RESULTS AND DISCUSSION
Description of Carmen Mountains White-Tailed Deer
Carmen Mountains white-tailed deer were described first by Goldman and Kellogg in 1940. The type specimen, an adult male, was collected by J. M. Dealey on 27 October l939 in Botellas Canyon, Sierra del Carmen, Coahuila, Mexico (Fig. 8).
The Carmen deer is small in size, with moderately spreading antlers which vary considerably in length and form. Many antlers form small baskets, with the main beams curving tightly inward, but sweeping beams forming wider arched antlers also are common. The Carmen deer approaches the color of the Texas white-tailed deer (O. v. texanus) and is described by Kellogg (1956).
Goldman and Kellogg (1940) and Kellogg (1956) reported skulls of Carmen deer to be characteristically smaller than Texas white-tailed deer. In relation to the Coues white-tailed deer (O. v. couesi), Carmen deer reportedly had smaller ears and antlers and a narrower and more slender rostrum. However, data collected during the present investigation revealed that Carmen deer have a broader rostrum than the Coues deer (Table 3), with longer tines and larger antlers than the Coues deer (Table 4). Additional measurements are compared in Table 5. The data presented by Goldman and Kellogg (1940) are from the type specimen, whereas others represent the largest male examined in the series.
TABLE 3. The mean skull size of adult males from 13 O. v. carminis and 11 O. v. couesi collected from their respective ranges.
TABLE 4. Mean comparisons of eight antler measurements from O. v. carminis and O. v. couesi collected in their respective ranges.a
TABLE 5. Comparative measurements between O. v. carminis, O. v. couesi, and O. v. texanus.a
Of the four subspecies of whitetails in Texas, Carmen deer are the smallest and may be smaller than any subspecies on the mainland of the United States. Teer et al. (1965) reported on the field-dressed weights of whitetails from the Llano Basin and Edwards Plateau of Texas. The whitetails in these Texas areas are small, but Carmen deer are smaller.
Ten adults were measured for the present whitetail examination. Adult males averaged 104 pounds (47 kg), and females in the 2.5- to 3.5-year class averaged 66 pounds (30 kg). The average weights of all sex and age classes sampled for Carmen deer are 67 pounds (30 kg) (Table 6).
Distribution and Status
According to Goldman and Kellogg (1940), the Chisos Mountains are the only North American habitat for Carmen deer. Kellogg (1956) included the adjacent Serranias del Burro, the Sierra del Carmen (Fig. 8), and other northern Coahuila mountain ranges as containing Carmen deer but still restricted (Kellogg 1956; Wauer 1973) their distribution in the United States to ranges in Big Bend National Park.
Diagnostic features important to whitetail distribution seem to be free standing water and areas of dense vegetation such as oak stands. In each instance where deer were located, within and outside of the park, these two habitat components have been present.
In the Chisos Mountains, whitetails are plentiful above 4,500 feet (1,373 m). Outside of the main Chisos complex, whitetails are common in the Panther Spring, Rock Spring, Cedar Spring, Ward Spring, and West Hills Spring areas (Fig. 2). These locations also are occupied by the desert mule deer. The lowest elevations Carmen deer were observed in this study were 2,970 feet (906 m) and 4,000 feet (1,220 m). A yearling male was observed at Dugout Wells which is 3 miles (5 km) from Rock Spring, the nearest locale where whitetails are common. The other low-elevation sighting occurred in Panther Canyon. An adult male was alerted at 4,400 feet (1,342 m) and ran down to 4,000 feet (1,220 m) before retreating to higher areas.
Chilicotal Mountain and Burro Mesa (Fig. 2) do not support Carmen deer although both areas are occupied by mule deer. These foothill mesas support vegetation associated with the SotolGrassland Formation on higher elevations, and Desert Shrub Formation plants on the lower slopes. Each range slightly exceeds 4,000 feet (1,220 m). Atkinson (1975) also investigated these ranges without observing whitetails.
