|
Post by Infinity Blade on Sept 4, 2024 7:24:23 GMT 5
Reach not being an enormous factor (at least depending on the animal) is something I hadn't considered before, that's an interesting example with the albatross and petrel. Maybe this counts too? A mink attacking a heron, where the bird most certainly has vastly more reach. Despite this, the mink is able to leap on the heron's back and attempt to bite it. Then again, it's possible this was an ambush, and the heron did manage to shake it off. www.thesun.co.uk/news/16179340/mink-sinks-teeth-into-herons-neck-bird-fights-back/
|
|
|
Post by Supercommunist on Sept 4, 2024 8:21:34 GMT 5
Reach not being an enormous factor (at least depending on the animal) is something I hadn't considered before, that's an interesting example with the albatross and petrel. Maybe this counts too? A mink attacking a heron, where the bird most certainly has vastly more reach. Despite this, the mink is able to leap on the heron's back and attempt to bite it. Then again, it's possible this was an ambush, and the heron did manage to shake it off. www.thesun.co.uk/news/16179340/mink-sinks-teeth-into-herons-neck-bird-fights-back/I think the issue with reach is that is require a lot of thinking and control. Humans can make use of its since we are intelligent, but an animal is less likely to figure out exactly where it can hit without getting hit in return. Human also have very short snouts, so our field of view isn't obscured by a muzzle which probably makes it easier for us to gauge our reach. I think storks/herons/egrets are the most obvious example but I tried not to use them as they aren't really the most impressive fighters so them getting bum rushed by carnivoras can be attributed to them being worse at fighting and being less competent on the ground. Fishing birds with long beaks can sometimes effectively space out other animals with their long reach but they are an outlier since they are basically designed for fast, jabs. I think a more relevant example would be spotted hyenas and AWDS. Spotted hyenas have much larger necks and reach than AWDs, but they don't seem to be intiate bites more effectively than the shorter necked AWDS. Obviously this is not a perfect example either, since their is a large size difference and the AWD's greater agility is a much more important factor than reach, but I don't see any evidence that a hyena's longer neck would help it get the first bite on a similar sized canine. The long neck is great for attacking back at odd angles and covering its flanks, but it doen't really seem to helo score "first blood". I think something like a terror bird could be able to use its reach effectively but its neck be better adapted to dealing with shorter quadrupedal animals. On another note, I think its worth pointing out that gigantosaurus has thinner jaws. Some while tyrannosaurus has a smaller gape, that's less of an issue if it manages to grasp the top of the giga's snout. EDIT: Also back to the reach discussion, IB do you know if there is any correlation between horn length and fighting ability? From what I understand, horn size is sexually selected because it is attractive, but longers horns generally aren't superior weapons.
|
|
|
Post by Supercommunist on Sept 4, 2024 9:07:54 GMT 5
Separate post because I am tired of editing the last one. I was thinking about the "lighter neck speed argument" and I was wondering how big of an advantage that is. I am not trying to trivialize every advantage giga has but it's hard for me to decide whether it is a noticeable advantage, unless we have some way to quantify it. I was looking at videos of shoebill storks. On paper, their larger bills and thicker necks should reduce the speed of their strikes compared to other storks or herons, but honestly I can't really see much of a difference, if at all. I am inclined to think that if we hyper analyzed the strike speed of a shoebill compared to a thinner billed, longer necked bird we would a difference in speed, or at least energy cost but I am not sure how much of a difference it makes in a fight to the death between similar sized foes. Unless the giga gets the first strike and quickly cripples the rex or engulfs the rex's jaws with its own, I think the rex would just bite it back a second later. Edit: Also wasn't 100 percent sure if a yellow nosed albatross' beak and neck was longer than a giant petrels but based on a quick google search it looks like their heads are longer: Albatross skull length: skullsite.com/skullpage/thalassarche-chlororhynchos-atlantic-yellow-nosed-albatross/Southern giant petrel skullsite.com/skullpage/thalassarche-chlororhynchos-atlantic-yellow-nosed-albatross/
|
|
|
Post by brobear on Sept 4, 2024 16:22:54 GMT 5
Coelurosauria appear to be more "aimed" at warm bloodedness considering so many proven to have had feathers. T-rex had Stereoscopic vision. A bigger brain. Stronger jaws. Where do you see religion. Have you discovered the "Book of Tyrannosaurus"? Well, in science, it is good practice to cite your sources and use precise language. "Warm-bloodedness" is a vague term, for example. "Homeothermy", "tachymetabolism", and "endothermy" are more specific. Also, when making statements like " T. rex had stronger jaws", it's best to provide numbers with references, as different studies can, depending on their methods, get different results. This is not to say that your statements are wrong. theropod's point is just that discussions are more constructive when we are clear and concise. The estimated bite force of Giganotosaurus is around 24,977 Newtons, or 2.5 metric tons. This estimate was made in a 2021 study by A. J. Rowe and Snively. ___________________________________________________ Adult T. rex: A bite force of around 35,000 newtons is commonly estimated for an adult T. rex. The strong bone-crushing bite force of T-rex should not have to be proven each and every time the subject comes up. Precise numbers are great, but so is a little common sense.
