BAKKER'S BRONTOPHAGY

[darren.naish@port.ac.uk]

I've been enjoying the comments on Uncle Bob's 
'brontophagy' paper (Bakker, R. T. 2000. Brontosaur killers: 
Late Jurassic allosaurids as sabre-tooth cat analogues. 
_Gaia_ 15, 145-158). I think whether you agree with him or 
not, his papers are always fun to read and do provoke 
thought and discussion. As to whether anyone actually 
believes his hypotheses, I recall being told that, following 
his talk on the same subject at SVP, [Jaime, it is this talk's 
title that was the inspiration for the 'fiery banana'] 
discussion went: 

"Did you see Bakker's talk?"
"Yes"
"Did you enjoy it?"
"Yes"
"Did you believe any of it?"
"No"

Apologies to the theropod worker I am paraphrasing.

I have some comments. Bakker's main premise is that 
extreme gape, reduced biting musculature, reinforcement of 
the caudal skull and big ventroflexor neck musculature = 
sabre-tooth style hunting. However, the ability to gape 
extremely wide does not necessarily tell you much about 
predation style seeing as thylacines (and apparently 
_Sarcophilus_ and some other dasyurids) have a gape of 
120 degrees, the same as that of most snakes. 

It should also be noted that the extreme gape of sabre-
toothed cats (more than 110 degrees compared to 65 for 
other felids) probably _isn't_ present so that they can strike 
down hard with an open mouth as Bakker posits, but to 
create clearance space between the tips of the upper and 
lower canines comparable to (and not bigger than) that seen 
in non-sabre-toothed carnivorans; in other words, the gape 
is a necessity evolved because of the exceptionally long 
teeth. As discussed below, because sabre-toothed cats and 
thylacosmilids (at least) were most likely throat-biters, they 
did not need to create more clearance space than non-sabre-
toothed taxa. The excessive clearance space between the 
tooth tips Bakker proposed for allosaurs is therefore non-
comparable and does not support analogy with sabre-
toothed mammals. The notable caudal skull reinforcement 
and neck musculature that Bakker cites as evidence for his 
model could be support for the hatchet-style attack mode 
proposed by Rayfield et al. and is not necessarily (so far as I 
can tell) only in agreement with his model.

The most important aspect of the paper though is that it 
relies on some dubious (and dare I say, outdated) 
assumptions about the way sabre-toothed mammals attack 
their prey. While Uncle Bob cites some recent literature on 
sabre-toothed cats (primarily Turner & Anton 1997) he 
seems to have missed/ignored the main point of those 
works... 

Firstly, that sabre-toothed mammals are not definitively 
specialised to 'harvest' mega-herbivores. Turner and Anton's 
contention is that sabre-toothed cats preyed mostly on 
horses, camels and other medium-sized prey; furthermore, 
some sabre-toothed mammals (including some nimravids 
and primitive thylacosmilids) are lynx-sized and could not 
have been mega-predators. 

Secondly, various lines of evidence indicate that sabre-
toothed mammals did not slash at the flanks of their prey, 
nor remove chunks as per the Akersten 1985 model (which 
incidentally is not cited by Bakker); it is looking more 
likely that they used their awesome pectoral and forelimb 
musculature to subdue prey on the ground before 
administering a carefully placed (and more conventional) 
throat bite. The unusual protruding incisor arrays, delicate 
nature of the sabre-teeth, mandibular flanges and other 
features further indicate that they were carefully guiding 
their sabre-teeth into a chosen spot and not hacking at 
random as proposed for the theropods. The latter 
observations suggest that sabre-toothed mammals are, 
again, poor analogues for big theropods.

On a minor mammalogical point, Uncle Bob says that the 
upper and lower canines of non-sabre-toothed carnivorans 
('from aardwolves to zorillas') are generally equal in size 
and shape. As a generalisation I suppose this is true, but the 
lower canines of most carnivoran groups I checked (dogs, 
bears, pinnipeds, mustelids, hyaenas, cats) are round about 
25-30% smaller than the uppers (this includes clouded 
leopards, elephant seals and servals, all of which have 
notably large upper canines)... only in Tigger Mamum-Ra 
the house cat were the uppers and lowers the same exact 
length (both 11 mm: she is exceptionally large). In the fossa 
(_Cryptoprocta_ the Madagascan proto-mongoose) the 
lowers are bigger than the uppers, but then this is a strange 
beast (and one that I will be seeing live, again, this 
Thursday).

While I don't reject Bakker's case that allosaurs may have 
had the anatomical specialisations he proposes (and in fact I 
agree with him that they probably _were_ sauropod-killers, 
as would most workers I believe), AND that they may have 
been >ecological< analogues of big cats, comparison with 
cats etc is otherwise misleading. If any theropod really were 
to dispatch prey in a manner at all analogous to that seen in 
sabre-toothed cats, creodonts and ameridelphians, it would 
be a form with hypertrophied maxillary dentition.. and of 
course that's _Ceratosaurus_. However, because theropods 
really aren't going to pin their prey down before carefully 
tearing through the trachea and carotid arteries etc, can they 
really be thought of as analogues for sabre-toothed 
mammals at all? 

Bottom line: it is perhaps misleading to see reptiles through 
mammalian eyes (and it was Uncle Bob anyone who once 
said "Mammal schmammal").

-- 
Darren Naish
School of Earth & Environmental Sciences
University of Portsmouth UK, PO1 3QL

email: darren.naish@port.ac.uk
tel: 023 92846045