The Sierra Quemada (Fig. 2) forms the southern portion of the entire Chisos Range and provides marginal habitat for the Carmen deer. Borell and Bryant (1942) reported whitetails above Smoky Spring (Fig. 2) at about 4,500 feet (1,373 m) elevation. Atkinson (1975) observed whitetails northeast of Punta de la Sierra and between Punta de la Sierra and the Dodson trail (Fig. 2). During the 12 days Atkinson spent in the Sierra Quemada, only six whitetails were observed. During the present study, Goat Mountain, Blue Creek, Trap Spring, Mule-Ear Spring, Smoky Spring, San Jacinto Spring, Fresno Spring (Fig. 2), and the east end of the Dodson trail were visited. Two whitetails were observed along Blue Creek at 4,100 feet (1,251 m). The occurrence of whitetails in the Sierra Quemada probably always has been, and will continue to be, marginal. Harsh environments, limited water and cover, and the presence of mule deer probably contribute to low whitetail numbers in this area.
The northern end of the Sierra del Carmen in Coahuila, Mexico, extends into the park and forms the eastern boundary. This portion of the massive del Carmens is called the Sierra del Caballo Muerto or Dead Horse Mountains (Fig. 8), and has elevations approaching 6,000 feet (1,830 m). Kellogg (1956) and Davis (1957) reported whitetails at the higher elevations of this range, but recent observations have been rare. McBride (1973 pers. comm.) reported their occurrence there in the late 1960s but only in small numbers.
Soon after the present study began, it became apparent that whitetails on some mountain ranges outside the park in Texas and Mexico closely resembled Carmen deer. Further, some of these inhabited areas approached the geographic range of the Coues whitetail in Chihuahua, Mexico, and the Texas whitetail in the Davis Mountains, Texas. Proper identification of members of this deer complex is essential for a description of the geographic range and numerical status of the subspecies in southwestern Texas.
Several mountain ranges outside the park were examined for the occurrence of Carmen deer. Carmen deer habitats were found in the Rosillos and Christmas mountains, and in the Chinati Mountains in Presidio County, Texas (Fig. 8). Although other isolated desert mountain ranges such as the Sierra Vieja, Del Norte, and Dead Horse Mountains (Fig. 8) may contain Carmen deer, the three ranges mentioned earlier were those examined. Mexican ranges examined included the Sierra del Carmen Range.
The Rosillos Mountains lie approximately 15 miles (24 km) north of the Chisos and attain a height of over 5,000 feet (1,525 m). The surveyed area was a 35° north slope dissected with deep washes. Vegetation is similar to the foothill vegetation of the Chisos, with ash (Fraxinus spp.) and oak providing the bulk of heavier vegetation.
Twelve miles (19 km) northwest of the Chisos Range are the Christmas Mountains (Fig. 9). The third range observed within the United States was the Chinati Mountains, which are approximately 90 miles (145 km) northwest of the Chisos. Chinati Peak is over 7,230 feet (2,205 m) (Fig. 10). The area is dissected with heavy oak washes with equally exposed east and west slopes of 25-35° The north side of Chinati Peak was examined. This area contained several small basins surrounded by hogbacks inclining to the larger mass of the range. Vegetation was similar to that of the Chisos, with pinyon pine and oak present.
The Rosillos and Chinati ranges contained free-standing water. Small springs were permanent and flowing during the dry part of the year.
Whitetails were observed on all ranges examined. Two adult females and an adult 8-point male were observed on the Rosillos Range; adult females, males, yearlings, and fawns were seen in the Christmas Mountains; and six adult females, five fawns, and one yearling were observed in the Chinati Range. Approximately 160 man-hours were spent examining these areas for whitetails.
Cranial and antler measurements were taken from adult Carmen, Coues, and Texas white-tailed deer by Krausman et al. (1978). These authors determined that there was sufficient justification for retaining Carmen deer as a separate subspecies apart from the Coues and Texas whitetail. Most cranial measurements revealed a gradual clinal increase, beginning in northern Coahuila and continuing northward into the Davis Mountains north of Big Bend National Park, and led Krausman et al. (1978) to conclude that white-tailed deer occupying the mountain ranges north and northwest of the park are best referred to as Carmen deer, although they show evidence of intergradation between Carmen Mountains white-tailed deer and Texas white-tailed deer. Size (Table 7), coloration, and antler configuration were also similar to the Carmen deer, and the habitat of whitetails in the areas examined was similar to whitetail habitat in the park.