|
|
|
Post by theropod on Sept 4, 2024 17:01:51 GMT 5
You mean the "Carnosaurs had few predatory specializations"-guy? lmao Still don't know how that one made peer review, geez Twice
|
|
|
Post by theropod on Sept 4, 2024 18:31:47 GMT 5
One thing I want to bring up is that if the Giganotosaurus holotype was asymptotic, because as it has an efs ( if it is not asymptotic please correct ), then it wouldn't really be fair to compare it to the average of all Tyrannosaurus specimens but only asymptotic ones, of which there are about 7, iirc, not sure how much the results would differ from that but I suspect the difference would be a lot less. When and where did someone find the Giganotosaurus holotype to have an EFS? I’m not aware of any histological work having ever been done on it. That’s a main reason why it is a major assumption to presuppose that (unlike the vast majority of theropod specimens in the fossil record) it represents what the online paleocommunity calls an "asymptotic adult". There aren’t even any photos of the fossil material that I’m aware of that are good enough to get an indication (depending on the preservation, you can sometimes tentatively tell if there is an EFS if there is characteristic superficial flaking-off of the very thin surface layers of the bone an EFS entails, but it’s obviously no replacement for actual histological sections). It is better to use the two specimens we do have to get an average than to make one up out of thin air. I already listed the p-value and acknowledged that it is not significant, but a 75% chance that Giganotosaurus was larger is enough for me to assume that it probably was, rather than go with the 25% chance that it wasn’t. I go into more detail on this thread: The question is, what’s the alternative. Sure, you can argue for a null of no difference and fail to reject that null, but see the discussion of statistical power in the thread I linked above. The question then, having a very low chance of significantly rejecting "no difference", becomes if "no difference" is so overwhelmingly probable that it deserves being treated as the null hypothesis. Usually when discussing the sizes of extinct animals nobody applies such stringent standards. We could, for example, not determine that Argentinosaurus was larger than Meraxes (or, if you want an even more ridiculous example, Epidexipteryx) if we required proof of statistically significant differences in size, because there’s just one specimen of each of them, meaning we are unable to do a significance test. Should that mean we must assume they were the same size? Clearly not, that assumption simply is no suitable null hypothesis in this situation. In practice, people generally accept that it is sufficient to demonstrate a tendency, i.e. demonstrate, on the basis of the best available size estimates, which taxon is most likely to be the largest. Often enough, people disregard even that, and any statistical considerations whatsoever, in favour of just looking at which one they can find the largest individual of (that is the approach that what I would guess is probably the majority of the online paleo community uses as a basis for celebrating T. rex as the largest theropod). And (this is by no means intended as criticism directed at you, it’s merely a general issue I keep observing) these are often the very same people who simultaneously complain that averages supposedly can’t be compared (talk about a double standard) because one or both have insufficient sample sizes, ignoring that is precisely why one should compare averages, which are uniquely robust to biases caused by sample sizes. We can assume that, but then we should apply the same standard to other, even more marginal, advantages some people see and like to throw around. For example the brain-size difference between a similar-sized T. rex and Giganotosaurus→ was less than the relative range of brain sizes within the species T. rex, or the species Homo sapiens for that matter. So how come the former is "smart" and "advanced" whereas the latter is "dumb" and "primitive"? On a different note, since Snively et al. 2019 was already brought up here regarding agility, there’s an issue I have with a detail of that study; Focusing on the axial segments (for simplicity, it makes little difference), for T. rex FMNH PR 2081, they estimate a mass of 9 130.87 kg and a rotational inertia of the axial segment of 28 847 kgm². For Giganotosaurus MUCPv-Ch1, they estimate a mass of 6,907.6 kg and an axial rotational inertia of 35 821 kgm², both based on Hartman’s skeletal reconstructions (but with a slightly higher mass due to superelliptical cross-sections). This is, frankly, entirely implausible. The two specimens are almost exactly the same length (12.4 vs 12.3 m), but the T. rex is heavier by almost a third. Rotational inertia scales with the second power of radius of rotation (i.e. distance from the center, proportional to body length) and the first power of mass, but even so, assuming the mass is even anywhere near similarly distributed along the length of the animal (which is the case here, they both have broadly comparable body shapes, one is simply more "stretched-out"), there’s no way the Giganotosaurus holotype could have a higher rotational inertia than sue, let alone by that much. Consider that the squared ratio of body lengths is (12.4/12.3)^2=1.016, so this would amount to about a 1.6% difference in rotational inertia, nowhere near enough to offset that the tyrannosaur specimen in this comparison is >30% heavier (1.31*1/1.016=1.29, i.e. we’d expect the tyrannosaur should have an inertia about 29% greater assuming equal mass distribution along the length). I’m not resorting to simplistic calculations here due to being too lazy to do the actual maths, just to illustrate the point more intuitively. If you want the precise figures: I can roughly replicate Snively et al.’s rotational inertia figure for Sue based on GDI-ing→ Hartman’s skeletal (I get 29 624.418 - 29 866.887 kgm² at an axial mass of 8754.846, very close to Snively et al.’s result of 28 847 kgm² ), but am