The Cave Lion (Panthera spelaea and Panthera fossilis)

[tigerluver]

(11-04-2018, 12:00 PM)GrizzlyClaws Wrote:

(11-04-2018, 11:21 AM)tigerluver Wrote:

(11-04-2018, 10:44 AM)GrizzlyClaws Wrote:

(11-04-2018, 12:25 AM)tigerluver Wrote: @Wolverine , I have not read of the released measurements for those fossils since their photos showed up years ago. However, I am certain larger P. fossilis are already on record. Let's use an ideal 200 kg lion as the isometric comparison.

The formula is:

Mass (fossil) = (Measurement of fossil)/(Measurement of extant specimen)^3 * Mass of the extant specimen.

Firstly, the 484.7 mm Chateau skull (Marciszak et al. 2014). A 200 kg modern lion could have a skull about 380 mm long. Applying these numbers to the aforementioned formula results in a mass of 415 kg.

Next, the 465 mm ulna (Reichenau 1908). A 200 kg lion could have an ulna about 385 mm long. Calculate... this P. fossilis weighed 352 kg.

Finally, the giant 192 mm MTIII (Marciszak et al. 2014). A 200 kg lion could have an MTIII of 145 mm. Calculate... a whopping 464 kg. Perhaps this is somewhat of an overestimate if P. fossilis was even longer limb-wise than the modern lion, but we have no skeletons to support this assumption other than the fact that its descendant, P. atrox, was indeed proportionally longer limb-wise. Moreover, MTIIIs vary a lot between individuals in relation to body size/mass. Nonetheless, the specimen was likely no less than 400 kg.

Both the MTIII and ulna can be confounded by an increased cursoriality in P. fossilis but the former much more so. The elongated skull of P. fossilis can also confound the isometric comparison to a degree.

From the scant fossil record, it is quite apparent P. fossilis consistently produced what would be giants by modern standards. For instance, many of the fragmented mandibles were likely from skulls that exceeded 400 mm. By probability and comparison of samples, P. fossilis was likely larger than P. atrox as well. The excessively large long bones of P. fossilis also hint that 130-140 cm at the shoulder would be comfortably attained by some of the largest specimens.

@epaiva I will post some photos soon, sorry about the wait.

Was the 192 mm MTIII specimen larger than the Chateau skull?

If we use the respective body parts of a 200 kg lion as the samples, it does appear so.

Maybe the 192 mm MTIII specimen could acquire a skull close to 500 mm which is the absolute maximum for any Panthera specimen?

The MTIII is a difficult one. One issue is that when calculating the range of weights produced by isometry, the range I found was as low as 400 kg to as high a over 500 kg. In other words, the bone is not that well correlated to body size, but is nonetheless from a top 10 Panthera specimen. The second issue is that if P. fossilis is more cursorial, the MTIII would be proportionately elongated and isometrically comparing the bone to that of an extant lion will heavily overestimate the weight. Thus extrapolating other body parts is filled with uncertainty. A skull length range of 450-500 mm would not be illogical in my opinion.

The issue can be applied to the skull. P. spelaea had a big skull for its bones, bigger than the extant lion. For instance, the specimen is Sabol (2018) has a femur to skull ratio of 0.99, which is distinctly less than extant lion generally as the number in P. leo is over 1.00. That means while applying the 484.7 mm skull to P. leo ratios, we'd think the femur would be 500 mm. Likely in reality the femur of the specimen was 470-480 mm, which is not that much bigger than the largest P. spelaea (470 mm femur from Germany mentioned by Deidrich). Of course still absurdly gigantic, but not what P. leo comparisons would lead one to believe. Another confounding issue is that long snouts usually means proportionately longer skulls for a body mass/size. We know that P. fossilis did have a longer snout that P. spelaea, likely making the long bone to skull ratio somewhat lesser in P. fossilis as compared to P. spelaea, further decreasing long bone length extrapolations by a bit. It is for this reason I feel the 465 mm has been underestimated via isometry or the 484.7 mm skull at least overestimated. Unless P. fossilis is exceptionally cursorial, a 465 mm ulna should give a femur of around 480 mm, matching the 484.7 mm skull. Now one can see why mass estimations vary so much, there are too many confounding variables. These two specimens were likely body size-wise just about the same as the 480 mm Ngandong femur for these reasons. Mass is a whole different rabbit hole in its own as the tiger and lion lineages hold mass differently. I have lately avoided addressing the complex topic and just give pure isometry estimates but hopefully this makes sense.

