In die-hard beetle collector), conducted field work in

In 1848 Henry Walter Bates, an entomologist (and die-hard beetle collector), conducted
field work in the Amazon, where he collated large collections of animals. Bates was a
friend and colleague of Wallace and continued to correspond with him during this time,
encouraging the development of his theory of evolution by natural selection. After four-
teen years in the Amazon, Bates published a famous paper detailing his observations of
colour patterns in butterflies. He observed that non-poisonous butterflies tend to mimic
the bright warning colourations of poisonous butterflies. Bates reasoned that the non-
poisonous butterflies must have evolved, by natural selection, to resemble the poisonous
ones (see chapter 6). Bates (1863) hinted at the board applicability of theories of warning
displays and mimicry.

Bates’ work was the first landmark in coevolutionary theory: mimicry was an appar-
ently intuitive example of reciprocal evolutionary change in interacting species. However,
although evolution was studied rigorously after Darwin’s theory was published, it took
longer for coevolution and interspecific interactions to receive the same broad attention.

Eleven years after Bates’ discovery, Fritz Mu ?ller contributed an extension to mimicry
theory. Influenced by his field work, Mu ?ller (1879) made an argument to show how two
poisonous animals might benefit from sharing a warning colouration. Whereas Bates re-
ported a parasitic type of mimicry (where one species benefits at the expense of at least
one another), the type of mimicry reported by Mu ?ller appeared to be mutualistic. Thus,
mimicry offered two forms of coevolutionary interactions: antagonism and mutualism

Despite the importance of coevolution to the fundamental questions in biology, it was
pushed to the wayside in the early 20th century for a long period of time. Thus, students
of evolutionary biology were taught only half the problem of evolutionary adaptation.
When interspecific interactions and coevolution was taught, it was done regarding only
specific cases, rather than as a general principle (Thompson, 1994). Thus, coevolution
was pushed back as a focus for evolutionary theorists. What caused this slump in studies
of interspecific interactions in evolution? First, evolutionary biology seemed to lack the
appropriate tools for modelling such interactions (although Mu ?ller had shown that simple
mathematical approaches could help). Second, and probably most influentially, there were
bigger fish to fry.

Although coevolution is a fundamentally important principle, there were even more
pressing and immediate problems for evolutionary theory. Two biological paradigms—
the naturalist/biometry (statistical study of biology) camp and the geneticist (study of
hereditary and transmission of genes) camp—were apparently at odds (for more detail
see Mayr & Provine, 1980). The stances taken by the two camps seemed incommensu-
rable. Many of the problems were down to communication problems and idiosyncratic
terminology. However, the problem plagued evolutionary biology. Further, questions
such as “If parents’ traits are combined in an offspring, then why have the colours of or-
ganisms not evolved into a grey blended mess like randomly mixed paint?” were given
as a problem for evolutionary theory. Questions such as this were used as sticks to beat
evolutionists.

The pathway was left open for a genetical theory of natural selection. This unified
theory was provided by one of the most influential evolutionary theorists: Ronald Aylmer
Fisher. In his highly influential book The Genetical Theory of Natural Selection Fisher
unified Darwinism and Mendelian genetics, reconciling biometry and genetics by show-
ing that genetics was actually reinforcing biometrics

The integration of Mendelian genetics into evolutionary theory also answered the
question of why everything does not evolve into a grey mess: parents’ traits are not
blended but are inherited in sections of genes from each parent’s chromosome. Thus
Darwin’s hypothesis that innate characteristics of an organism were somehow passed on
to its offspring had finally been paired with a plausible mechanism.

Although work on coevolution in general made little progress in the first half of the
twentieth century, mimicry was a notable exception. It carried the flag for coevolution
with Fisher (1930) devoting an entire chapter to it. The chapter was the only one to deal
with interspecific interactions in detail.

After the shock-waves of the evolutionary synthesis had settled, coevolution made its
way back into the limelight. Not only were the fundamental pre-synthesis questions out
of the way, but new techniques had emerged which were well suited to studies of simple
coevolutionary interactions.

Game theory is a mathematical technique that was initially developed for use in eco-
nomics. In particular, the use of game theory to find Nash equilibria1 was proving useful
to evolutionary thinkers. Due to the parallels between economic change and evolution, it
became apparent that game theory could be applied to theoretical studies of coevolution-
ary phenomena. This application was championed by the highly influential and respected
biologist John Maynard Smith (see e.g., Maynard Smith, 1982). The availability of game
theory as a tool (and later, fast computers and other techniques) led to an increase in work
on these topics.

An important example of the success of the game-theoretic approach was in work
on the evolution of communication. In an often-cited paper, Krebs and Dawkins (1984)
described interactions between species in terms of mind-reading and manipulation. (The
perspective introduced by Krebs and Dawkins will prove useful later on in this thesis
when we develop a theoretical understanding of warning signals and mimicry.) Krebs and
Dawkins maintain that when there is a conflict of interest, receivers (the animals receiving
the signal) are under selection pressure to critically assess the behaviour of the other in
order to exploit any telltale signs of their intentions, thus extracting useful information.
On the other hand signallers are under selection pressure to manipulate the receiver into
doing what they want.

Biologist Leigh Van Valen proposed what is now known as the Red Queen principle. The
name comes from the analogy to the Red Queen chess piece in Lewis Carroll’s Through
the Looking Glass, who explained to Alice “. . . here, you see, it takes all the running you
can do, to keep in the same place.” Essentially, the principle explains that for an evolving
organism, continuing development is needed simply in order to maintain its fitness relative
to the species with which it is coevolving (Van Valen, 1973). The only way predators can
compensate for a better defence by the prey (e.g., gazelles running faster) is by developing
a better offence (e.g., leopards running faster). In turn, prey need to evolve a better defence
to escape predators, which will again be met by an improvement in predator offence; and
so the cycle continues. Of course, there is often an escalation limit; the cycle has to stop
somewhere due to physiological, genetical or environmental constraints. Thus, we do not
see animals running quickly enough to break the sound barrier.

The red queen effect is not ubiquitous in coevolution. As Krebs and Dawkins noted in
the case of conspiratorial whispers, some coevolutionary interactions are mutualistic, in
which the species involved benefit from complementary interactions. Darwin tackled mu-
tualistic coevolution explicitly in discussing his ideas on how bees and flowers interact.
Darwin dedicated his follow-up to the Origin to this issue and backed up his ideas with
some observations (Darwin, 1877). He imagined the size, length and body form of hon-
eybees evolving to better match the shape and length of the corolla tubes surrounding the
pollen so as to more efficiently exploit the resource. Further, he imagined the shape and
length of the corolla tubes to be modified to best allow bees to pollinate them 

BACK TO TOP