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

RE: Gangs, packs, etc.

Bonnan gives an excellent summary of the situtaion. I basically agree with
him, but differ on a couple key points.

The "mammilian pack hunting" model is the dominant popular view (i.e.
Discovery Channel et. al.).  Older scientific work also supports the notion
of close cooperation in mammilian carnivores like wolves, lions etc.
However, more recent field studies have largely discredited this view.
There is no hard evidence that a deep level of cooperation exists.  

The most important such study was Schaller's work on lions in the Serengeti.
He witnessed 523 hunting events and recorded the statistics of success for
various numbers of lions.  Here is the results of hunting success for all
kinds of prey:

1 lion  15%
2 lions 29%
3 lions 27%
4-5 lions 32%
6 or more lions 33%

(from table 59, The Serengeti Lion)

So, two lions are basically just as effective hunting together as they would
be if they hunted separately (actually, slightly less successful, but that
is within experimental error).  There is NO difference in effectiveness due
to hunting "cooperatively".  They get the SAME results they would get if
hunting alone - they just happen to do it together.

After two lions, the story get worse.  Three lions are about as effect as
two, and WAY less effective than if they split up.  Adding additional lions
does not help. The statistics above are for all prey, but Schaller's tables
breaks it down by each prey animal and the same result applies.

Since Schaller's study (1967) numerous field biologists have documented this
with increasing accuracy for lions and (particularly work by Craig Packer)
and in wolves (Thurber and Peterson, J Mammology 74(4):879-889, 1993) and
other group hunting carnivores.  The same conclusion holds - "packs" have NO
benefit, and usually a large COST when compared with the same number of
animals hunting alone. 

You might think, OK, that is hunting success, but what about number of Kg of
meat, or reduction in variance of hunting success, or size of prey.   Maybe
they get the same percentages, but they can kill larger prey, or hunt more
often, or more reliably...

All of these measures, and many more, have been tried and failed.  There is
no detectable advantage to so called cooperative hunting.   The best current
theories of social hunting (particularly Packer's work s see Heinsohn &
Packer, Science 269:1260-2, 1995) are not based on hunting success at all,
but rather other social interactions (holding a territory, reproduction).

So, if you believe that "pack hunting" mammals have deep cooperation, you
must explain how it is that they get ZERO benefit from it.  In fact, they
frequently get very high COST compared to splitting up and hunting
individually.  If it does not have any measurable impact in terms of
success, is it really cooperative?

The same field biologists who gather these statistics also have described
hunting in detail.  Their reports are consistent with the conclusion - there
isn't a lot of deep cooperation.  For every case where it appears that a
lion flushes prey into the waiting grasp of a pack member, there is a case
where the lion flushes prey the wrong way.  We don't tend to see that flim
clip on the Discovery Channel.  By counting only the sucesses, we make them
look very good.

Detailed field reports are consistent with lions, wolves and other group
living mammilian carnivores being what Bonnan calles "gang hunting" - i.e.
pack members hunt together, and rush to grab the same prey.  Selective
reporting and anthropomorphism lead us to conclude that great intellegence
is involved.  

Of course, these animals ARE individually quite intelligent, and they are
highly social with a rich set of social behaviors, some of which are
expressed during hunts. They are excellent hunters, and do have the ability
to anticipate their prey in hunting (there is good evidence of this).  

So, I am not suggesting that lions and wolves are dumb, or that they are
dumb hunters.  However, available evindence suggests that this intelligence
is expressed invidually, and that they do not have deep levels of
cooperation in hunting. 

The implication is that high intelligence is required if you want to be a
wolf or lion, but there is no indiciation that high intelligence is required
to cooperate in hunting as well as a wolf or lion.  "Pack hunting" is NOT as
effective as people think, and does not require high intelligence.  

Lions and wolves are social creatures and hang out together when they hunt,
just as they do when they sleep or do other functions.  When hunting
together they probably are "gang" hunters - i.e. they each follow an
individual strategy, with little or no input about what their companions are

There are MANY well know examples of simple, individual behaviors giving
rise to apparent complex group behavior.  Schooling of fish, flocking of
birds, and many other things have been shown to be due to this. 

Note that I am not saying I have hard evidence that this is preciesly what
lions and wolves do - that is a stronger statement.  However, the evidence
is very consistent with this, and it is NOT consistent with deep levels of

Gang hunting exists across a large swath of the animal kingdom, all the way
down to army ants.  In fact, it is quite likely that ant hunting statistics
would show that ants probably get far more benefit from their gang hunting
than lions or wolves do from their pack hunting.

There are birds which "pack hunt" (again, most likely gang hunting), notably
Harris' hawk.  Most predatory birds are solitary, or live in pairs, so
broadly one can concude that are not group hunters.  However, there are many
species of social living birds (albeit not raptors).  

In conclusion:

-  Bonnan is absolutely right that definitions are essential to this debate.

-  Pack hunting by mammals is not supported by hard data.  If it exists, it
does not effect hunting success, which is very hard to explain.  Most
likely, so called cooperative pack hunting is really just gang hunting -
simple piling on.

-  Theropods cannot be ruled out as gang hunters, on the basis of
intelligence since gang hunting occurs in such a wide variety of animals.

-  In particular, pack hunting (which is probably gang hunting) does occur
in some birds (although not in most predatory birds).  As a result, the
extant phylogenetic bracket method cannot be used to rule out gang/pack
hunting for theropods.

-  At the very least, the discussion above should make the point that it is
suprisingly hand for field biologists to get hard data on how well LIVING
animals cooperate in hunting.  Determining this with any scientific accuracy
for extinct animals is far harder yet.