Biology Reference
In-Depth Information
applied across a range of organisms, facilitating empirical testing through both
experiments and comparative studies.
(iii)
There may be coevolutionary arms races (Chapter 4), as for example between
predator and prey; if one side is ahead in the arms race, the other will appear to
be poorly adapted to its environment, for example hosts that are killed or
debilitated by pathogens and parasites (Rothstein, 1986).
(iv)
Individuals may be limited in the extent to which they can process relevant
information. Most ESS models implicitly assume that individuals have complete
information about the environment. For example, that the ducks and
sticklebacks of Chapter 5 are able to assess the relative rate at which food can be
acquired in different patches and adjust their foraging appropriately, or that the
parasitoid wasps of Chapter 10 can estimate the number of other females laying
eggs on a patch and adjust their offspring sex ratio correspondingly. If individuals
cannot assess such variables perfectly, or make errors, then we would not expect
them to behave 'perfectly'. For example, worker ants can only tell if their queen
has mated multiply, and adjust the sex ratio accordingly, if the queen mated
with males who smell differently (Fig. 13.12). Even if individuals could assess
the relevant features of the environment perfectly, it may not be in their best
interest to do so, if the resources that would have to be invested (e.g. time) are
too costly. Incorporating such informational constraints into ESS models, and
modelling information acquisition itself, remains a major task that would allow
ESS models to be tested more quantitatively.
… arms races …
… and
information
constraints
A general point about all of the possible constraints on adaptation is that they do not
invalidate the ESS approach, but rather suggest care in its application. Indeed, by
emphasizing clear and testable predictions, the ESS approach provides a clear method
for identifying constraints.
4
Quantitative tests of theory are often not possible
.
The critical reader will have noticed that although we stressed the value of making
quantitative predictions from optimality and ESS models, most of the tests of these
predictions were qualitative. The animals were usually seen to do 'approximately the
right thing': the dung flies in Chapter 3, for example, copulated for 36 minutes instead
of the predicted 41 minutes. Some might ask whether it is worth developing quantitative
arguments if the tests are only qualitative. There are three issues here.
Quantitative or
qualitative tests
of theory
In some cases, tests are only qualitative because of limitations in our
understanding or the techniques used to carry out the tests, and these can be
overcome. Once the quantitative predictions can be tested accurately,
discrepancies between observed and predicted results help to tell us what is
wrong with the models. For example, in Chapter 1 we discussed how allowing
for the cost of reproduction moved the predicted clutch size closer to the observed.
(i)
In some cases, quantitative predictions rely on so many biological details
that we would be unlikely to ever make quantitative predictions, and it
would not be very useful to try and do so. For example, in Chapter 2
we showed how variation in sexual dimorphism across primate species could
(ii)
