[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Scaling problems in Hutchinson 2004

I was going to wait until SVP to comment on Hutchinson's J. Morph papers, but 
 some of the assertions concerning scaling of locomotary performance are so 
disturbing that refutations follow. 

H states that even at small sizes increasing size degrades absolute running 
performance, so it is more difficult for even large ratites to run faster than 
small birds. This is abjectly incorrect. 

The late Thomas McMahon was arguably the leading researcher on the scaling of 
locomotary performance, and his premature death (along with that of C R 
Taylor) has injured the field of animal biomechanics. In a series of innovative 
papers McMahon showed that as size increases absolute speed potential 
to the point that it is as if the largest ?animals had the opportunity to be 
the fastest as well as the biggest but didn?t care to try? (McMahon and 
Bonner1983, On Size and LIfe). H does not cite McMahon's extensive work on this 

The absurdity of the belief that small size favors high speed is exposed by 
noting that if this were true then ants, cockroachs and shrews would easily 
outrun cheetahs, ostriches and rhinos. That this is not true, that top absolute 
speed attainable at any given size increases with size, as long been understood 
and is well documented, as are the reasons why. 

Consider walking your chihuahua. You are strolling at an easy walking pace 
with a low stride frequency that requires little effort because it exploits 
a modest minority of your top exercise capacity. Meanwhile the peepy little 
chihuahua, its tonque hanging out for cooling, is working hard to keep up with 
a fast, high stride frequency running trot that requires it to engage a major 
portion of its top exercise potential. 

The reasons it becomes easier to move at a given speed as size increases 
involves many factors, including improving muscle efficiency due to decreasing 
stride frequency and limb excursion arcs. Basically it is a matter of moving 
lengthes per unit of time. To achieve 50 km/h (30 mph) an ant must run its 
own body length a couple of thousand times each second, a chihuahua must do 50 
body lengthes, and Tyrannosaurus about one. It is simply impossible for small 
animals to move their bodies so many multiples of their length in so short a 
time, for big animals it is easy. Obviously, as size goes up so does the length 
of the limbs, and the increase in stride length is more rapid than the 
decrease in stride frequency as shown my McMahon, so top speed rises (unless 
the big 
animal is poorly adapted for high speed).
Here we come to a key issue. In all of his papers on speed scaling in bipedal 
theropods (incl the Nature study) H cites the well known fact that the power 
that limb muscles can deliver scales to the 2/3s power, implying that unless 
relative muscle mass increases speed will decline with increasing size, and 
that this is true at all sizes. Yet this is obviously incorrect since it is 
small animals that cannot move fast, so what is going on? The solution is 
simple. It is also well documented that the power required to move at a given 
speed too scales to the 2/3s power, and does so pretty much regardless of limb 
configuration. Known since Fedak and Seeherman's 1979 Nature paper, this 
improvement in energy efficiency has since been measured in animals up to the 
size of elephants (in Langman et al. 1995, who deliberatley extrapolate the 
exponent to the largest extinct land animals). In none of his papers has H 
cited this crucial information. 

Crucial because if both the power that muscles can deliver and the power 
needed to move at a given speed scale to the 2/3s power, then the proportion of 
body mass that needs to be locomotary muscles in order to move at a given 
absolute speed is constant at all sizes. This is why the largest running 
including racehorses (the fastest accurately measured runners aside from 
cheetahs) and galloping rhinos (whose small legs imply relatively small leg 
both of which dedicate large portions of their body mass to extensive 
digestive tracts do not have exceptionally large limb muscles compared to small 
animals. H did not provide a large data set showing an increase in limb muscle 
in running land animals from insects to rhinos, and the limited available data 
that exists suggests there is little or none. 

H cites increasing proportional leg muscle mass in larger lizards and birds 
as evidence that it becomes harder to run as size increases. In doing so he 
contradicts Leahy 2002 (which is really Leahy and Paul 02 but for a typo) who 
stated that relative hindlimb muscle mass does not increase in birds as size 
increases. Where did that data come from? Well, in H's nature paper they cite 
leg extensors of a chicken as making up about 18% of total mass, which is 
similar to values for emus and ostriches published elsewhere, and also to the 
in the H's papers. Also, Maloiy et al. (1979 J Zool 187:161) observed no 
increase in proportional leg muscle mass in a large sample of birds, H did not 
cite this study. H's data is based on a small sample. On the other hand the 
fairly large sample in Hartman 1961 indicates small leg muscles in small birds. 
whether or not there is an increase in this parameter requires a fresh 
determination based on a large sample size. Even if ratites do have relatively 
leg muscles, H does not discuss the probablity that this is due to their not 
having a large portion of their body mass tied up in flight muscles, and even 
more importantly, their being much faster than the smaller birds. Get a well 
motivated guinea fowl, chicken, turkey, emu and ostrich on a racetrack. Anyone 
want to bet on which will win? A human can literally catch a chicken in a 
field for dinner, but not an ostrich.   