ending up with a much lower figure of "only" around 22 000 kgm² (21584.902 to 21723.954, depending on whether I assume elliptical vs rectangular cross-sections, at an axial mass of 6897.687 kg) for MUCPv-Ch1. That is still slightly simplistic because I ignored pneumaticity here and assumed a unit density, but this doesn’t result in any biases (both would have had pneumatic structures, especially in the anterior part of the body, which would somewhat lower these inertia figures and explains why my figure for sue is slightly higher than Snively et al.’s). The figures also match up very closely with what would be expected based on length and mass, with the tyrannosaur having an inertia about a third higher than Giganotosaurus (just a little more than would be predicted assuming identical mass distributions, likely because T. rex has a little more of its mass concentrated near the anterior end of its body axis due to the heavy skull and neck). So I think it is safe to say that the figures do not match up, and that MUCPv-Ch1 would have had a considerably lower rotational inertia than Sue, not higher as Snively et al. estimated. I have provided the script and silhouettes here, if anyone wants to replicate or modify this analysis: rotI.zip (23.54 KB) Should you still be unconvinced, here is a simple visual aid. Both silhouettes are to scale and their respective center of mass (i.e. of rotation) is marked: Does it look plausible for the blue/lower model to have a rotational inertia as high as, let alone 20% higher than, the red/top model? My guess is that there must have been a mix-up somewhere, which is further corroborated by the fact that the full-body rotational inertia for Giganotosaurus, as given in the same table, is much lower than the axial one (though still higher than I think is realistic), when it should actually be marginally higher (as it includes the limbs). If we translate the implications of all this into the "agility index" as defined by Snively et al (lateral Ilium area/rotational inertia), that shrinks down what their estimates suggest to be a ~70% difference to more of a ~40% difference, although it should be noted that this index doesn’t account for the entire hindlimb torque, as the most important leg retractor, the caudofemoralis, is not included (neither are differences in the limb muscle moment arms). That being said I am not disputing that T. rex would have had lower rotational inertia than a carnosaur of equal mass (on account of being shorter), but certainly not at equal length (a scenario in which T. rex would be by far the heavier animal, and thus necessarily have the higher rotational inertia). Nor am I disputing that tyrannosaurids would have likely had better turning performance than carnosaurs at similar masses, although how big this difference is, and how relevant, requires some further considerations (e.g. about axial flexibility). Keep in mind that by the same metric Snively et al.’s estimated, a dog or bear would be considered to be much more agile than a felid of similar mass (due to more compact body shape resulting in lower rotational inertia. I can attempt to quantify that suspicion if there’s sufficient interest). I’m not so sure if this picture matches up with real-world perceptions of how agile these respective animals are, let alone how they would do in a fight. What I also don’t see discussed nearly enough compared to this is that, while tyrannosaurids would have had better turning ability than carcharodontosaurs, the latter would have had better stability and sidestepping ability (something that would not necessitate turning the entire body) due to their dorsally angled femoral heads giving their legs a wider transverse range of motion. Ultimately this is just a classic case of locomotory differences between a more cursorially adapted vs a less cursorially adapted animal. But whereas most of the time people seem to agree that cursoriality is of limited or no use in direct confrontations (cheetah vs leopard, wolf vs cougar, etc.), the doctrine of Tyrannosaurus exceptionalism has convinced people that this specific case is the exception to that and blown the study results in question out of proportion in the ongoing quest to list all the ways in which T. rex is "superior" to other theropods. -- Hartman, S.A. 2013. Mass estimates: North vs South redux. Dr. Scott Hartman’s Skeletal Drawing.com. Downloaded from www.skeletaldrawing.com/home/mass-estimates-north-vs-south-redux772013 on 16 August 2023. Snively, E., O’Brien, H., Henderson, D.M., Mallison, H., Surring, L.A., Burns, M.E., Jr, T.R.H., Russell, A.P., Witmer, L.M., Currie, P.J., Hartman, S.A. and Cotton, J.R. 2019. Lower rotational inertia and larger leg muscles indicate more rapid turns in tyrannosaurids than in other large theropods. PeerJ 7: e6432. Attachments:
|
|
|
Post by dinosauria101 on Sept 4, 2024 19:03:55 GMT 5
If you ask me, the more I think about it, the more these 'same size assumptions' come across to me as either conscious or subconscious double standards to always favor T. rex that you were talking about yesterday, and ultimately a cop-out (intentional or not) to avoid addressing the probable substantial size advantage for Giganotosaurus. Same goes for the brain size thing I don't know why I didn't notice that sooner. Something else that might reduce the relative discrepancy between rotational inertia of T. rex and Giganotosaurus is that the Giganotosaurus holotype would be even heavier relative to its length than it was in the study if we follow the estimate from Spinoinwonderland. The increase from the 6.9 ton to the 8.2 ton estimate would be a big reduction already, and then the size increasing factors that were talked about in the comments would presumably place it right within the range of Sue size estimates (to my knowledge 8.4 tons to 9.1 tons). With a Giganotosaurus holotype of similar length AND weight to Sue, might it be reasonable to assume that that would sufficiently reduce the agility difference for the T. rex not to have a worthwhile advantage? Glad to see this mentioned by someone else.
