Horns, Antlers, Tusks, and the way evolution works
3 Sep 1996 07:49:57 -0600
I've been thinking about a concise way to describe how most
adaptationists (particularly this adaptationist) view(s) the relative
roles of constraint and adaptation in the evolutionary process.
Horns, tusks, and antlers all serve (or served) similiar evolved
functions: they are adaptations for fighting--in many species, fighting
between males for access to mates. Comparative data and behavioral
ecological observations support this adaptationist view; I'd be happy to
review them if necessary.
But why do some creatures (like deer) produce their weapons with bone,
while others (like elephants and bores) produce them with teeth, and
still others (like bighorn sheep) produce them with hair?
Most adaptationists would say that these differences are examples of
constraint; natural selection had different raw materials to work with in
each species. Convergence of function (trait analogy) can occur on many
'adaptive slopes/peaks'...which one a given animal 'climbs' is largely a
matter of what new genetic opportunities present themselves. Deer
ancestors had rudamentary antlers (bone) before forming proto-horns
But what about species in which females also possess these weapons? Does
this negate the functional explanation described above? Perhaps. Perhaps
not. We could retreat to arguing that developmental constraints apply;
males need the things, and through some embryonic developmental
constraint, the hardware for producing them are forced upon females as
well. Or we could argue that this was initially the case, and that
selection pressures favored the retention of female weapons (this is
But what about antlered animals having to regenerate them every year?
Aren't continuously growing horns "better"? Isn't this an example of
phylogenetic constraint? Or might there be an adaptive advantage to
shedding antlers? The latter is testable; the former is not.
To see if the apparent evolved 'purpose' of such a trait is being
retained in a population by current selection, behavioral observations
and perhaps experiments are called for. If it is found that the trait
affords possessors with fitness-improving effects, it is not unreasonable
to posit that the trait is being maintained by current selection.
The alternative paradigm, of emphasizing drift, random fixation, or
side-effects, do not predict such direct benefits, and are themselves
untestable. Hence one of my two big gripes with Gould. (The other being
his straw-man characterizations of the adaptationist program.)
This, in fact, was the dominant paradigm through the early part of this
century; antlers were declared to be functionless, just as beetle's
"horns" were thought to be "side effects of developmental processes." No
research is generated from such nihilistic proclamations. Only after
re-appraisal by adaptationists assuming selection on individuals for
inclusive fitness, for instance, was the likely function of beetle horns
as levers for removing rivals from females discovered.
It is, in my opinion, only possible to approach constraint and drift
explanations by falsifying adaptationist alternatives.
The dismissal of adapationist hypothesizing as "just-so story telling" is
only justified when those doing so cite no supporting evidence and make
no falsifiable predictions about their hypotheses. Dismissing working
hypotheses (which necessarily come before tested hypotheses) is