and evolved different head shapes for different feeding strategies depending on the family (remember, they were taking over niches from older sauropod groups that became extinct). However their teeth, braincases, and the rear parts of the skull were pretty similar across the board. Where they differed was mainly in the shape and size of the nostrils and the jaws. Basically the front 60% of the head (aside from the teeth) was where you find the big variations.
Amphicoelias is not that closely related to Diplodocus. Several new cladistic anlyses in the last few years place it close to the basal end of diplodocoidea. Somewhere near the rebbachisaurids in other words. So it ends up being a much shorter-bodied animal than it would have been if it were proportioned like Diplodocus. Still huge though.
On a smaller-scale extinction like the LJ-EK boundary, which killed off the true diplodocids but plenty of dinosaurs and other warm-bloods still made it through, it was the more "all-purpose" types that survived. Dicraeosaurs with their rounded mouths had more diverse feeding options than true diplodocids, so less risk of starving to death in case one food source is wiped out. Early rebbachisaurs (before they became vacuum-mouths) were probably much the same. Brachiosaurs and basal somphospondyls were more derived than any diplodocoids, but they did well because they had rounded mouths full of teeth, plus much more neck range than dicraeosaurs, so literally they could eat anything with leaves. We don't know when the very first titanosaurs appeared, but likely they also had a similar mouth shape, as Malawisaurus retains a basal oval mouth with well-distributed teeth. So in this case it isn't so much about who is "basal" or "derived" but rather who is better suited to a wide range of feeding and habitat options. If the K-T extinction had been less severe, more like the LJ-EK one where many big dinosaurs survived, it's likely that basal and intermediate titanosaurs (wide variety of feeding options) would have dominated and dedicated low-grazing saltasaurs would have been in huge trouble.
The best explanation is that features or sets of features that we label "basal" and "derived" don't necessarily equal "outdated" and "updated". The basal offshoots are often (but not always) less specialized than the derived forms in a clade, which means that they don't do as well in times of high biodiversity and niche specialization, but if there's a big extinction, usually the more over-specialized animals die out first. This means usually the derived members of a clade are the most vulnerable to a mass-extinction, and the basal ones suddenly have an edge.
In normal circumstances the derived ones dominate, but when a mass-extinction occurs it's actually often better to be a generalist "basal" species. So it's likely that rebbachisaurs and dicraeosaurs didn't out-compete diplodocids so much as they were more adaptable to the LJ-EK mass-extinction (i.e. less specialized feeders that could eat more types of plants - keep in mind the hyper-specialized "vacuum mouth" rebbachisaurs like Nigersaurus came millions of years later).
Interestingly, the meat eaters had the opposite happen: ceratosaurs are considered more basal than allosaurs, but most ceratosaurs died out in the LJ-EK extinction because they were the ones that got too specialized (extremely oversized teeth, much more front-heavy, highly reduced arms) whereas allosaurs were more derived in their internal engineering, but retained a pretty general-purpose predatory tool kit (big arms and hand claws, smaller teeth, more flexible jaw joints). Thus Allosaroids dominated the immediate aftermath of the the extinction, with Acrocanthosaurus and the carcharodontosaurs and neovenatorids being top predators. Ceratosauroids made a comeback later as abelisaurids and noasaurids, but these were nearly always 2nd-tier or 3rd-tier predators (unless you're in India or Madagascar).
I think you aren't giving Turiasaurus enough credit. Based on the circumference of its humerus, it's in the same size class as Paralititan.
I have to say I was initially shocked at how low Antarctosaurus giganteus was, but after looking at the estimates in the Notocolossus paper, I guess you're right. It must be fairly gracile by titanosaur standards.
What makes you say Tornieria is larger than Barosaurus or Diplodocus? I admit to not being particularly familiar with it, though.