|
|
|
Post by theropod on Sept 4, 2024 19:23:56 GMT 5
I think the issue with reach is that is require a lot of thinking and control. Humans can make use of its since we are intelligent, but an animal is less likely to figure out exactly where it can hit without getting hit in return. Human also have very short snouts, so our field of view isn't obscured by a muzzle which probably makes it easier for us to gauge our reach. Perhaps, but then the same could be said for dodging or outmaneuvering, which is essentially contingent on the same abilities. Something else that might reduce the relative discrepancy between rotational inertia of T. rex and Giganotosaurus is that the Giganotosaurus holotype would be even heavier relative to its length than it was in the study if we follow the estimate from Spinoinwonderland. The increase from the 6.9 ton to the 8.2 ton estimate would be a big reduction already, and then the size increasing factors that were talked about in the comments would presumably place it right within the range of Sue size estimates (to my knowledge 8.4 tons to 9.1 tons). With a Giganotosaurus holotype of similar length AND weight to Sue, might it be reasonable to assume that that would sufficiently reduce the agility difference for the T. rex not to have a worthwhile advantage? The error in this line of thought is that making something more massive won’t increase the agility, it will decrease it, because when you put more mass on something, all else being equal, the rotational inertia will go up, not down. What it would indeed do would be to decrease rotational inertia for a given mass, since increasing the mass decreases the length at mass parity, therefore resulting in a more favourable mass distribution to minimize rotational inertia. But that’s not really relevant here, since the mass wouldn’t be constant, it would go up in such a scenario. However, the impact on turning performance ultimately depends on not just the rotational inertia, but also the strength and mechanical advantage of the limb muscles, which determines how much torque the animal can excert on its mass to turn it around (the more torque, the higher the turning acceleration). As I wrote, the "agility index" paints an incomplete picture in this regard, since it doesn’t take into account the important caudofemoral musculature (or its moment arms) and is exclusively basesd on the area of the ilium as a proxy for limb muscle size (which is naturally a metric that is biased on favor of the tyrannosaur, as a taxon that has a more "bird-like", knee-driven locomotor mode than non-coelurosaurian theropods, and therefore needed more thigh musculature). Making the animal heavier of course could imply larger thigh and tail muscles, hence more torque, hence offsetting (part of) the increased rotational inertia. However, due to how force (proportional to the 0.67th power of mass) vs inertia (proportional to mass, given that dimensions stay unchanged) scaling works, it is unlikely this will fully compensate for the increase in mass, let alone overcompensate (therefore resulting in greater agility). I.e., unless the mass increase is disproportionately concentrated in the musculature used for turning, this pretty much always means that if you make something bigger and heavier, it will become less agile. That is also why the larger you make an animal, the more of its musculature needs to be devoted to locomotion in order to maintain the same level of athleticism. Scaling simply works against animals in this regard (that’s also why the agility indices drop off so much the larger the animals get). In any case, all that would apply to both taxa equally, and the proxy used by Snively et al. (and therefore the agility index per their definition) was not based on estimated muscle volumes or cross-sections, but on iliac area as a simple, unchanging skeletal measurement. If you increase the mass, you’ll simply increase the rotational inertia (by the same percentage as the mass, given that the distribution stays the same), therefore decreasing agility. That being said, the estimates you discuss are likely based on higher density or other assumptions that would also make T. rex heavier (and increase both taxa’s rotational inertia relative to a more pneumatized model). I know it has recently become fashionable to estimate Sue and Scotty, as well as MUCPv-95, at over 10 tons, as opposed to the 8-9 t back in the 2010s.
|
|
|
Post by dinosauria101 on Sept 4, 2024 19:55:21 GMT 5
Something else that might reduce the relative discrepancy between rotational inertia of T. rex and Giganotosaurus is that the Giganotosaurus holotype would be even heavier relative to its length than it was in the study if we follow the estimate from Spinoinwonderland. The increase from the 6.9 ton to the 8.2 ton estimate would be a big reduction already, and then the size increasing factors that were talked about in the comments would presumably place it right within the range of Sue size estimates (to my knowledge 8.4 tons to 9.1 tons). With a Giganotosaurus holotype of similar length AND weight to Sue, might it be reasonable to assume that that would sufficiently reduce the agility difference for the T. rex not to have a worthwhile advantage? That being said, the estimates you discuss are likely based on higher density or other assumptions that would also make T. rex heavier (and increase both taxa’s rotational inertia relative to a more pneumatized model). I know it has recently become fashionable to estimate Sue and Scotty, as well as MUCPv-95, at over 10 tons, as opposed to the 8-9 t back in the 2010s. Thank you for the explanation on mass/length/inertia. As for the part about the estimates vs assumptions, if it helps, I did look a little more into it and Spinoinwonderland said on the Discord that he estimates Sue at 9.4 tons (2023 and 2024). This is presumably apples to apples with his Giganotosaurus for density and other assumptions since they are from the same author.
|
|
|
Post by Infinity Blade on Sept 5, 2024 5:56:47 GMT 5
EDIT: Also back to the reach discussion, IB do you know if there is any correlation between horn length and fighting ability? From what I understand, horn size is sexually selected because it is attractive, but longers horns generally aren't superior weapons. Apologies that it took me a while to get to this, I've been at work or school most of the day. Also hope it's okay for me to post this here, given that it's only tangentially related to the topic at hand. So, this isn't for vertebrates, but a 2014 study found that male Japanese rhinoceros beetles sometimes fight vigorously enough to break their horns, and by measuring safety factors for the entire range of horn sizes, it was found that safety factors decrease with horn length, and that the longest horns are at the highest risk of breaking. This suggests that there is, in fact, a mechanical limit to horn length ( McCullough, 2014). Mind you, rhinoceros beetle horns are made of chitin, which is, weight for weight, stronger than bone (which vertebrate horns are partly made of) and even steel ( link->). " Chitin forms extremely strong fibrous microfibrils that are “stronger, weight-for-weight, than bone or steel” (Lenardon et al., 2010)." If that's the case, there definitely has to be a mechanical upper limit to horn length in vertebrates that use them in combat. I've posted info before about how elephant biologists believe that super long tusks are unlikely to be good weapons. Also, in bovids, it's the short, dagger-like horns that are specialized for potentially lethal stabbing behavior, whereas long horns with long reach are more for wrestling and fencing ( Lundrigan, 1996). Obviously a long horn could still stab too if enough force was put behind it without being enough to snap it in half (you posted accounts of gemsbok skewering lions IIRC), but stabbing capacity isn't so much a function of weapon length as it is other functions (like hitting your foe like you're armed with a quarterstaff, or some measure of wrestling). After all, a horn can be as long as it wants, but if it misses vital regions, the foe can survive. A short horn that still punctures deeply enough to damage vital organs all the same will be lethal.