Actually, it is small animals that require proportionally larger leg muscles 
in order to run fast. That the power that muscles can deliver per unit cross 
sectional area scales to the 2/3s power means that smaller animals are 
producing many times more watts per unit of muscle mass. This is a critical 
because muscle can produce only so much power per unit mass, about 400 W/kg. 
Because the power required to run scales to the 2/3s power the power a unit of 
muscle mass needs to run at high speeds (over 30-50 km/h) becomes so high in 
small animals that it is impossible for them to achieve it. This is why little 
animals are not able to produce the extremely high relative power levels needed 
to achieve the extreme stride lengthes and frequencies required to run their 
body lengthes at the extreme multiples needed to achieve speeds easily reached 
by larger, longer striding, more energy efficient animals. If the locomotary 
muscles of juvenile and adult tyrannosaurs made up about the same portion of 
body mass as ratites, then they would have easily been able to produce the 
needed to run at about the same top speed. If tyrannosaurs had a much larger 
portion of total mass as leg muscles, then they would have been far faster 
than ratites. 

H also implies that at a given size animals need much larger limb muscles to 
achieve a given speed if they run on two legs rather than four, but again he 
does not provide the necessary supporting data. Why running on fewer legs would 
require more muscle volume is not at all obvious, especially since it has 
been known since Fedak and Seeherman (1979) that the power required to run at a 
given speed at a given size is the same in bipeds and quadrupeds, and snakes 
and multilegged arthropods for that matter. Back when I read the Roberts et al. 
(1998) paper claiming that turkeys needed to have 2.5 times more muscle volume 
than similar sized dogs I was taken aback by their combination of illogical 
analysis and sloppy data gathering. They did not actually compare the over all 
limb muscle volume in the two tetrapods - which would have required some 
simple tissue dissection and weighing. According to the data in H's new work 
and in 
Grand (1977) the percent of body mass tied up in locomotary muscles is not 
higher and may be actually less in turkeys than it is in dogs, cats and even 
baboons. Likewise the locomotary muscles of running dogs are not much smaller 
than those of ratites. Those who wish to argue that bipeds need much larger leg 
muscles than quadrupeds need to produce a large data set showing this is true, 
rather than citing a couple of poorly conducted studies. 

There is no doubt that basic absolute speed potential increases with 
increasing size, and this speed rise is operative in small animals. There 
probably is 
a speed limit at somewhere over 10 kg in that further increases in absolute 
speed are barred by certain physical constraints. As the McMahon-Bonner quote 
implies, the potential to remain fast should continue into the realm of the 
gigantic. That no animal does so today does not mean it cannot or did not 
anymore than the absence of modern land animals over10 tonnes or with necks 10 
m long bars such creatures from evolving in the past. Since the largest 
living land animals, elephants, evolved from slow ancestors, have no anatomical 
adaptations for fast running (and do not begin as fast juveniles and slow down 
they mature) and have small leg muscles, they are useless for answering the 
question (just as tortoises do not prove that animals of 50 kg cannot run 
fast). We tell whether past forms exceeded modern giants in size and neck 
by examining their dimensions. We determine whether past giants were faster 
than modern giants by examining their locomotary adpatations. The question 
whether certain extinct forms that show a full set of adaptations for running 
in terms of body and limb anatomy plus attachment areas of large leg muscles 
were achieving speeds many times higher than those seen in elephants, and as 
high or higher than in rhinos. 

Because H's papers continuely fail to cite basic research in numerous studies 
that favor high speed in large animals, and do not appear to understand 
important fundmentals of the scaling animal locomotion, his research cannot be 
presumed to answer this question. My confidence in the peer review system 
continues to decline. 

I completely agree with H that after decades worth of a vast body of papers 
on sometimes obscure aspects of animal locomotion, there remaining yawning gaps 
in terms of basic, critical knowledge of the subject that should have been 
taken care of years ago. This occurs because such simple data collection does 
not do much for establishing tenure or getting grants. Will someone please cut 
apart a horse (preferably thoroughbred) that was healthy when it died and weigh 
all its locomotary muscles? Same with a wild elephant? 

G Paul