The Cave Lion (Panthera spelaea and Panthera fossilis)

[tigerluver]

(11-04-2018, 08:08 AM)GuateGojira Wrote:

(11-04-2018, 04:54 AM)Wolverine Wrote:

Quote:Calculate... a whopping 464 kg. 

464 kg... P. fossilis was really a titanic predator dwarfing all felines we have on the planet now! So 140 cm shoulder height is not impossible digit.

Thanks a lot for the answer, every your post in the area of paleontology has a value equal to the value of all of our posts combined together.

We must be carefull with the estimates based in only one bone. For example I will consider that of the skull the most reliable, but it will depends of the database used (the more specimens the better), for example there is a problem with the lion, as this species had a large head in comparison with its body and consequently with its body mass, this means that a large skull do not represent a large specimen all the time. Also, 380 mm will be "average" for lions in South Africa but is the maximum for lions in East Africa.

I will recoment to use more specimens like Christiansen & Harris (2009) which use many lions, tigers and jaguars to provide a wide spectrum. The method used by Sorkin (2008) is the same formula but using only the largest specimens (commonly not related between each other): If we use the maximum figures for modern lion in scientific records, which is a skull of 401 mm and a weight of 250 kg, will produce a mass of 441.5 kg; if we use the maximum in hunting records, which is 419 mm skull and a weight of 272 kg, will produce a mass of 421.1 kg; if we use average for South Africa lions, which is 380 mm skull and weigh of 190 kg, the mass will be 394.3 kg; finally if we use the average of East African lion, which is 363 mm skull and 175 kg, will produce a mass of 416.6 kg. The average of all figures is 418.4 kg, which is a real giant amoung the great cats and probably the largest Felidae ever!

Now about the limbs, some bones may be very long and other very wide, for example the longest femur from Panthera atrox is not the widest. Christiansen & Harris (2005) noted this with the Smilodon, where some bones provide lower estimations than others, even amount one single specimen, check this: Single bones produced different estimations, and this is clear when we see the result of the complete (associated bones) of Smilodon fatalis and Smilodon populator fossils used by Christiansen & Harris (2005):

* Smilodon fatalis - LACM PMS1-1:
Range: 195.1 - 279.0 kg, mean of 4 bones= 241 kg (weighted).

* Smilodon populator  - CN52
Range: 231.2 - 316.2 kg, mean of 4 bones= 258.2 kg (weighted).

So the best form will be to have many bones from the same specimen in order to make an "average" of the estimations made, but as we know, this is normally not the case, as most of the specimens are fragmentary.

The skull is likely the least reliable, maybe in competition with the MTIII. In proportion, P. spelaea had a long skull for its long bones (Sabol 2018). The ulna is weight bearing and the variation in mass to bone length proportions is much less in such bones (Christiansen and Harris 2005). 

Moreover, this is a good opportunity to explain the flaws of regression and ideas behind estimation a. The Christiansen and Harris (2005), or any interspecific regression for that matter, should be taken with a grain of salt. By grouping together a plethora of species to produce the regression, the scaling terms are confounded by the fact that the equations essentially assume that all cat species are proportioned the same. Think of it this way: a jaguar is bulkier than a lion for its bones. When you plot these two species together, there becomes the illusion that as body frame increases, mass does not increase at the same rate (negative allometry, scale factor <3). In reality, this happens because the lion is naturally leaner for its frame, not because there is an actual negative allometry. Conversely, put together the leopard and the tiger in the way Christiansen and Harris (2005) did for regression. Suddenly, mass is growing much faster than the bone frame. This is because the tiger is bulkier, not because bigger framed cats are by some law heavier for their bones. When one uses regression and scale factor comparisons within the species, isometry rather than negative or positive allometry becomes the trend with a large enough sample size. That's why in my opinion, if an extant analog to an extinct species is available, using isometry for mass prediction is likely best, as Christiansen and Harris (2009) did.

Also please note that the Sorkin method is just the conventional isometric comparison method, nothing more and nothing unique. The author just uses the largest extant specimen for isometric comparison, but that is an arbitrary measure. 

These giants are so far out of the extant range that the negative allometry in the Christiansen and Harris (2005) predisposes to inaccuracies not represented by the percent errors provided. The weighted method is not much of an actual correction either and has not been used since. Weighting percent errors is akin to an organization investigating itself, the results don't mean enough. Reply