to more complicated ones having sharp multivariate filters that exclude all
other possible signals but those of conspecifics (Enquist and Arak 1998).
Such a simple recognition strategy might have some advantages of neural
economy relative to a more complicated one that relied on a multitude of
signal parameters, and in the simple environment of only one heterospecific,
a more generalizing strategy would be as effective as a more specific one—
effective, that is, until there appeared a new heterospecific with an even
longer signal and the simple recognition strategy was foiled because the
receiver now made recognition errors by responding to the longer heterospecific signal. Responses to this longer heterospecific signal would be an
incidental and maladaptive consequence or response bias of the original, and
adaptive, mate-recognition strategy. This is bound to happen because selection cannot anticipate future situations but can only judge among alternative
phenotypes in a current context.But if the costs of making a recognition error
toward longer signals exceeded the benefits of such a simple recognition
strategy, we assume that the recognition strategy would evolve further under
selection (Dawkins and Guilford 1996; Ryan 1999).
Such response biases are the basis of another exploitative, interspecific
interaction similar to the orchid–insect example reviewed above. Cuckoos
do not raise their own young. They place their eggs in the nests of other
species, where their young are raised by the host, often to the detriment of
the host’s own young. Reed warblers are one such host. Even though the
host young produce begging calls to elicit feeding from their parents that
are quite different from the begging calls of the cuckoo (Fig. 5.7), reed warblers feed cuckoos preferentially to their own young. The begging call of
the cuckoo does, however, mimic the sounds of a group of begging reed
warblers (Fig. 5.7). In a series of elegant experiments, Davies et al. (1998;
see also Kilner et al. 1999) showed that it is this cuckoo signal that exploits
the receiver system of the reed warbler.
In many other cases, we might not expect response biases that emerge
from recognition strategies to be maladaptive. Ryan and Keddy-Hector
(1992) reviewed numerous cases of female preferences based on longdistance signals. If female preference deviated from the population mean,
it was usually in the direction of greater signal energy—more intense, longer
signals produced at higher repetition rates. If such preferences represent
general response biases of many auditory-recognition systems, they might
still continue to guide females toward conspecific males and, incidentally,
toward males that are in better physical condition, being able to marshal
more energy to support calling. An alternative is that the general pattern
of bias toward greater signal content is an evolved response to choose such
males in the first place.
6.3.2. Habituation and Song Preference in Birds
Songs of many oscines are characterized by their signal complexity. One
explanation for such complexity, the antimonotony hypothesis, is based on
256
M.J. Ryan and N.M. Kime
other possible signals but those of conspecifics (Enquist and Arak 1998).
Such a simple recognition strategy might have some advantages of neural
economy relative to a more complicated one that relied on a multitude of
signal parameters, and in the simple environment of only one heterospecific,
a more generalizing strategy would be as effective as a more specific one—
effective, that is, until there appeared a new heterospecific with an even
longer signal and the simple recognition strategy was foiled because the
receiver now made recognition errors by responding to the longer heterospecific signal. Responses to this longer heterospecific signal would be an
incidental and maladaptive consequence or response bias of the original, and
adaptive, mate-recognition strategy. This is bound to happen because selection cannot anticipate future situations but can only judge among alternative
phenotypes in a current context.But if the costs of making a recognition error
toward longer signals exceeded the benefits of such a simple recognition
strategy, we assume that the recognition strategy would evolve further under
selection (Dawkins and Guilford 1996; Ryan 1999).
Such response biases are the basis of another exploitative, interspecific
interaction similar to the orchid–insect example reviewed above. Cuckoos
do not raise their own young. They place their eggs in the nests of other
species, where their young are raised by the host, often to the detriment of
the host’s own young. Reed warblers are one such host. Even though the
host young produce begging calls to elicit feeding from their parents that
are quite different from the begging calls of the cuckoo (Fig. 5.7), reed warblers feed cuckoos preferentially to their own young. The begging call of
the cuckoo does, however, mimic the sounds of a group of begging reed
warblers (Fig. 5.7). In a series of elegant experiments, Davies et al. (1998;
see also Kilner et al. 1999) showed that it is this cuckoo signal that exploits
the receiver system of the reed warbler.
In many other cases, we might not expect response biases that emerge
from recognition strategies to be maladaptive. Ryan and Keddy-Hector
(1992) reviewed numerous cases of female preferences based on longdistance signals. If female preference deviated from the population mean,
it was usually in the direction of greater signal energy—more intense, longer
signals produced at higher repetition rates. If such preferences represent
general response biases of many auditory-recognition systems, they might
still continue to guide females toward conspecific males and, incidentally,
toward males that are in better physical condition, being able to marshal
more energy to support calling. An alternative is that the general pattern
of bias toward greater signal content is an evolved response to choose such
males in the first place.
6.3.2. Habituation and Song Preference in Birds
Songs of many oscines are characterized by their signal complexity. One
explanation for such complexity, the antimonotony hypothesis, is based on
256
M.J. Ryan and N.M. Kime