If you go by limb-bone circumference (which is always only the midshaft circumference, endpoint sizes be damned), you get hilarious things like Dale Russell's 15-ton Giraffatitan. You think Greg Paul's version is bad, you ain't seen nothing! It's a vastly unreliable measure, and says nothing about how much muscle actually went on the bone. Limb-bone circumference is one measurement out of many, it fails, because it literally fails to take into account differences in rib cage proportions (highly variable between basal sauropods and derived ones), muscle crests on not just the humerus but all the other bones, and every other factor that feed into overall volume. Limb-bone circumference is pretty much bankrupt as a mass estimate method since the 1980s, a GDI based on full skeletals or even the basic water-dunk method will yield better results than that as long as your model is accurate. Different species and families can vary in mass by up to a factor of two for any given humerus circumference. What is your "universal" reference ratio for humerus circumference-to-body mass based on anyway? Elephants? Giraffes? Antelopes? Humans?
The more basal a sauropod is, generally the more heavy-boned it is (less pneumaticity, wider humerus midshaft relative to the ends, etc.). Turiasaurus may have had a big humerus, but this is very different than "being in the same size class" (turiasaurs appear to be unusually long-limbed for their volume) which is usually understood as either length or mass. Most of Turisaurus' body length was tail. And it had a relatively slim torso as is common for basal sauropods. It's basically an oversized Patagosaurus look-alike. Huge tail, on a midsized body (comparable to the biggest camarasaurs), short neck and very long legs. Far from the same size class as Paralititan mass-wise, even if the humerus may rival Paralititan's in midshaft circumference. With more derived sauropod groups, you will always get abnormally and unrealistically low masses if you only rely on limb-bone circumference to estimate mass. They are just more efficient than basal sauropods at supporting more mass with less bone, even in the robust forms (which is why a 60+ ton Dreadnoghtus or Futalognkosaurus is actually likely, looking at the muscle crests and the overall volume of the rib cage and other structures when they are properly de-crushed and restored).
Compare them to medieval cathedrals. Turiasaurus is Chartres. Very solid-boned, more than is necessary for its weight. The giant titanosauriformes are like Amiens. Bigger, more spacious and much more pneumatized with generally lighter limb shafts and larger muscle crests, a lighter skeleton (cubic inch for cubic inch) which is more efficient in supporting heavier weights.
BTW, a lot of the low mass estimates for Dreadnoughtus seem to be based on midshaft circumference too, which again is an unreliable indicator across clades. I acknowledge it may be less than 65 tons, maybe even as low as 50 tons, but until I do a restoration and have one of my trusted associates do the GDI, it's not certain. Most important is how the Paleo-King skeletal will compare to those for Argentinosaurus, Futalognkosaurus, etc. and that is skill an unknown though you are welcome to guess. I do know that people are lowballing it at 20 tons of some other ridiculous number, and it does rival many of the giants, so I skew high because too many are skewing low. Certain "experts" (they know who they are) think that paleontology is like politics, and have no interest in either facts or compromise and are constantly hogging attention, so getting their feathers ruffled is often more useful than spitting out some modest number they can easily ignore as they feed the ego.
There are some pretty big Tornieria remains which exceed Diplodocus, Bricksmashtv has more data on those and estimated their large sizes. Since Barosaurus itself is turning out to be bigger than once thought, it would make sense that Tornieria was also a lot bigger than Diplodocus.
I’m going to reply to both of your replies in one reply for simplicity’s sake, as both cover similar subjects.
I’m afraid you misunderstand why limb bone circumferences are used to estimate body mass. It isn’t that the sum of circumferences of the midshafts of the humerus and femur (let’s call this the combined stylopodial midshaft circumference, or CSMC for short) are simply an easily-measured proxy for body size, but that these values necessarily are closely correlated with body mass. The midshafts of the humerus and femur are a limiting factor in body mass, as the limb bones must be thick enough to provide the structural support to hold up the entire body. There is also a strong selective pressure to not make the limb bones any bigger than they need to be to adequately support the body at all times, due to the energetic costs of growing thicker bones. Therefore, the CSMC would be expected to correlate closely with live body mass, a hypothesis that has been strongly supported across almost all quadrupedal amniotes based on empirical evidence (Campione and Evans 2012). This equation has been tested and applies accurately to animals as diverse as turtles, crocodiles, frogs, elephants, deer, and mice. Equations based on limb bone dimensions remain accurate regardless of the build of the body. By contrast, there are multiple levels of informed speculation involved in calculating a volumetric mass estimate, such as reconstructing the skeleton and estimating the densities and proportional volumes of different soft tissues. For example, the effect of pneumaticity on body mass is substantial, but requires extensive guesswork (Wedel 2005). The overall volume of soft tissue is also uncertain, so volumetric estimates of sauropods can vary by nearly a factor of two across a range of plausible soft tissue volumes (Bates et al. 2015). In other words, while CSMC scaling is doubtlessly an imperfect method—no method is perfect, after all—it is a fairly reliable technique that requires minimal guesswork, unlike any volumetric method. It allows for a testable method that can be used to compare disparate taxa and give a general idea of how massive different species were relative to one another.