The pattern of variation in antler measurements is somewhat different from that exhibited by cranial measurements. The antlers of Carmen deer in the park are intermediate in size between Coues deer and Texas white-tailed deer (Krausman et al. 1978). However, antlers of deer from the Rosillos (Table 8), Christmas, Chinati, and Davis mountains average smaller in most measurements than those of Carmen deer in the park; antlers of these deer are intermediate in size between Carmen deer and Coues deer although they are more similar to the former than the latter. Antler measurements are much more variable than are measurements of the skull. Furthermore, the size and configuration of antlers in deer is known to be markedly affected by nutritional factors (French et al. 1955) which relate to range conditions. These two factors make antler measurements of less value than cranial measurements in making geographic comparisons.
TABLE 8. Mean comparisons of 8 antler measurements from whitetails in known O. v. couesi and O. v. carminis ranges, and the Rosillos Mountains.a
Classification of whitetails in the Chinati, Rosillos, and Christmas Mountains, and perhaps other scattered ranges such as the Del Norte, Sierra Vieja, and Dead Horse should be framed around the concept that Texas, Coues, and Carmen deer are valid subspecies and that the isolated populations between the ranges of these subspecies are intergrades between Carmen deer and Texas white-tailed deer. Geographical isolates thus occur between the ranges of the recognized subspecies and are characteristically small, with low populations. The assignment of whitetails from the Davis, Chinati, Christmas, and Rosillos mountains to Carmen Mountains white-tailed deer extends the range of this subspecies about 124 miles (200 km) northward into Jeff Davis County and northwestward about 118 miles (190 km) into Presidio County.
Cases of primary intergradation are often caused by related fluctuations of environmental conditions. Suitable habitat of Carmen deer on isolated mountain ranges may be the result of such changes. Wells (1966) has produced evidence that pluvial climate of Wisconsin time allowed for extensive growth of the present montane pinyonjuniperoak woodland zone over much of the available span of elevation in regional lowlands of the Chihuahuan Desert. Over time, the invasion of desert vegetation may have pushed back the pinyonjuniperoak woodlands to their present relic populations with deer following, resulting in their scattered distribution.
Environmental conditions are a most important factor in the development of new forms. Some animals such as members of the family Didelphidae have remained morphologically unchanged over millions of years (Doutt 1955), while some rodents take less than 300 years for subspecific differences to develop (Simpson 1944). For some microgeographic races, the differences may show up in relatively few years.
According to Mayr (1963), populations separated from the parent population can form either another subspecies, reestablish with the parent group, or die out. The possibility of new subspecies formation in the isolated ranges is remote and the desert separating them prohibits reestablishment with the parent populations. Environmental and man-caused impositions placed on Burro Mesa, Chilicotal Mountain, and other hills and mesas close to the Chisos Range have caused vegetational changes and water reduction, resulting in the disappearance of former whitetail populations inhabiting this area. This may also be the fate of small, isolated whitetail populations in the Rosillos, Christmas, and Chinati mountains, and other small ranges in southwest Texas which contain Carmen deer.
White-tailed deer populations in Coahuila may be stable at higher elevations. Deer are hunted heavily by local inhabitants and venison is a main food item when available. Hunting in Mexico is common for sustenance and is a definite part of the residents' living (Taylor et al. 1945; Baker 1956; Davis 1957; Leopold 1959). While in the Sierra del Carmen in August 1973, discussions with local residents revealed that the situation has not changed.
The entire area sampled in the Chinati Mountains was heavily overgrazed by sheep, especially on the lower foothills and slopes. Mule deer also occur in this range. All whitetails seen were close to water and few signs of deer were observed in drier areas. These factors along with lower population levels and hunting pressure place the future of this deer in an uncertain position.
