30 PROCEEDINGS OF THE NATIONAL MUSEUM "ol. 122 



When we turn to social interactions, again it appears that the 

 presence or absence of certain movements as an arbitrary criterion 

 will not always suffice to delineate species or serve as a taxonomic 

 tool. This results, no doubt, from the fact that small nocturnal 

 rodents rely less on visual display in communication and, hence, 

 less on visual releasers in sexual isolation with the result that spec- 

 tacular differences in the fixed action patterns, observable in some 

 avian and fish taxa, are not to be found. Certainly visual communica- 

 tion by movements, postures, and coat color are important in some 

 mammalian taxa such as the primates and ungulates, but they are 

 obviously less important in the small, nocturnal mammalian taxa. 

 Again and again in this study, species-specific behavior patterns 

 are found to be variations on universal neuromuscular patterns. 

 Differences and similarities in the frequency of expression are more 

 reliable indicators of nonrelationship or affinity than is the criterion 

 of presence or absence of expression. In some cases even a frequency 

 analysis fails to indicate affinity above the generic level, for example, 

 in the temporal patterning of copulation. Thus, employing the 

 presence or absence criterion, some behavior patterns will be family- 

 specific, others genus-specific, and even more rarely a pattern will 

 be species-specific; however, analysis of the discrete temporal pat- 

 terning or relative frequency of occurrence almost always will demon- 

 strate species differences although the adaptive significance of such 

 differences may be unclear (consider again sandbathing). Since such 

 an equipotentiality exists in the repertoires of these mammals — 

 although frequency of occurrence serves to clearly delineate species — 

 I can only conclude that differences in thresholds exist that account 

 for the differential frequencies of expression. Thus, the behavior 

 of a given species may be described in terms of the most probable 

 sequence or set of patterns rather than in terms of its total potentiality 

 for expression. In this sense, species-specific behavior describes the 

 normal expression for a given environment and conforms to Ley- 

 hausen's "Verhaltenshaufigkeit" (Leyhausen, 1965). 



By applying this definition of species-specific behavior to the social 

 structure observed in different species, it is possible to characterize a 

 given species as belonging to one of several categories of social be- 

 havior. For example, Perognathvs 'parvus can be kept in groups with 

 little overt fighting. This is especially true if the group is composed 

 of littermates. Although they can tolerate one another, they do 

 not reproduce. Utilizing reproductive success as the criterion for a 

 natural social grouping, we can conclude that, although P. parvus has 

 a range tolerance that permits group life without overt aggression 

 leading to wounding, it does not have the ability to reproduce when 

 subjected to social contacts above some minimum level. In the case 

 of this species, the minimum appears to be only the pair association 



