Skull, proboscis musculature and preorbital gland in the saiga antelope and Guenther's dikdik (Mammalia, Artiodactyla, Bovidae)
The convergent formation of a proboscis-shaped nose in Saiga tatarica and Rhynchotragus guentheri has led to a reduction of the rostral parts of the bony nasal region. The maxillae, in particular, form a marked dorsally directed saddle-shaped concavity accommodating the connective and muscle tissues of the proboscis. The lacrimal bones as well as the bony nasal opening of the skull are of a considerable dorsoventral height. A likely reason for this may be the reduced length and caudal displacement of the more clearly tiered turbinate bones. The completely different muscular morphology of the proboscis in both species points to different phylogenetic origins. The ability for an extraordinarily precise spatial orientation of the proboscis in R. guentheri is due to a ''muscular hand'' on both sides, each consisting of four extremely long ''fingers''. The necessity for an exact space-oriented smelling (''directional smelling'') is explained as an adaptation to a highly structured habitat and by the specific feeding habits of the species, which selects certain plant species and even their specific parts from a large variety of plant species occurring in the habitat, primarily by olfactory stimuli. In contrast, the open and mostly arid habitat of Saiga tatarica does not require such precise orientation of the nostrils. Saiga antelopes are characterized as an intermediate feeding type with a tendency to grazing. Their range of feeding plants comprises a smaller number of species which, however, occur in greater numbers and over larger areas and, therefore, are more easily located by visual and olfactory stimuli. Consequently, the musculature of the proboscis is less differentiated, but its large-area fan-shaped muscles facilitate the discharge of concretions of dust and nasal mucus, regularly found in its nose as a result of dust-filtering. The evolution of a proboscis in S. tatarica and R. guentheri is assessed as an example of morphological convergence without functional and adaptational convergence. Hence the observed differences between the probosces of the two species, particularly with respect to musculature, may be called paradaptations. The different position of the preorbital gland with respect to the surrounding facial muscles is related to differences in the social structure and marking behaviour of the two species. Guenther's dikdiks an monogamous and live in permanent territories, the boundaries of which they mark several times a day by actively pressing the gland against prominent plant parts, thereby ''affixing'' the preorbital secretion. Compression of the gland is by contiguous facial muscles. As the preorbital gland is of considerable size throughout the year, large preorbital pits have formed in the skull. Saiga antelopes are polygamous and do not occupy any marked territories, not even during their brief rutting season. During the rut, preorbital glandular secretion is released more or less continuously without substantial action from surrounding muscles and is spread diffusely by a tuft of hair below the eye. It is only during the rutting season that the preorbital gland grows to substantial size. For most of the year it remains small. The gland is positioned in an inconspicious and shallow depression of the lacrimal bone.