Biology Reference
In-Depth Information
woodpeckers could not tell in advance which were the empty logs, so they had to use
information gathered at the start of foraging on each log to decide whether or not it was
likely to be empty and therefore should be abandoned. When the logs contained 0 or 24
seeds the task was easy: looking in a single hole in theory gave sufficient information to
decide and the woodpeckers, in fact, took an average of 1.7 looks in an empty log before
moving on. The task was more complicated when the two kinds of log contained 0 and
6 or 0 and 12 seeds: finding one empty hole is no longer enough to reject a log, but there
must be some point at which the information gained from seeing a succession of empty
holes makes it worthwhile giving up. Lima calculated how many empty holes the
woodpeckers ought to check before giving up on a log in order to maximize their rate of
food intake. The calculated values were 6 and 3 while the observed means were 6.3 and
3.5; thus, the woodpeckers use information gleaned while foraging in a way that comes
close to maximizing their overall rate of intake.
The risk of starvation
Two kinds of currency for foraging animals - rate of food intake (starlings, great tits)
and efficiency (bees) - have come up so far. Another currency that may be important for
foraging animals is the risk of starvation. This is especially likely to be important when
the animal lives in an environment that is unpredictable; the exact amount of food the
animal will obtain is uncertain.
For example, imagine you are offered the choice of two daily food rations: one is fixed
at 10 sausages per day, the other is uncertain; on half the days you get five sausages and
on the other half, 20 sausages. Although the
average
of the second diet is higher than
that of the first, it is a riskier option. Which is the better option? The answer depends on
the benefit (or 'utility' in economic jargon) of eating different numbers of sausages per
day. If a diet of 10 is enough to survive on while five is not, then nothing is to be gained
by choosing the risky option. If, on the other hand, 10 is not quite enough to survive on,
the only viable option may be to take the risk and hope for 20 sausages. This option
offers a 50% chance of survival while the certain option offers no chance.
In short, animals should be sensitive not only to the mean rate of return from a
particular foraging option but also the variability. Whether or not animals prefer high
variability should depend on the relationship between the animal's needs (usually called
its
state
) and the expected rewards. If energy requirements are less than the average
expected reward, it pays to choose the less variable option (
risk-averse behaviour
) whilst,
if requirements are above average, it usually pays to choose the more variable option
(
risk-prone behaviour
).
This idea has been tested in an experiment by Caraco
et al
. (1990). They offered
yellow-eyed juncos (
Junco phaeonotus
) (small birds) in an aviary a sequence of choices
between two feeding options: one variable and one with a fixed pay-off. For example, the
variable option in one treatment was either 0 or 6 seeds with a probability of 0.5 each,
whilst the corresponding fixed option was always three seeds. The experiment was
carried out at two temperatures: 1 and 19°C. At the low temperature the rewards from
the fixed option were inadequate to meet daily energy needs, whilst at 19°C they were
sufficient. As predicted by the theoretical argument, the birds switched from risk-averse
behaviour at 19°C to risk-prone behaviour at 1°C. An equivalent result was obtained by
Risk-averse versus
risk-prone
behaviour