As a whole, the closer the species temporally to an extant species, the more accurate the single bone estimates and vice versa.

As for the rank of the cursoriality for the Pleistocene pantherine cats, it could be something like that?

Panthera fossilis > Panthera atrox > Panthera spelaea intermedia > Panthera spelaea spelaea > Panthera tigris (Pleistocene)

I still need to read up on the proposed P. s. intermedia but looking through old notes I tried comparing P. fossilis and P. spelaea cursoriality by comparing the MTIII to femur ratio. The data used is from Marciszak et al. (2014) for P. fossilis and P. spelaea I took from a few other places too.

As a disclosure, both the small sample size and fact that we do not have a multi-bone specimen of P. fossilis may (and probably do) affect the accuracy of the conclusions. 

I first averaged the lengths of the femurs (n=4) and MTIII (n=11) of P. fossilis. The same was done for P. spelaea (femur: n=12, MTIII: n=44). Now it is not possible to do a t-test for ratios as again, the data is not paired by individual but rather just a population average. The next best thing I could think of would be to bootstrap in a way (certainly not real bootstrapping as the data does not allow) and recombine the femur to MTIII ratio to produce multiple numbers. To do so, I used the confidence intervals of the averages and considered them their own unique data point. So for P. fossilis, we had the following data:

Avg. femur length: 382 mm
     - Low 95% CI: 339 mm
     - High 95% CI: 424 mm

Avg. MTIII length: 156 mm
     - Low 95% CI: 145 mm
     - High 95% CI: 166 mm

For P. spelaea:
Avg. femur length: 407 mm
     - Low 95% CI: 391 mm
     - High 95% CI: 422 mm

Avg. MTIII length: 141 mm
     - Low 95% CI: 139 mm
     - High 95% CI: 143 mm


Then the ratios were populated as follows: 
Low 95% CI MTIII/Low 95% CI Femur
High 95% CI MTIII/High 95% CI Femur
Avg. MTIII/Avg. Femur
Low 95% CI MTIII/High 95% CI Femur
High 95% CI MTIII/Low 95% CI Femur
Avg. 95% CI MTIII/Low 95% CI Femur
Avg. 95% CI MTIII/High 95% CI Femur
Etc.

So essentially a whole lot of permutations (9).

The range for P. fossilis was 0.37-0.47 (Avg. 0.41) . It was 0.33-0.37 for P. spelaea (Avg. 0.35). A t-test shows significant difference but there has already been too many theoretical calculations for me to put too much value in the t-test.

This analysis means P. fossilis had a longer MTIII, meaning it was more cursorial. This could very tentatively be supported by the fact that the distal long bones of P. fossilis are generally larger than those of P. spelaea, but the proximal long bones are not as consistently larger.


Far from conclusive as I'd rather have actual paired data but it's better than nothing.

On that note, I'll again stress the risk of extrapolating one bone length from another with the lack of data on the extinct species. By the ratios described above, the 192 mm MTIII would have a femur measuring c. 470 mm. This would not produce a 500 mm skull in all likelihood but rather a skull equal to or slightly smaller than the Chateau giant. The previous mass estimations I provided were under the assumption that P. fossilis was a P. leo clone but as people seem to be interested I've detailed all the caveats.

Again I will stress the aforementioned is all very theoretical, we have 4 P. fossilis femurs only. Nonetheless, food for thought. Reply