Applications of Search in Biology: Some Open Problems - Search Theory: A Game Theoretic Perspective - page 285

butterflies, and to correlate these with environmental factors, but to infer the mech-
anisms driving the observations is a much greater challenge. Further practical chal-
lenges of predicting the conservation consequences of changes in habitat structure
can be answered only speculatively at this stage (Preston et al., [ 25 ]). State-based
models offer some natural avenues of progress; the key challenge is more likely to
be the framing of the biological question and interpreting the data, rather than the
mathematical technicalities.
Some Individuals Are Not Individuals
An ant colony might contain thousands of foraging non-reproductive clones, serving
the interests of a single fertile female. In effect the only individual of interest is the
queen, and ecological success of foraging workers needs to be measured in this
context. This is an extreme case, but sociality is common in animal studies and is
perhaps mirrored at smaller scales by quorum sensing in bacteria. This is a large
and active research area (see, for example, [ 27 ], which contains excellent ecological
background spanning the remit of this article) and notions of inclusive fitness in
social systems are well developed, though not without controversy [ 20 ].
Taking plants as a less obvious example, each root tip could be thought of as
an individual forager (Problem 2, below). Neighbouring roots are not independent,
however, and the metric of interest is the cumulative foraging success of each plant's
dynamic population of roots. The scope for practical problems and interesting math-
ematics is enormous.
Sometimes Being Optimal Is Not Good Enough
Much of the theory of search and rendezvous concerns finding mathematical solu-
tions which minimise the expected value of some property. This makes sense for
long-lived and “valuable” humans, but is probably too anthropocentric a view for
biology in general. Evolution by natural selection involves, by necessity, probabilis-
tic forces. “Survival of the fittest” is a statement about extremes, not averages. Put
simply, if an individual is almost certain to die soon anyway, then it has no interest
in maximising its long term average fitness - it only needs to get very lucky, very
quickly. This has been long understood in human societies [ 21 ], and so-called “risk
sensitive foraging” is not a new theory in ecology [ 26 ].
This is not necessarily a difficult problem to overcome mathematically; one sim-
ply optimises the property of interest in the surviving tail of the distribution rather
than globally [ 9 ]. However, its consequences are possibly very large (see Problem 1,
below). New technologies are revealing that natural systems can evolve dramatically
so as to exploit stochasticity [ 2 ]. In this author's biased opinion, of the complica-