An Analysis of Hair Structure and Its Phylogenetic Implications among Heteromyid Rodents

-H~~~ morphology of 36 species of the family Heteromyidae including the genera Dipodomys, Perognathus, Microdipodops, Liomys, and Heteromys was studied using both light and scanning electron microscopy. Variables investigated included length and width of hair, imbricate scale pattern, external and cross-section form of hair, and medullary characteristics. Although the hair of individual species could be characterized with detailed study, we do not believe that hair structure will be of value in evolutionary studies of this group below the generic level. The overhair of heteromyid rodents falls into two morphological types-hair which is round to oval in outline and hair which has a trough along the dorsal surface. Hair of the first type is found in most perognathines, and all members of the genera Dipodomys and Microdipodops. Troughed hairs are found in all chaetodipines, Liomys, and Heteromys and the perognathines, P. ampulus and P. formosus. As a taxonomic character, hair structure has been used primarily in studies of bats (Cole, 1924; Nason, 1948; Benedict, 1957; Miles, 1965). These authors concluded that hair characters were of no value below the generic level. However numerous keys have been written for mammals, which are based solely upon hair characteristics (Toldt, 1935; Mathiak, 1938; Mayer, 1952; Moore et al., 1974). Such keys have been used in regional surveys and investigations of food habits (Day, 1966, 1968; Stains, 1958; Williams, 1938), forensic medicine (Hausman, 1944; Stoves, 1942), and archaeology (Douglas, 1965; Appleyard and Wildman, 1969). Short (1978) has recently studied the cuticular scale patterns of the hair of a variety of species of mammals. He concludes that species could not be characterized solely on the basis of cuticular scale pattern but they could be if a variety of characters of the hairs were used. The thermoregulatory aspects of mammalian hair covering have been studied by several workers (Herrington, 1951; Carpenter, 1966; Ghobrial, 1970). All hair is comprised of three major portions-cuticle, cortex, and medulla. The cuticle or covering of individual hairs is made up of overlapping scales, the distal edges of which are free. The patterns made by these scales have been classified by Hausman (1920) into two types-coronal (scales surrounding the hair shaft) and imbricate (scales not surrounding the hair shaft). The shape of the individual scale is also an important feature. An inverse relationship between the width of a hair and the proximodistal length of each scale has been reported by Hausman (1930) and Noback (1952). The primary function of the cuticle is protection of the hair (Rudall, 1941; Azzola and Shurmann, 1969). After treatment with trypsin, the structural elements of the cuticle are found to be a network of what appears to be fibers of various widths, which appear randomly oriented (Lundgren and Ward, 1963). Another structural component of the cuticle can be seen in some overlapping scales where small projections from the underscale protrude into the overlying scale (Rogers, 195%). The development of scanning electron microscopy and other techniques have allowed detailed study of the ultrastructure of cuticular scales. Each scale is composed of an epicuticle, an exocuticle, and an endocuticle. The outermost layer, the epicuticle, has received much attention (Haly et al., 1970; Bradbury and Leeder, 1970; King and Bradbury, 1968; Leeder and Bradbury, 1968). This layer influences the surface properties of wool, as well as probably other types of hair, including wettability and frictional characteristics. J. bfornrn., 59(4):74~760, 1978 740 November 1978 HOMAN AND GENOWAYS-HETEROMYID HAIR 74 1 The middle layer of a hair shaft, the cortex, is made up of fusiform or spindle-shaped cells, which interdigitate with each other along the long axis of the shaft. In some kinds of hair there are air spaces or fusi between the cells (Hausman, 1932). Each cell contains a nucleus and pigment granules. As keratinization takes place in a growing hair, the cytoplasm is replaced by fibers of protein (alfakeratin). A keratinized cortical cell contains rounded subunits (macrofibrils) 3,000 A in width. Within each of these units are still smaller units (microfibrils) 60-80 A, which are separated and delineated by thin layers of dense matrix (Anderson and Lyeder, 1971; Anderson and Lipson, 1970; Rogers, 1959a, 195%). A microfibril contains "protofibrils" 20 A wide and is composed of smaller filaments, possibly single protein molecules (Johnson and Speakman, 1965). Another division of the cortex is that of the orthoand paracortex in some wools. Chapman and Bradbury (1968) concluded that the differences between the two areas are based on (1) the configuration of the microfibrils, (2) the microfibril to matrix ratios, and (3) possible variation in amino acid sequences in the two areas. The central portion of the hair shaft, the medulla, is made up of cells of various shapes, which are often interspaced with air pockets. The presence and patterns of these cells have been used to distinguish various kinds of hair (Hausman, 1920, 1944; Dearborn, 1939; Day, 1966; Mayer, 1952). Bradbury and O'Shea (1969) analyzed amino acid components of the medulla in various monotremes, marsupials, and placental~ and discovered remarkable similarity of composition in all medullas. Pigment is normally concentrated in the medulla cell, but may be lacking entirely. The present study provides data on various aspects of the morphology of the hair of members of the rodent family Heteromyidae using both light and scanning electron microscopy. Heteromyid rodents form a unique and autochthonous component of the North American mammalian fauna. Systematic studies of the various genera of the Heteromyidae