The diversity of both scent-marking motor patterns and the anatomy of scent-glands may be explained as examples of reproductive isolating mechanisms. These mechanisms commonly evolve in order to decrease the likelihood of interspecies mating which would waste the female's reproductive energies [Dobzhansky, 1937]. If we assume that one of the functions of muroid rodent pheromones is to attract potential mates, and there is abundant behavioral evidence for this, then it is to the advantage of each species to develop different types of pheromones. Ultimately, of course, the process ought to require the production of different enzymes and chemically differentiable pheromones [cf O'Connell et al, 1978], but in the process of differentiation it is not surprising if various species accentuate the development of certain combinations of skin glands different from those of sibling species. From this it would follow that with a change in skin glands, there would be a strong natural selection force for changing the motor patterns of scent-marking so that they would optimally broadcast the secretions of the glands. Just as it has been proposed that each species "chooses" several from a "list" of potential skin glands potentially available to all muroid rodents, so too it may be proposed that each species "chooses" several from a "list" of scent-marking motor patterns that are, potentially available.
At this point we have no evidence concerning the extent to which the differences in scent-marking across species reflect differences in genetic instructions for brain mechanisms and to what extent they reflect developmental differences dependent upon the anatomical location of scent glands. It is entirely conceivable, for example, that an animal actively secretes pheromones from certain glands during patrol/marking and that it learns to scent-mark by "rubbing where it itches." Although presence of the gland is not necessary for scent-marking motor patterns in the adult, at least in the case of the ventral rub of Mer unguiculatus [Blum and Thiessen, 1970], it remains to be determined if presence of the gland plays a critical role in the ontogenetic development of the motor pattern.
Whatever their genetic and developmental basis may be, the behaviors of scent-marking are unique among agonistic behaviors of muroid rodents in serving as reproductive isolating mechanisms. For that reason, they are the only clear example of "species-specific behavior" comparable to the unique songs of bird species [Beach, 1960] .
Submissive motor patterns and their effectiveness in inhibiting or precluding offense vary across the various muroid rodent species. The contrast is particularly well documented between the laboratory rat [Blanchard et al, 1977a] and the laboratory mouse [Blanchard et al, 1979]. In the former, the submissive animal uses a full submissive posture and the offensive animal is inhibited from delivering a bite-and-kick attack because it does not get releasing stimuli from the dorsal surface of the opponent. In the latter, the submissive animal uses an upright submissive posture instead; if he were to use a full submissive posture it would not inhibit offense because offensive biting in the mouse unlike in the rat can be directed at the ventral surface of the opponent. The full submissive posture is lacking in certain other muroid rodent species, although it has not been determined if bite-and-kick attack also depends on different releasing stimuli in those species.