You make several specific points which I doubt. While Turiasaurus and Paralititan did not exactly match in proportions, it beggars belief that Paralititan could support over twice the mass of Turiasaurus despite having slenderer legs. Turiasaurus may have been narrower, but it also had a longer torso than a comparably-sized titanosaur, and was likely denser due to the lesser pneumaticity. Much of the apparent relative smallness of Turiasaurus comes from its neck, which is far shorter than the neck of Dreadnoughtus or other giant titanosauriforms. As a ballpark estimate, if Turiasaurus had a titanosaur-like neck, it probably would have been at least 25, maybe 30 meters long. Your argument that CSMC methods underestimate titanosaur masses relative to other techniques is demonstrably incorrect. The highest published estimate of the mass of Dreadnoughtus is also the only estimate of its mass based on midshaft circumferences, whereas the highest volumetric estimate is only two-thirds that —your statement that “a lot of the low mass estimates for Dreadnoughtus seem to be based on midshaft circumference” has it completely backwards (Bates et al. 2015; Lacovara et al. 2014). I especially find it odd that you claim that the same method wildly overestimates mass in Turiasaurus and underestimates mass in Dreadnoughtus. Finally, the largest reported femora of Tornieria africana are substantially smaller than those of Diplodocus, Galeamopus, or Barosaurus (McIntosh 2005; Janensch 1961). I may have missed something, but I would argue that Barosaurus lentus or Diplodocus hallorum would be more deserving of third place, even if one considers Brontosaurus ineligible.
I would like to know what basis you have for your own mass estimates, out of curiosity.
I completely agree that some “experts” have no interest in facts, though. One comes to mind right now.
Bates, Karl T., Peter L. Falkingham, Sophie Macaulay, Charlotte Brassey, and Susannah C. R. Maidment. 2015. “Downsizing a Giant: Re-Evaluating Dreadnoughtus Body Mass.” Biology Letters 11 (6): 20150215. doi:10.1098/rsbl.2015.0215.
Campione, Nicolás E., and David C. Evans. 2012. “A Universal Scaling Relationship between Body Mass and Proximal Limb Bone Dimensions in Quadrupedal Terrestrial Tetrapods.” Bmc Biology 10 (1): 60.
Janensch, Werner. 1961. “Die Gliedmaszen Und Gliedmaszengürtel Der Sauropoden Der Tendaguru-Schicten.” Palaeontographica Suppl. 7 (September): 177–235.
Lacovara, Kenneth J., Matthew C. Lamanna, Lucio M. Ibiricu, Jason C. Poole, Elena R. Schroeter, Paul V. Ullmann, Kristyn K. Voegele, et al. 2014. “A Gigantic, Exceptionally Complete Titanosaurian Sauropod Dinosaur from Southern Patagonia, Argentina.” Scientific Reports 4 (September): 6196. doi:10.1038/srep06196.
McIntosh, John S. 2005. “The Genus Barosaurus Marsh (Sauropoda, Diplodocidae).” In Thunder-Lizards: The Sauropodomorph Dinosaurs, edited by Virginia Tidwell and Kenneth Carpenter, 38–77. Life of the Past. Bloomington: Indiana University Press.
Wedel, Mathew J. 2005. “Postcranial Skeletal Pneumaticity in Sauropods and Its Implications for Mass Estimates.” In The Sauropods: Evolution and Paleobiology, edited by Kristina A. Curry Rogers and Jeffrey A. Wilson, 201–29. Berkeley: University of California Press.