|
|
|
Post by dinosauria101 on Sept 5, 2024 18:41:37 GMT 5
Just noticed some edits to these comments. I think I might have a thing or two to add about them. On the matter of size, I asked Spinoinwonderland what he thinks would be a reasonable upsize from his 8.2 ton Giganotosaurus holotype estimate with the anatomical corrections and he hopefully will get back to me soon. In the meantime, moving forward from my deliberately minimalistic initial comments using only the holotype for comparison (just to demonstrate that even with only it, Giganotosaurus is estimated to be markedly larger), I will factor in MUCPv-95 to be more fair to Giganotosaurus. Spinoinwonderland said (if I remember right) that it is around 7% larger, so that would give us 10 tons for it, making the average of those 2 specimens 9.1 tons and therefore around 40% larger than the estimated 6-7 tons average for T. rex. This discrepancy would get even larger with Giganotosaurus' anatomical corrections, if these estimates are accurate then that reinforces my earlier verdict that Giganotosaurus wins a large majority on account of being much bigger. I imagine knowing this would change a lot of people's opinions on the relevance of the metrics that study used. Especially considering that, as you wrote, another way the study is biased towards tyrannosaurids is by using a method that favors their method of locomotion. Although I replied to the top statement yesterday, I think these 2 statements prove this statement you made on 'Animal fighter tier list'. The turning adaptations of T. rex are blown out of proportion to make it seem "superior", while we don't even have a study - let alone a comparative one - on the superior sidestepping ability of Giganotosaurus. I am inclined to think that thanks to all the hype and bias towards T. rex causing this discrepancy, Giganotosaurus might have some unstudied advantages over T. rex as well. Therefore, if we go off of comparative studies to determine advantages, it is likely to be misleading overall.
Additionally, while reading 'Animal fighter tier list', I came across another statement you made that Supercommunist might want to know about regarding face biting. Going by what tiger sharks can do to sea turtles, it might be possible for Giganotosaurus to cause severe damage to Tyrannosaurus' skull bones. And there is more: on the Discord server I was talking about, it has been mentioned that the force of the neck muscles would drastically increase the bite force of Giganotosaurus, such as if it were to bite down while also pressing down. There's also the matter of facial nerves (considering how tactile tyrannosaurid faces were it is likely they would be more vulnerable to this than your examples such as moray eels): if those nerves are severed, they might not render a T. rex physically unable to fight but it seems plausible that the shock from the pain would incapacitate it.
Even if we were to disregard all of that, would you still favor T. rex at a size disadvantage of over 40 percent?
|
|
|
Post by Supercommunist on Sept 5, 2024 20:50:55 GMT 5
Doging is a pretty basic concept that doesn't require advanced thinking. For instance, flies are great at avoiding fly swatters and they have almost no combat ability to speak off. To be fair, I probably downplayed an animals ability to understand the concept. I am sure that predators do have an idea of where they can safely approach a prey item or when to expect a quick, but this most likely is a practiced or instinctive skill. Gigantosaurus and tyrannosaurus never met so they wouldn't be able to identify a reach difference and trying to take advantage of that probably wouldn't occur to a gigantosaurus. Reach is an important factor in human fights, but humans measure their reach and pros specifically train to deal with each opponent. Average human's don't seem to be able to indentify whether a stranger has a reach advantage on them, unless its obvious due to a height difference or if the other guy has lanky, orangutan arms or something. I went back to the thread and do so we are retreading old ground with this topic. Someone posted Scott Hartman's take on the subject Carr also mentioned that he also thinks a sample of 25 animals is still rather small. Tyrannosaurus may have a much larger sample size than giganotsaurus but it is rather small. And as Scott points out, animal sizes can vary based on the time. Siberian tigers for instance were historically the largest extant cats, but nowadays bengals are larger on average. It is statisically unlikely that we found the largest specimen of either animal. So that's one Hartman's takes that I wouldn't consider valid. Both of these animals seem to be approaching the practical size limits of bipedal predator which is why I would be cautious in assuming that either predator has a decisive weight advantage on average. In fairness to the people that look at the largest specimen to determine the largest theropod, trying to research specific fossil specimens can be a bit of doozy. Most of them don't have convenient nicknames like Stan or Scotty and if you type something like MUCPv-Ch1 in google the first results are usually reddit or other social media sites. These two cases seem to be apple and orange cases. I understand you are trying to use an obvious example to illustrate your point but point but Argentinosaurus is clearly vastly larger than any theropod, so it's clean that no population of Argentinsoaurus is going to be smaller than a Meraxes unless someone happens to disocver a tiny dwarf population but no one in good faith would argue that a dwarf variation of animal would be a factor in a discussion like this. www.livescience.com/tyrannosaurus-rex-size-age.htmlI am not good at math, so I am not qualfiied to argue against your numbers. I will note, however, there are some instances when larger animals can be more agile than