species versus among species rather than implying unvarying speciesspecific stereotypy. It seems to us that in many if not all cases the variation
in long-distance signals within the species is substantially less than the variation of the same kind of signals among closely related or ecologically sympatric species. Thus, the properties of the signal can be used for correct
identification of the species producing it, as anyone who has tried to identify a bird, frog, or a cricket sight unseen might know. The females of the
respective species are even more unerring in their identification, and it is
this interaction of signal and receiver that has put long-distance communication at the center of research on speciation theory (e.g., Blair 1958;
Alexander 1962; Andersson 1994; Howard et al. 1998).
Preferences of receivers for conspecific versus heterospecific signals lead
to assortative mating, reduce the opportunity for reproductive interactions
between incipient species, and contribute to the genetic divergence among
populations that is critical for speciation. The preference for conspecific
versus heterospecific signals is a species-isolating mechanism because it
restricts reproduction among different species. The preference is also a premating isolating mechanism because the effect takes place prior to the act
of mating, unlike a postmating isolating mechanism, whose effect on reproductive isolation takes place after the mating act; hybrid sterility would be
an example of a postmating isolating mechanism.
5.1.1. Species Discrimination as an Incidental Consequence
The use of the term “mechanism” might imply that the function achieved,
in this case preference for a conspecific versus a heterospecific signal,
evolved under selection. Although there appears little doubt that selection
can result in the evolution of such discrimination patterns (see Section
5.1.2), the preference for a conspecific signal versus many heterospecific
signals may also be an incidental consequence of signal preferences already
in place. For example, if an animal expands its range into a new area and
encounters other species for the first time, it might have no problem
recognizing the signals of its own species from that of the newly encountered heterospecifics. This particular set of preferences for conspecific
versus heterospecific signals did not need to evolve; it was already present
as a consequence of how this recognition mechanism—this interaction
between properties of the signal and the receiver—happened to evolve in
the past. It is adaptive (the current effect on fitness) to the animal to prefer
the conspecific signal versus the novel heterospecific ones, but this particular preference is not an adaptation (the evolved function). The fact that a
communication system results in effective species recognition need not
mean it evolved for that purpose.
5.1.2. Reinforcement and Reproductive Character Displacement
A second possibility invokes direct selection in the evolution of materecognition signals. When two forms of a single species exist in sympatry,
5. Selection on Signals
243
in long-distance signals within the species is substantially less than the variation of the same kind of signals among closely related or ecologically sympatric species. Thus, the properties of the signal can be used for correct
identification of the species producing it, as anyone who has tried to identify a bird, frog, or a cricket sight unseen might know. The females of the
respective species are even more unerring in their identification, and it is
this interaction of signal and receiver that has put long-distance communication at the center of research on speciation theory (e.g., Blair 1958;
Alexander 1962; Andersson 1994; Howard et al. 1998).
Preferences of receivers for conspecific versus heterospecific signals lead
to assortative mating, reduce the opportunity for reproductive interactions
between incipient species, and contribute to the genetic divergence among
populations that is critical for speciation. The preference for conspecific
versus heterospecific signals is a species-isolating mechanism because it
restricts reproduction among different species. The preference is also a premating isolating mechanism because the effect takes place prior to the act
of mating, unlike a postmating isolating mechanism, whose effect on reproductive isolation takes place after the mating act; hybrid sterility would be
an example of a postmating isolating mechanism.
5.1.1. Species Discrimination as an Incidental Consequence
The use of the term “mechanism” might imply that the function achieved,
in this case preference for a conspecific versus a heterospecific signal,
evolved under selection. Although there appears little doubt that selection
can result in the evolution of such discrimination patterns (see Section
5.1.2), the preference for a conspecific signal versus many heterospecific
signals may also be an incidental consequence of signal preferences already
in place. For example, if an animal expands its range into a new area and
encounters other species for the first time, it might have no problem
recognizing the signals of its own species from that of the newly encountered heterospecifics. This particular set of preferences for conspecific
versus heterospecific signals did not need to evolve; it was already present
as a consequence of how this recognition mechanism—this interaction
between properties of the signal and the receiver—happened to evolve in
the past. It is adaptive (the current effect on fitness) to the animal to prefer
the conspecific signal versus the novel heterospecific ones, but this particular preference is not an adaptation (the evolved function). The fact that a
communication system results in effective species recognition need not
mean it evolved for that purpose.
5.1.2. Reinforcement and Reproductive Character Displacement
A second possibility invokes direct selection in the evolution of materecognition signals. When two forms of a single species exist in sympatry,
5. Selection on Signals
243