include the following: Dipodomys (Burt, 1936; Lidicker, 1962; Lackey, 1967; Stock, 1974); Perognathus (Patton, 1967a, 1967b); Microdipodops (Hall, 1941); Liomys (Genoways, 1973); Heteromys (Goldman, 1911). None of these studies utilized structure of hair as a systematic character. Variables investigated in the present study included length and width of hair (as defined by Mayer, 1952), imbricate scale pattern (Hausman, 1920; Miles, 1965), and medullary characteristics (Hausman, 1920; Mayer, 1952). MATERIALS AND METHODS Thirty-six species from the family Heteromyidae were chosen for study. Among these species are representatives of all subfamilies, genera, subgenera, and species groups recognized by Hall and Kelson (1959) and Genoways (1973). A number of authors stress the use of underfur in studies of hair (Hausman, 1920, 1930; Cole, 1924). Others have preferred the use of the guard hair (Williams, 1938; Stoves, 1942; Mayer, 1952; Stains, 1958). The overhair or guard hair was used in the present study because this type is far more abundant than underfur in most heteromyids (Williams, 1938) and because guard hair is much more regular in shape and easier to work with than the fine underhair. Hair samples were obtained from museum specimens and were taken from an area of the dorsum at the level of the scapulae. Additional samples were taken from the rump for comparison as well as to characterize the so-called spines prevalent in some species. A dissecting needle was used to draw the hair toward the head (against the normal grain), exposing a distinct line of fur. Samples were taken along this line with curved forceps with fine tips. Care was taken to pluck the hairs close to the skin to insure that the base of each hair was included in the samples. In this manner one or two samples provided enough hair for the study, but in no way altered or damaged the specimen. Where possible, hair was taken from specimens collected soon after the spring molt. Hair samples were labeled and stored in one-dram vials with plastic caps until 742 JOURNAL O F MAMMALOGY Vol. 59, No. 4 A preliminary series of scanning electron micrographs revealed the presence of a trough or depression along the dorsal surfaces of hair of certain species ofPerognathus. Consequently, we thought a study of hair impressions in nail polish might reveal not only the scale patterns, but perhaps also the presence and extent of the trough. A modification of Weingart's (1973) technique was employed. A thin layer of inexpensive clear nail polish was spread evenly on a glass microscope slide and allowed to dry for at least 1 h. A second coat, which needed at least 3 h to dry sufficiently, was applied to the slide for study of the coarse hair of Liomys and Heteromys. Hair was washed in acetone and placed between the coated surfaces of two slides. The slides were then placed between two wooden blocks and compressed with C-clamps for up to 24 h. The best impressions were obtained when the clamps and blocks were undisturbed; bumping the apparatus or tightening the clamps often blurred the impressions. Photographs of the base, midsection, and tip of each hair were taken on a Leitz-Wetzlar Ortholux microscope at 5 0 0 ~ magnifications. Because the scale impression technique failed to give complete information on the extent of troughs in the perognathines, we used the scanning electron microscope to explore this feature. This technique was also used to elucidate any peculiarities of the surface structure of the hair, which were not seen with light microscopy. Ten to 12 hairs, including a few from the rump, were cleaned in acetone, dried, and mounted on double-sided tape on metal stubs. They were then coated with gold in a vacuum evaporator. Each sample was examined with a Kent-Cambridge S4-10 scanning electron microscope (SEM). All hair was observed carefully before representative photographs were taken. Coated samples were stored in small, dust-proof boxes for future reference. Because of some discrepancies in our observations between light and scanning microscopy, and to aid in the study of medullary patterns and cortex-medulla relationships, cross sections of the hair of all species were made. Several hairs of each species were embedded in Spurr's plastic (Spurr, 1969; see also Mathiak, 1938, and Williams, 1938, for other methods of sectioning). Small plastic boats were used for embedding; only a thin layer of the plastic was necessary. We found that hair was never as well infiltrated as softer tissue would be by the medium; therefore, the hair was placed directly into Spurr's plastic after cleaning in ethanol. Sections through the middle of the hair shafts were cut at one to five microns on an LKB Ultratome I11 microtome, utilizing glass knives. Some sections were treated with Richardson's stain, but this proved unsatisfactory because many details became too obscure to photograph. Photographs were taken through immersion oil under a cover slip. In order to isolate various cell components of hair (especially cortical cells), a few hairs from each species were boiled gently in concentrated sodium hydroxide on a drop slide (Hausman, 1932). The cells of the hair were teased apart with fine needles and mounted on clean, glass slides in water or immersion oil for study. Whole-mounts were prepared by placing several individual hairs in Permount under a cover slip on a clean slide. These slides were placed in a 40°C oven overnight to dry; this procedure drives out any air bubbles formed when mounting cover slips. Each sample was examined with a Leitz Dialux microscope at various magnifications. Drawings and measurements were taken from three to five hairs. Drawings were made with the aid of a Leitz drawing tube. Average width and maximum and minimym lengths were recorded. Medulla patterns were noted and drawn or ~hotographed for comparison.