smallers. For instance, there are many anecdotal accounts of ravens being much more skilled and agile flyers than crows. Obviously these animals are vastly smaller than either theropod and aren't a great analogy but there are times when the seemingly more cumbersome animal may be more agile than the other. It also doesn't seem that rotational speed was just based on length and weight, leg strength was a factor too. I believe IB has posted many examples of tyrannosaurus adapations for agility and speed. Two, a big component of feline agility is their ability to leap around. Both of these animals are able to jump so turning abiility is one of the main factors here, whereas there are additional factors to consider for smaller animals. In addition, cats, bears, and canines have vastly different tail lengths and anatomy. I am not super knowledgeable on cats, but from what I understand, a cat's tail helps with balance and steering, whereas a canines tail is mainly for communication and a bear's might as well be noexistant. In contrast, tyrannosaurus and gigantosaurus' tails served similar functions. To be honest, it does seem you are being quick to downplay Eric Snively's findings while being to accept untested information that appears to be sourced from a blog post. Can we get Cau's exact statement on the subject? Because, saying that giganotsaurus appears to have adapations for enhanced lateral movement is not the same as, gigantosaurus how more efficent lateral movement than tyrannosaurus. I am inclined to take theropod's word for it, but again, its hard to say if it was a feasible when it hasn't been quantified or even tested yet. Based on all my years of watching animal footage, I highly suspect tiger sharks require a good grip and a decent amount of time, they usually aren't able to to just chomp into a turtle's shell like a cookie. In addition, many of the animals they scavenge are already dead or dying which would naturally make it easier to break the caparace. phys.org/news/2016-08-tiger-sharks-opt-scavenging-dead.htmlFirstly, this statement would need support before I accept it. More importantly, the eel in the example I posted was pretty much facing a worst case scenario. It's attacker was vastly larger than it, so it had the advantage of a much more powerful bite, longer teeth, and a greater gape. It was visibly missing large chunks of flesh from both its upper jaw and lower jaw, it suffered extensive damage on other areas of its body, and it was likely disoriented from being rolled around against its will, yet the eel was still able to bite back. So yeah, I do think it actually rather unlikely a gigantosaurus would be able to disable a tyrannssaurus' jaw fast enought to prevent the tyrannosaurus from fighting back at all. In some scenarios it might be able to maintain a grip that prevent the tyrannosaurus from biting back but that would be due to positioning rather than sheer damage output. Here is a longer version of the komodo video I posted before. At the start you can see the komodo working on the deer's back leg but at 2:25 not only is the deer able to run on that leg it kicks out with them as well. There are studies that indicate that komodos can induce shock in prey fairly quickly but in my experience, weapons are not as potent when actually tested on a living target. It seems pretty rare for animals to kill another similar sized animal with a single quick bite. They usually require multiple bites, or apply pressure for a good amount of time. On paper, a ziphodont theropod would just be able to bite a prey item on the back of the leg, cripple it, and watch it bleed to death for an easy GG, but I am inclined to think killing and wounding prey was easier said than done. For example, you would think that this leopard had this dog dead to rights, and in fairness it probably would have died if the owner didn't come out. But on paper a leopard should have been able to inflict a fatal wound on a sleeping dog's throat. In fact, one study indicates jaguars often kill prey outright. Again, human intervention is probably a key factor, but you would think that animal designed for skull crushing would be able to instantly kill smaller, sleeping dog at a higher rate. www.sciencedirect.com/org/science/article/pii/S2236377721000221#:~:text=A%20total%20of%2020%20attacks,other%20half%20in%20the%20south. Ripping out a "blood lusted" tyrannosaurus jaw and preventing it from using it at all in retaliation seems like it would take quite a bit of precision and luck. I haven't seen any examples of that happening. In fairness, there aren't many ziphodont animals around today, but I never heard of a case of a shark ripping out another sharks jaw muscles. If you scour across the internet you may find an example of a shark tearing off a smaller sharks jaw but suspect you're not going to find much evidence of sharks disabling other similar sized sharks jaws.
|
|
|
Post by Supercommunist on Sept 5, 2024 22:08:03 GMT 5
Also while its fair to say that, overall gigantosaurus' teeth are better at inflicting tissue damage than a rex's, the tyrannosaurus' teeth should have better penetration ability.
According to a study on knives. Stabbing injuries are much more likely to inflict fatal wounds than incised ones.
I am going to link the study but please be aware that there are graphic images in it.
iraqijms.com/upload/pdf/iraqijms57012c3a2ce9d.pdf
This is another reason why I think a tyrannosaurus would be able to disable a gigantosaurus' faster than vice versa. A tyrannosaurus' tooth might not be able to cut but it could embed itself deeper into the gigantosaurus' flesh.
Obviously, a gigantosaurus tooth also inflicts penetration damage as well, and part of the reason incise knife wounds are less effective is because knives lack mass, but even if you scale things up, thrusting weapons tend be more lethal than ones designed to cut.
