Information Technology Reference
In-Depth Information
However, competition (red queen), due to the exhaustion of resources, soon
changes the tness of dierent genotypes. In particular, the location corresponding
to the origin of life becomes the one where competition is maximal, due to the large
connectivity. The competition induces speciation, so that a hypothetical \movie" of
the ancient lineages will show a diverging phylogenetic tree, in which the locations
corresponding to ancestors become minima of tness (extinction) after a speciation
event. As we have seen, competition also promotes the grouping of phenotypes and
the formation of \species", even in an asexual world.
In experimental and simulated evolutionary processes, mutants soon arises try-
ing to exploit others by \stealing" already formed products, like for instance the
capsid of viruses. This is the equivalent of a predator/prey (or parasite/host) rela-
tionship. Predators induce complex networks of relationship, for which even distant
(in phenotypic space) \species" coevolve synchronously. When such an intricate
\ecological" network has established, it may happen that the extinction of a species
may aect many others, triggering an avalanche of extinctions (mass extinctions).
The opposite point of view is assuming that chance and incidents dominate the
evolution. This could be a perfect motivation for punctuated equilibrium [72] (inter-
mittent bursts of activity followed by long quiescent periods): random catastrophic
non-biological events (collision of asteroids, changes in sun activity, etc.) suddenly
alter the equilibrium on earth, triggering mass extinction and the subsequent re-
arrangements. These events certainly happens, the problem is that of establishing
their importance.
It is quite dicult to verify the rst scenario computationally. A good starting
point is the food web [73]. We can reformulate the model in our language as follows.
A species i is dened by a set of L phenotypic traits i
1
; : : : ; i
k
, chosen among K
possibilities. The relationship between two species is given by the match of these
characteristics, according with an antisymmetric KK matrix M. The score
W (ijj) accumulated by species i after an encounter with species j is given by
X
k
X
k
1
L
W (ijj) =
M
i
n
;j
m
:
n=1
m=1
A species accumulates the score by computing all possible encounters. A monotonic
function of the score determines the growth (or decreasing) of a species. When the
system has reached a stationary state, the size of a species determines its probability
of survival (i.e., the survival probability is determined by the accumulated score).
When a species disappears, another one is chosen for cloning, with mutations in the
phenotypic traits.
A special \species" 0 represents the source of energy or food, for instance solar
light ux. This special species is only \predated" and it is not aected by selection.
This model is able to reproduce a feature observed in real ecosystems: in spite of
many coexisting species, the number of tropic level is extremely low, and increases
only logarithmically whith the number of species.