Allometry based methods rely in a reationship on how much structural support has a limb to provide to support an animal in earth, but it is not like a limb with the same dimensions can just support an animal of an exact weight, it is well known that allometry based estimations have a relative error of up to over 50% svpow.com/2011/01/20/tutorial-… ; so the same limb bone can support an animal 50% heavier or 50% lighter than the median value , applying this to a median value of 6 tonnes, the same bone can support a 3 tonnes and a 9 tonnes animal. Of course it would also be able to support an animal of less than 3t, but here the selective pressures that you talked about come into play.
Answering your question; can the same bone support certain animal and another one twice it's mass? Yes, definately. It is a bit extreme, but still possible. 4.5 t and 9 t for example fall within the range proposed above. So do 3 t and 6 t, or 3.5 t and 7 t. Specially since we are comparing only one bone and the animal has four limbs with more bones than just the humerus, the surprising thing here is that how both estimates relate based on the humeri doesn't have why to be the exact same as to how the femur based estimations relate, or the tibia or ulna based ones. This can be seen in Mazzeta 2004, in which some bones yield different estimations than the others. This is also why Campione and Evans use an equation that takes both the humeri and femur into account. (Having said this, Paralititan only yields 44 t scaled from Asier's Elaltitan, of which the mass seems just an educated guess, so in the end it is possible that it wasn't that much bigger than Turiasaurus, it certainly doesn't seem to be over twice the mass)
Which one would be the right weight in a range with a 50% uncertainty? The answer is that a single measurement isn't enough to determine it and the uncertainty is just way too big to pick the median value, as Paleo-king said, that's why you have estudies like Mazzeta 2004 which plot values obtained with allometry equations based on several bones. Most of the mass of a dinosaur, is centered in the torso. The dimensions of the torso aren't taken into account with allometry methods; and here is where vollumetric methods shine and have the advantage because provided a model is based on a complete enough animal with a meticulous restoration, the error bars of the method are much smaller, less than 20%, and if you anylize it with a computational model it is lower than that. Most of that error would rely on how the ribcage is restored and in the ammounts of soft tissue, but putting this as an example (journals.plos.org/plosone/arti…) it appears clear that not all soft tissue levels are equally likely, and a variation of approximately 8% from the best fit model should be enough to cover all of the most likely estimations unless the skeleton the model is based on is severely flawed. journals.plos.org/plosone/arti…
The correct conclussion addressing this is that the allometry based equations remain approximate; and with wider error bars than meticulous vollumetric methods, not accurate nor certain. The reason they apply correctly to modern day extant fauna is becase they have been modified and derived from regressions based on those very same animals of which the values have been ploted. The allometry based methods can be used as a beliavility test for the vollumetric ones and also can be used when the animal being analized is very poorly known and a vollumetric estimation is not possible. In chase that both are possible, the vollumetric method should be favored as defended by some experts like Asier Larramendi (2015). Femur circumference specifically can prove very useful when scaling from a vollumetric model of a closely related animal with a known femur, since the error margin gets greatly reduced and the discrepancy between limb circumferences cubed is still much better to scale mass from another, more complete animal than let's say the proportional differences in the height of a vertebrae cubed.
I don't see any point of this discussion in which it seemed like didn't understand what you were saying so please try to be respectful. I understand that you disagree with some of the estimations here as do I but that doesn't invalidate the fact that he has made a lof of good points, although you made good points as well, all this in my humble opinion.
I disagree with a Dreadnoughtus much over 40 t considering what we have , and it is very likely that it wasn't this extremely far from Turiasaurus. I think it is also likely that Turiasaurus was a bit over 30 t since I believe Asier's estimations are a bit in the low end of the plausible estimations, reading his latest book about theropod dinosaurs. They are still much better than the estimations that any allometric method yield for theropods, they are extremely bad for them because of the bipedal gait; even when they tried to fix this in Campione 2014, the method, although better, still yields some fails). And yes limb allometry usually overestimates the size of the sauropods because most modern day taxa has pathetically thin limb bones compared to them, sauropods have very big limb circumferences in part because the bones have a much lower global compactness index than in mammals so they develop proportionally wider circumferences than needed because the tissue is distrubuted differently due to the limb bones being heavily pneumatized.
Thank you for your thoughtful response. You’ve raised some important points.