Broken bones tend to hamper movement more than torn tissue, and even if the tyrannosaurus' doesn't manage to quickly break bone, its teeth are likely to be deeply embedded into a muscle.
|
|
|
Post by theropod on Sept 5, 2024 22:26:37 GMT 5
I imagine knowing this would change a lot of people's opinions on the relevance of the metrics that study used. Actually I very much doubt that it will. I would say that this is a somewhat imprecise description. But yes, it is likely that the differences in the "agility index" would be less if the tail musculature were included in the calculations, and not just the iliac musculature. Every study has its limitations though. What 40% size disadvantage are you talking about? It’s most likely that Giganotosaurus is somewhere around 10% larger than Tyrannosaurus. That will no doubt readily be dismissed as irrelevant in favour of the Tyrannosaurid’s many highly praised superiorities, however it should be noted that that is a larger relative difference than there is between boxers within one weight class, and these weight classes exist for a reason; that size is one of the most decisive factors in a fight, and that you just won’t get fair or interesting competitions if the combatants differ too much in terms of it. ----- Neither does quick striking or biting. Nobody seems to consider snakes to be particular geniuses (they certainly have lower EQs and brain masses than any large theropod), but they do it without much trouble. Trying to attack an animal as soon as it gets into striking range is pretty basic behaviour, trying to keep a potential threat at as much of a distance as possible while still being able to engage it is too, it does not require advanced understandings of the biology of one’s opponent. That would then impart a disadvantage to the party with the reach disadvantage, which will tend to underestimate the opponent’s reach, not on the party with the reach advantage, who will be fairly well-aware of their own reach at the very least. I am afraid I cannot see the relevance of any of that. Sure 25 (or 33 in this case) is a small sample if you compare it to some we have for other (especially extant) animals. But that has no bearing on the appropriateness of comparing average sizes, on the contrary, it is particularly BECAUSE we are comparing small and variably-sized samples that we have to compare average sizes over anything else, as we simply do not have the data to know maximum (or minimum) sizes or conclude about any differences in regard to them. Is that a statement based on quantifiable data, or just a personal feeling? I for my part have been reading statements like this for decades, but never have I actually seen a quantification of what the "practical size limits of a bipedal predator" (or similar) actually are. I’ve also frequently seen similar lines of criticism leveled at size estimates for giant sauropods ("size limits for terrestrial animals") or giant marine predators ("size limits of macrophagous predators"), but they always appear to be statements merely based on personal incredulity. Recently there was a study that proposed some T. rex individuals could hypothetically have reached 15 t (that was actually the study in response to which I made the graph we discussed before, because as usual, this study was of course done on T. rex and misconstrued by some people into saying that it was even bigger than we thought, which is not what it was about at all). They didn’t discuss any issues they had with that on the basis of it being larger than supposed "practical size limits", so it would appear they were as unaware of these limits as I am. Well, maybe one of them is actually a bit closer to the practical size limits of a bipedal predator, whatever those limits may be. Arguing they were both equally close and thus there could be no decisive size differences is basically stating the circular argument that "they were both the same size, ergo they were both the same size". As you say yourself, I am using an obvious example to illustrate the point. The point is that people (often the same people who will happily compare single specimens to determine which is larger with no regard for statistical rigor) will happily apply near-impossible standards of evidence when it comes to the suggestion that some other theropod might be larger than T. rex, standards that aren’t generally applied anywhere else. If we had the opposite situation, if the sample of adult T. rex specimens was larger than the sample of Giganotosaurus specimens with p=0.25, we would be standing knee-deep in claims of T. rex being clearly and demonstrably the larger taxon. How do I know that? Well, we already are, even though really people who claim that are doing so merely based on comparing two specimens (most commonly: the smallest known Giganotosaurus and the largest known Tyrannosaurus). I did not mean to imply that larger animals can never be more agile than smaller ones. But not without major differences in body shape or locomotory adaptations. What I wrote was that, if you simply make an animal larger without such differences (e.g. strong positive allometry of limb muscle size), you will make it less agile. This is simply because mass (and therefore inertia) scale to a higher power than area (and therefore strength) does. Of course depending on their locomotory adaptations, some larger animals may have specializations that enable them to outperform smaller animals (though usually at some sort of cost), but that doesn’t mean size doesn’t have a strong, inverse correlation with agility (in this case specifically defined as "turning ability"). You will find that I have already addressed this above. Snively et al. estimated an "agility index", calculated as the ratio between the lateral area of the ilium and the rotational inertia of the body. I simply discussed one of the two components of said index (the other, iliac area, being a direct skeletal measurement, I simply see no reason to question at this time). Turning ability undoubtedly is also a main factor for smaller animals. Linear accelleration and the lateral flexibility of the axis and limbs is another, probably for any terrestrial animal. Also RE: "Two", I’m afraid I didn’t see "one". We can happily ignore the tails if you want. A cat has a much longer tail, which would of course impart a vastly higher rotational inertia if treated as a stiff beam. However it isn’t a stiff beam, and will be flexed as needed to optimize turning ability, so I would not treat it the same way I would a theropod tail (although of course even treating a theropod as a stiff beam is inherently a simplification, real theropods were obviously not stiff beams, and as I believe I’ve mentioned elsewhere could, for example, have significantly lowered their inertia by dorsiflexing their neck and tail). But a cat will still have a considerably lower agility index than a canid at similar overall mass, yet I don’t see people lining up to claim a wolf would beat a leopard in a fight at the same size die to the former’s presumed superior agility. Maybe so, but it’s also symptomatic of the issue I’ve been going on and on about that there