|
|
Post by Grey on Oct 29, 2013 23:15:20 GMT 5
There is that large tooth : www.fossilsonline.com/index.php?main_page=product_info&cPath=264_155&products_id=1835Not technically a posterior but apparently massive. Based on the given width measurements of 3 3/4 inches, I've measured the crown height to be about 7.4 cm (using paint). I'm not sure where to place that tooth in the dentition, and thus to calculate the body size of the shark. If my measurement is good (7.4 cm), if my position is good (L5 ?), if my calculation with Shimada's method is accurate, I've got a TL estimate of about 20 m. If someone wants to verify it. In anycase, it seems that tooth came from a very large shark according to the description. |
|
|
|
Post by theropod on Oct 30, 2013 3:20:40 GMT 5
I get almost the same figure for the crown height, 7.388cm ( method of measurement?), however since the width is rounded to quarter inches, the actual size could lie anywhere within ~3% in either direction. Based on Shimada's formula for an L5 (TL[mm]=-5.778+26.381*CH[mm]) the result is 19.44m (7.4cm yield 19.46, so im prepared to argue the result is roughly 19.5m), thus definitely a very big specimen and among the largest known (at least assuming this position). How did you deduce this was a L5? Judging by what you posted here?, L4 or even L6 seems equally possible. Using these immediate neighbours gives us highly fluctuating results. For future use I've attached an .ods-file for the formulas: shimada-method.ods (14.32 KB) |
|
|
|
Post by Grey on Oct 30, 2013 4:22:20 GMT 5
I've used L4 and got around 13.5 m TL, which is big but not exceptionnal for a megalodon, as the source talks about a "large posterior" tooth coming from a "very large shark", I've assumed this was not a L4. And L6 gives totally unreasonnable estimates...once again if I've performed the calculation right. Looking at meg tooth set I also think L5 is most accurate. BTW, can be remarked in the last file I've put in the megalodon thread that Pimiento considers, using Gottfried, as well <14 m individuals potentially juveniles based that in Gottfried table, the largest immature meg reaches ~14 m TL. Then in the paper, she states >10.5 m individuals as adults, based that 10.5 m corresponds to the size of the smallest mature meg. At the end, any meg between 10.5 m and 14 m can be potentially an immature or a mature individual.
Also, Pimiento has posted on twitter, a poster of her actual on-going research about meg size trends she's going to present at the 2013 SVP.
|
|
|
|
Post by theropod on Oct 30, 2013 15:55:05 GMT 5
It's easily possible that fossil-selling page is not a trustable source regarding its estimations on tooth positions. Even the experts gave ranges of possible positions in Pimiento et al., 2010, so I doubt that page can possible be sure about its size guesstimate.
But yes, some 35m shark is not among the most probable explanations here, and its quite possible this was indeed from a 19m+ specimen.
|
|
|
|
Post by theropod on Oct 30, 2013 16:10:58 GMT 5
However, when substituting the tooth width in the assembled dentition from Hubbel (assuming an L5 position) the widest tooth measures 13.29cm in total width, which would suggest a total body lenght of 16.7-18.2m. So either way a large specimen, but whether its exceptional heavily depends on the method.
|
|
|
|
Post by Grey on Oct 30, 2013 16:52:50 GMT 5
That is why I m just asking directly to sharks experts. But fossils collectors can be skilled at placing teeth. They even precise this is not technically a posterior, hence less likely a L6 than a L5 but a L4 is still possible, though not representing then a large/certainly adult meg. A L5 or L4 are equally possible awaiting confirmation.
I guess your 16.7 m figure comes from Kent, but I recall this was not intended to give a method based on tooth width for meg. However, estatimates based on jaws perimeter (seen in the Parotodus and Cardabiodon paper) would interest be but they require a number of factors (total teeth width, spacing between teeth, ratio jaws/body in lamniforms...). If one wants to try study this...
In any case, seems that tooth comes from a massive predator, I hope to get a rigorous about its position, that is the determining factor.
But given the coloration, I guess it comes possibly from South Carolina where numerous big teeth are found, so that meg being 18-19 m wouldn't be surprising.
It is good that finally a new paper discussing size trends is coming this winter but I don't think I will be present at this time, you guys could discuss it without me.
|
|
|
|
Post by theropod on Oct 30, 2013 18:28:57 GMT 5
16.7m bases on Kent's regression for tooth width of great white sharks. Intended for this use (which it is of course, what purpose else would it be for?) or not, it's at least as valid a method as the unpublished ratio from Jeremiah. Since it is a good proxy for jaw perimeter, tooth width is also probably the best methodology in general.