First of all, the reason why stylopodial circumferences are used is because, due to the simple fact of the way sauropods stand, nearly the entire body weight needs to be supported by the narrowest parts of the humeri and femora. Therefore, the body mass and stylopodial circumferences are closely correlated and the only question is what the margin of error on that correlation is. Speaking of margin of error, the odds that the equation simultaneously vastly underestimates the mass of Paralititan and vastly overestimates the mass of Turiasaurus are quite low (and it’s equally likely that the equations could be dramatically overestimating the mass of Paralititan and underestimating that of Turiasaurus, based on the equation alone).
The paper Wedel cites as showing that allometric methods are inaccurate doesn’t appear to test models that exclusively use stylopodial circumferences. Stylopodial circumferences correlate well with body mass across a range of taxa, whereas humerus and femur length show significant variation in their correlation with body mass.
There is still substantial variation in body masses calculated from volumetric techniques, even in nearly completely-known individuals. In the case of Dreadnoughtus, it is based solely on different convex hull volumes, without even considering density, but especially in sauropods, body density is highly uncertain. Pneumatic volumes alone can account for 20% or more variation in body mass. Once you combine the triple challenges of correctly reconstructing the skeleton, correctly estimating soft tissue volumes, and correctly estimating variable densities, I have no doubt that the range of error is quite large for volumetric techniques as well.
The result of any body mass estimate is only as good as your assumptions, and volumetric methods include far more assumptions, and therefore far more potential sources of error, than allometric methods. With allometric methods calculated from both limb bones, the only source of error is the prediction interval of the regression equation (which can be quantified), whereas potential sources of error involve dozens of small differences in reconstruction, such as the angle of ribs or size of air sacs, that add up to a large and difficult-to-quantify amount of error. With a little stats work, one could easily show whether Paralititan or Turiasaurus is significantly bigger than the other based on regression equations.
This is not to say that volumetric methods are wrong and allometric methods are infallible, but rather that allometric methods are “good enough” relative comparisons in a subject where definite answers are nearly impossible to find. Until somebody does a volumetric study of reconstructions of both Paralititan and Turiasaurus, with carefully quantified estimates of pneumatic volume and tissue distribution, and shows that Paralititan cannot realistically be smaller than Turiasaurus, our best guess is that Turiasaurus was larger. There are reasons why a simple GDI as paleo-king proposes is insufficient to overturn the estimates from the bone thicknesses. Simply performing GDI with a density of 0.9 for the body dismisses the likely substantial difference in pneumaticity between the two species, for example.
I’m fairly certain that sauropod limb bones are not pneumatized, and so wouldn’t have “wider circumferences than needed” for that reason. As far as I know, sauropod limb bones haven’t been shown to have significantly different mechanical properties than those of extant quadrupeds, so there’s no good reason why the equation would consistently overestimate their masses.
As far as your circular reasoning comment goes—I decided to see what happens if you use an equation derived from the herps on the mammals, and vice versa. The mammal equation, on average, finds body masses 82% the actual masses in amphibians and reptiles, whereas the amphibian+reptile equation, on average, finds body masses 134% the actual masses of mammals. Not great, I admit, but less than the factor of two Bates et al. found for their volumetric calculation.
"Volumetric methods in general, and GDI in particular, have one more huge advantage over allometric methods: they’re more precise and more accurate. In the only published study that compares the accuracy of various methods on extant animals of known mass, Hurlburt (1999) found that GDI estimates were sometimes off by as much as 20%, but that allometric estimates were much worse, with several off by 90-100% and one off by more than 800%. GDI estimates were not only closer to the right answers, they also varied much less than allometric methods. On one hand, this is good news for GDI afficionados, since it is the cheapest and easiest of all the mass estimation methods out there. On the other hand, it should give us pause that on samples of known mass, the best available method can still be off by as much as a fifth even when working with complete bodies, including the flesh. We should account for every source of error that we can, and still treat our results with appropriate skepticism."
- Henderson, D.M. 1999. Estimating the mass and centers of mass of extinct animals by 3D mathematical slicing. Paleobiology 25:88-106.