is a comprehensive study on every single advantageous trait of tyrannosaurids, but not when it’s a trait of other theropods. Certainly locomotory specializations of large carnosaurs are understudied compared to those of tyrannosaurids, but why is that the case? Cau’s blog is still in italian and doesn’t have a search function, so it’s a nightmare to search for a particular post. But IIRC he noted that based on a direct comparison between Tyrannotitan and Tyrannosaurus, but finding the post again may take a while. To be fair, neither are Alligators. It’s a myth that blunt-toothed bone crushers can just effortlessly crush anything in their path, just as much as it is a myth that animals with slicing teeth can just effortlessly slice through anything. Yes. I think this statement should be highlighted. But it also applies to animals with bone crushing bites, which have at least as much, if not more, difficulty inflicting lethal or debilitating damage to large animals. I would add though that, simply on account of size, it would likely be easier for giant theropods to inflict lethal damage, because their tissues would be less resistant compared to the inertia of their body masses pulling on them or resisting the pull. In the same way, you’ll have an easier time slashing water bottles with a sword when there’s actually water in them to keep them in place than if the bottles were empty. Why? Tyrannosaurus teeth are much thicker and would thus encounter much more resistance when penetrating prey, in addition to having serrations that do not cut efficiently the way those of sharks or varanids do (Abler 1992). No doubt, due to simply applying way more force, it would still have no problem penetrating with its teeth, but why would it be better at it? I don’t think you can compare a knife stab to a bite. The teeth in both these animal’s jaws are oriented the same way and would be driven into the prey when biting, most likely completely. However, one of them ( Giganotosaurus) has undeniably the much sharper teeth and lacks a large secondary palate (tyrannosaurids have this to increase torsional strength of the snout) blocking penetration, which would make it much more likely to actually incise a wound that would penetrate deeper than the length of the tooth crowns (which are not that different between these two taxa on average, don’t let the pictures of Stan’s skull with teeth that have slipped unnaturally far out of their sockets fool you). Also keep in mind the effectiveness of stabbing vs cutting depends a lot on the body region. Cuts are generally a lot more debilitating and dangerous than stabs when on the limbs, while on the torso stabs are more dangerous. TBH I think I’d rather have my neck stabbed with a blunt spike than slashed with a knife; Both are undeniably bad scenarios, but I think the chances of the latter failing to fatally damage any vital structure are a lot lower than those of the former. There’s a reason most bladed weapons made for fighting have some sort of edge on them rather than just a blunt spike (even though the latter would be far easier to manufacture), unless they are specialized for armor penetration.
|
|
|
Post by theropod on Sept 5, 2024 23:22:30 GMT 5
Actually what the heck. I realize that this goes way off-topic and may be more relevant to the mammal enthusiasts out there, but I was curious so I did a model calculation to demonstrate the point I made about canid vs felid agility indices when it comes to interpreting these indices for theropods, as calculated by Snively et al. : I used this wolf skeletal and this leopard: Of course, do note that I lay no claim to absolute accuracy at all, and that the results cannot be directly quantitatively compared to similar indices for bipedal animal because of the different locomotor mode (of course the quadruped will be able to excert vastly more torque for rotation around its COM because it has two sets of limbs that are spaced quite far from the COM). It is simply to demonstrate the kind of effect variations in morphology will have on the resultant agility indices with an example of a comparison between animals for which most likely nobody ever thought to quantify the same metrics given quite a lot of weight in discussions involving theropod dinosaurs. From these skeletals, it’s easy to trace body outlines and the lateral area of the ilium and (included because they are quadrupeds that use their forelimbs for locomotion) scapula, in order to end up with a representation like this: Doing a gdi and simplistically assuming a 1/2 width/depth ratio for the body throughout, you get two models, each with a volume of about 35 l (=a mass around 35 kg), that look like this: The leopard has a rotational inertia of 2.87 kgm², the wolf has a rotational inertia of 2.59 kgm². Both are probably underestimates, because a uniform width/depth ratio underestimates the width of the head and neck region, especially for the leopard, but they are good enough to get an idea of the relative values (the agility index’ main use is for comparison between taxa, it doesn’t have much use for quantitatively estimating actual turning speeds or the like because the assumptions made about musculature are very simplistic, reducing them down to a single number quantifying the attachment area, or part thereof) in the absence of a dorsal view. The leopard has a combined scapular and iliac area of 187 cm², while the same measurement for the wolf is 234 cm². As a result, you get an agility index of 65.2 for the leopard, but 90.0 for the wolf, quite a noticeable difference, mirroring the difference you get between carnosaurs and tyrannosaurids. If these were extinct theropods, we would get a study about them similar to Snively et al., noting how canids were apparently much more agile than similar-sized felids based on their rotational inertia and the size of their limb muscle attachment sites. This is just to highlight the difficulties and caveats of (over)interpreting such an index when it comes to the overall mobility of different organisms. There are of course many factors not covered by it, in both examples (the exact line of action and moment arms of the muscles involved, the limbs ability to impart torque on the body, the impacts of spinal flexibility and changes in posture, the type of muscle tissue and its specific tension, the error associated with the proportionality between muscle size and the area of attachment area, even large parts of the limb retractor musculature that Snively et al. left unquantified). In any case, I don’t believe I’ve ever seen someone favor a wolf over a leopard at weight parity due to the former’s perceived greater agility, although admittedly I am not at all very invested in that scenario, so who knows. And if I were to suggest that "T. rex vs Giganotosaurus is like wolf vs leopard", I would ironically probably be laughed at quite thoroughly in most online paleo circles, because that just doesn’t fit the preferred perception (even though, at least in terms of relative mobility, it is possibly a good analogue). As usual, here’s all the R-code and raw data to replicate this: rotI_wolfleopard.zip (86.83 KB)
|
|