Mortal enemies form powerful alliances, discovered György Szabó
(Research Institute for Technical Physics and Materials Science, Budapest,
szabo@mfa.kfki.hu, 36-1275-4991)
and Tamás Czárán (Hungarian Academy of Sciences)
in a simulation of biochemical war among nine competing bacterial strains.
The study was inspired by recent discoveries of bacteria that excrete
toxins affective against their microbial cousins. The researchers began
by hypothesizing that there exist three types of related bacteria: toxin
producing Killers (K); Resistants (R) carrying genetic factors that
protect them from a given toxin; and Sensitives (S) that neither produce
toxins nor carry resistance factors. A population of S-type bacteria
are clearly at the mercy of K, but K-type can be overrun by R which
do not expend precious resources creating toxins. S-type can in turn
out-compete the R bacteria because they do not shoulder the burden of
producing toxins or resistance factors. This is a biological version
of the "rock-paper-scissors" game well known in various natural
populations.
The researchers were surprised by the consequences when they complicated
things by imagining that bacteria could produce two distinct toxins
and could carry their corresponding resistance factors. The hypothesis
leads to nine types of related bacteria with various combinations of
K, R, and S characteristics (KK, KR, KS, RS, RR, etc.). Under some circumstances,
trios of the very worst enemies form alliances that eliminate any of
the other six species they encounter, essentially as a side effect of
their vicious rock-paper-scissors game.
For example, the alliance KK-RR-SS, is one of three stable combinations
in which each strain entirely dominates one ally and is entirely dominated
by the other. While an invading RS species could out-compete the RR
strain, it is susceptible to attack by the KK strain and is overrun
by the SS strain. Similar troubles threaten any other invading strain,
and it is this multi-level attack that protects the alliance from all
other bacteria.
The researchers were also startled to find that varying the rate of
bacterial mutation in their model could dramatically affect the balance
of power on a bacterial battleground. At low mutation rates, one of
the three stable alliances will eventually wipe out all non-alliance
bacteria, with each alliance equally likely to win the day.
Above a certain critical mutation rate, bacteria change identities
frequently enough (e.g. KK turns into KR) that alliances break down,
and all nine strains are present in equal proportions. Changing the
mutation rate, therefore, can lead to a phase change as the system jumps
from a soup of all nine bacteria to a stable, cyclic alliance of mortal
enemies.
The complexity of even this simple system suggests that population
dynamics in nature is much richer than predicted by previous models
based on equations initially developed to study simple systems of predators
and their prey. (G. Szabó and T. Czárán, Physical
Review E, June 2001; text at Physics
News Select.)