Not inaccurate enough, however, that it's likely that Turiasaurus was less than 30 tons and Paralititan was almost 70. For the type specimens to have such a large difference in mass would require Paralititan and Turiasaurus to both be on the extreme fringes of the equation in different directions. The general biomechanics of sauropod limb bones should remain closely similar in the two taxa, so they would not be as different as, say, a crocodile and an antelope, and it seems unlikely that the equation would be off by such a large percentage in opposite directions (elephant and antelope have a difference of like 30% in the scaling equation, this would require a difference of roughly 150%).
If you could show me a properly-done GDI of both (with provided multi-view skeletals) which showed a dramatic difference I might believe it; until such a time I will think it's likely that Turiasaurus and Paralititan were of similar size.
Or you could just use your method of blind maths without any inference of the thousands of bone sections OTHER than a humerus midshaft that affect soft tissue volume, and end up with laughable things like Dale Russell's 15-ton Giraffatitan. And it's a well-known fact that basal sauropods were built like Chartres and more derived ones far more like Amiens or Rheims.
Also, the equation gives an estimate of roughly 35 tonnes for HMN SII, not 15. In 1985 Russell was saying 30. What's the reference for 15, again?
Jacobs, LE. (ed.). Aspects of Vertebrate History. Flagstaff:
Museum of Northern Arizona Press. p. 257-268.
Note that Asier Larramendi did a skeletal of T. riodevensis based on the measurements in the paper, & got a total length of only 21m for it:
Yes "A." giganteus is probably not as large as some estimates claimed. Roughly 25m or so giving it Bonitasaura-like antarctosaurid proportions.
I'm not sure about Tornieria, given that it's published measurements don't suggest a length greater than Barosaurus or Diplodocus, & using Supersaurus proportions (as per Tschopp, et. al.'s analysis), it should be smaller than both AMNH 6341 or CM 84.
Length estimates don't mean much with sauropods, given how much the neck and tail affect them. It's best to compare mass estimates derived by the same methods when comparing sauropod body sizes.
It's weird that A. giganteus is so small, though, given that it has the longest described femur of any animal. It must have been the indricothere to Dreadnoughtus's mammoth.
Still I wouldn't consider 65 tons to be "small" by any means. It's heavier than any of the Giraffatitan specimens likely were (another long-limbed animal).
As for length... when you actually consider how much of the length is neck and tail, the 21m length of Turiasaurus looks even less impressive. And if Larramendi did a top-view of Turiasaurus, I would expect the rib cage to be maybe HALF the width of that of most derived titanosaurs when you scale them to the same length.
Mass estimates "using the same method" in turn are only as reliable as that method itself. Limb-bone circumference is a horribly unreliable method. Just assuming a mass based on a single section of a single bone, reminds me of the old tale of the blind men and the elephant, each trying to guess its true size and form based on just one body part, and consistently getting it wrong. And even the blind men had a living, fleshed out animal. Midshaft circumference doesn't even take into account the muscle crest sizes of the ends of the singular limb bone in question, let alone the likely extent of flesh on all the other body parts. (see below for more details). A much better method is a volumetric GDI from a good multiview skeletal, or even the old-school "make a model and dunk it in a tank" method with the neck density estimated separately from the rest. At least that way you're not assuming mass based on a single cross-section of a single bone and telling the rest of the animal's body and its thousands of other dimensional variables to bugger off and not be seen.
Also queries are on Barosaurus (I do believe that BYU 9024 was a Barosaurus specimen due to their similarity more than to Supersaurus), and on the footprints from Bolivia.
Giraffatitan's foot size.... depends on the specimen you're talking about. For HMN SII is is 75cm, for HMN XV2 (possible "adult" specimen) it is 85cm. Still smaller than Breviparopus.
For another comparison, the biggest Paluxy River prints (which may be Sauroposeidon) are 87cm wide on average.