Its pretty difficult to say what this tooth really is, there are too many variables. Given some of them, it could be extremely large, given others, not that much, that's all we know.
|
|
|
|
Post by Grey on Oct 30, 2013 19:00:52 GMT 5
No I had posted Kent's mail, he just wanted to show the potential error in that, no methodology for estimate behind it. He suggests the best method is to calculate the upper jaw perimeter, though he did not perform it for meg but for Parotodus, and Siversson did it for Cardabiodon.
I m doing some research regarding that tooth, and others. Will post the results.
|
|
|
|
Post by theropod on Oct 30, 2013 19:20:16 GMT 5
What for did he fit the function to the dataset and give the formula then? That is a model that can be used for estimating values, just like Jeremiah's.
|
|
|
|
Post by Grey on Oct 30, 2013 19:38:17 GMT 5
To show the potential huge error with Jeremiah's. At least with white shark teeth.
The best is to use total perimeter but that's trickier to do. If you want test it...
|
|
|
|
Post by theropod on Oct 30, 2013 23:20:56 GMT 5
In that he as well created a regression line that can be used for estimates, nowhere does it say this was not valid. There is not even a sample given for Jeremiahs, so we cannot really say which one is "the one deviating from the other"...
I don't see what's the big difference between tooth width and total jaw perimeter. Does it matter whether one uses tooth width to calculate jaw width and then jaw width to calculate total lenght, instead of just using tooth width directly for estimating total lenght? Most experts seem to regard tooth width as a good proxy in absence of preserved jaws.
|
|
|
|
Post by Grey on Oct 30, 2013 23:34:22 GMT 5
I'm the one here discussing with Kent since two years so I fairly know what he explained to me.
The purpose of this scatter is not to be a method of his own but to show the limits of tooth width isometry.
Jaw perimeter is less prone to error than a single isolated tooth, especially if based on others species than white shark (which shows greater variations).
I couldn't perform successfully any calculation based on Hubbell's set yet. I'd ask datas to Siversson but I know he's busy at the moment..
|
|
|
|
Post by creature386 on Oct 31, 2013 0:11:27 GMT 5
|
|
|
|
Post by Grey on Oct 31, 2013 2:27:48 GMT 5
Pimiento's tweet :
Sad there is only a glimpse, the information seems extremely interesting. Coming in the new paper certainly.
Other than that creature, the fact Ward, Kent and others (discussion with Ehret) agree that tooth width is reasonnable does not mean Jeremiah's value is 100% reliable, nor even 90%, the problem being it has not been tested, added that, Kent demonstrated it, tooth width and body growth are weakly correlated. Kent did not intend to propose a new method but to show the issues using the width of isolated teeth. However, data based on complete set of teeth ( the number of teeth in the dentition, their combined width and the space between the individual teeth) do however strongly correlate with the size of the mouth. Mike Siversson seems to have performed unpublished works about this for meg, and have performed it for Cardabiodon. I don't think Siversson formula and value for meg are similar to Jeremiah's... One can contact him about this and get more clues.
In the meantime, all the new datas by Pimiento and Balk using Shimada's method will be interesting. I know they use collections from various museums in US.
|
|
|
|
Post by theropod on Oct 31, 2013 3:07:21 GMT 5
In demonstrating the variation that is found in the majority of animal samples, he at the same time calculated a regression line (and no, this is not "Jeremiah's formula", its the regression line based on the given sample, that is given in the figure).
This is of course at the same time a valid formula, and since Jeremiah's method is not even properly documented there's no reason for it to overthrow this one, which bases on a decent sample. To date, you should be content with using both.
Of course he did not create a new method, he rather made a proper/new model on one that already exists.
A correlation coefficient of 0.65 is not exactly great, but its much better than nothing. I don't think arguing about weak correlation is any good, other methods (eg. tooth height) likely have even worse ones.
The point I don't understand is that either way we are stuck with isolated teeth, we do not have several evidently belonging to the same specimen in the majority of times, so either way our estimates won't get better than ones based on isolated teeth.
There is no difference between being strongly correlated with another thing that in turn is strongly correlated with a third, and being strongly correlated with that thing in the first place.
|
|