These figures are from Paul 1988, which estimates foot pad sizes based on elephant foot pads as well. 85cm for XV2 vs. 88cm (SV-POW estimate, if indeed it is for XV2) is not that big of a difference. I would go with the smaller figure since brachiosaur hindfeet are more conservatively proportioned than those of diplodocoids and most other sauropods. (i.e. even a downright HUGE brachiosaur may have hindfeet that are no bigger, and possibly even a bit smaller, than those of an average-sized diplodocoid). Despite Paul skimping out on some soft tissues, he does keep the foot pads at a reasonable size, at least width-wise. Also you can reverse-scale the size of a Giraffatitan foot pad by comparing its foot bones - to - shin bones ratios, and cross-scaling with prints that preserve good toe prints like those at Paluxy (not true brachiosaurids but at least close), and then use the foot-to-shin ratios to scale it to XV2 which is known from the fibula but not the foot. Again you get similar figures to Paul. There's no radical change in foot design between brachiosaurs and chubutisaurs (there is a 4th ungual in chubutisaurs apparently, but that doesn't change any of the proportions of the foot overall) so you're still good with 85cm for XV2.
Whatever Breviparopus is, it's huge. Not a record-breaker, but still huge.
Brick said that you and him are working on a "B.nougeredi" report as well for an analysis there.
Ultrasauripus, like many footprint taxa, is very obscure. Many footprints are eroded, I've never even SEEN pictures of the prints for Ultrasauripus, so you can see how messy it is to guess the sizes of the animals. And even then, not all brachiosaurs had identical proportions. Even my proportions for Fusuisaurus are highly speculative since so little of it is known.
But I do agree there is a possibility it could be more like Fusuisaurus than other brachiosaurs due to its location. But Fusuisaurus is already very Brachiosaurus-like. In fact its remains resemble Brachiosaurus far more than any of the "Giraffatitanine" or "Pleurocoeline" (cretaceous USA) forms, or for that matter the "angelfish" forms. The problem is that it actually appears MORE basal than Brachiosaurus based on the caudal neural spines and the lack of rib pneumatization... so no telling if it looked more like a giant Atlasaurus than a giant Brachiosaurus (the lack of neck material doesn't help either), but I went with the latter because based on other sauropod families, any macronarian pushing 100 feet is more than likely going to be feeding VERY high up, and be VERY long-necked.
The Bolivian footprints are Aptian and likely belong to a Somphospondylid-same with Broome.
Brick mentioned that the closest to "B.nougeredi" we have to compare to sacrum-wise was the Archbishop, and even then it's a stretch, so whatever it is will have very unusual proportions.
Nearly all the species on this list are FAR more different from one another than what you get with Apatosaurus and Brontosaurus. There really isn't a valid comparison there. And Apatosaurus and Brontosaurus were NEVER considered to be the same species, merely the same genus. There were at least three species assigned to it (ajax, louisae, excelsus).
However, in the past few years, some really good scientific papers have come out that re-examined all the material in "Apatosaurus" and "Brontosaurus" and ended up splitting them apart again. It is now clear Apatosaurus consists of just the type species, Apatosaurus ajax, which has longer arms and neck, and smaller hips than the others. The shorter "Apatosaurus" excelsus and the super-chunky "Apatosaurus" louisae are both now firmly classed as Brontosaurus. The other less well-known "Apatosaurus" species like parvus and yahnapin are also now grouped into Brontosaurus. The thing is, Dr. Bakker was already splitting them decades ago back when most other paleontologists assumed they knew all there was to know about Apatosaurus. The evidence is pretty clear, you just have to look at it closer.
There are many differences in the vertebrae of different species apart from size. Their shapes are extremely different in many cases. For example nobody would ever confuse a brachiosaur dorsal vertebra for a diplodocid one, if you laid the two side by side. Even within families there are distinct differences, for example Puertasaurus and Futalognkosaurus are closely related to each other, but the shape of the neural spines on their neck vertebrae is very different, and the positions of the laminae also differ. Not to mention they come from different regions and time periods.
If any of the species listed here could actually be lumped together with another, I would have made that clear. But generally you don't see erroneous splitting in most of these new discoveries (most of these dinosaurs are pretty recent discoveries, and the scientists are a lot better at telling species apart now than they were 100 years ago when Apatosaurus and Brontosaurus got mixed up.) I read all the papers and I agree with the authors on at least this much, all of these animals are distinct species. There's a lot you can tell from a species even based on a single bone, as long as it's well-preserved enough to be diagnostic (i.e. show a unique set of characters that differs from other dinosaurs). Puertasaurus is known from only 4 bones, only 2 of them have been published, but even those 2 bones have all the data needed to prove it is a unique genus, even when comparing it to other lognkosaurian titanosaurs.