Adequate habitat on the Rosillos Range is limited to north slopes, with oak and ash in the washes. Only one free-running spring was found which apparently runs all year. Deer were observed, and deer remains were collected in the vicinity of this spring only. Again, free-standing water appears to be important. Population numbers are low in this range and it would not be surprising for the few remaining whitetails to die out.
The real-estate interests presently active around the Christmas Mountains include free hunting as a privilege for landowners. Most hunting has been for mule deer in the past 4 years, resulting in an extremely heavy over-harvest. During the 1973 season, very few adult males were taken, but there was a heavy hunter harvest of yearlings. With fewer mule deer, excessive harvests of whitetails may result.
Because of hunting pressure, competition with livestock, natural decimating factors, limited water supply, low population levels, and small range size, whitetails in the small, isolated habitats have a very uncertain future. Ranching activities encourage water production and, as all the whitetails were observed in areas of free-standing water, it appears that this has enhanced their survival. If the water is not maintained, their status will become more vulnerable.
Big Bend National Park provides the most stable environment for the Carmen deer. Hunting and livestock grazing are eliminated. The deer's habitat is stable and human impact on the area is limited at present. Recent decisions by park administrators are aimed at limiting human use of the main wooded Chisos area even further, and excessive use by man of free-standing water has been discouraged since the park's creation. Big Bend is probably the only area remaining where the Carmen deer may continue to maintain a healthy, reproducing population in a natural state.
During the late 1950s, O. C. Wallmo made the first quantitative estimate of deer in Big Bend (Davis 1957). The estimated numbers of whitetails and mule deer on exclusive range were 25 per square mile (2.59 km2) and 7 per square mile, respectively. With a 20 square mile (52 km2) exclusive whitetail range, Wallmo's estimation was 500 whitetails in exclusive habitat.
More recently, the park staff of Big Bend established new transect lines for both deer species in the latter part of 1968 and continued to read them until the middle of 1972. One 0.10-acre (0.04-ha) plot and eight 0.2-acre (0.08-ha) plots were established in exclusive mule deer habitat. Six were in the River FloodplainArroyo Formation, and three were in the SotolGrassland Formation. Whitetail numbers were estimated by nine 0.10-acre (0.04-ha) pellet plot series and two 0.20-acre (0.08-ha) pellet plot series in the Woodland Formation of the Chisos.
The estimates derived in this study were from pellet plot transects established by Atkinson (1975) (Table 9). Figures 11 and 12 show the SotolGrassland and Woodland formations in which some transects were located.
Transect number 1 was the only exclusive mule deer transect. Other transects located in the SotolGrassland Formation contained both mule deer and whitetails but sightings of whitetails were infrequent.
Distinct pH qualities of mule deer and whitetail pellets (Krausman et al. 1974) assisted in determining species occurrence in this formation. A sample of 10 pellet groups was collected on overlapping range transects in February and May 1974, providing 100 pellet groups for analysis. All were from mule deer so transects in the SotolGrassland Formation were eliminated in establishing whitetail numbers. Other areas in this formation, such as Panther and Rock springs, had higher mule deer:whitetail ratios where pellet plots were not established.
Deer estimates from National Park Service data and pellet plots read during this study are presented in Tables 10 and 11. Estimates of mule deer were not attempted over their entire range; the results (Table 11) may be misleading without population estimates that include the rest of their range.
TABLE 10. Estimates of white-tailed deer in Big Bend National Park between 1968a and 1974.
TABLE 11. Estimates of mule deer on the Chisos Mountains foothills between 1968a and 1974.
When the whitetail pellet plot transects were established by Atkinson (1975), a minimum number of reasonably accessible areas were sampled. The areas also were visited frequently by deer, and I suspect that the mean density estimates were too high. Numbers of transects were too low for precise estimates, as reflected in the wide confidence intervals given for the deer estimates (Tables 10 and 11). Many parts of the range not utilized as heavily by deer were not sampled. However, I do feel that the lower limits of the 90% confidence intervals (Table 10) are representative of whitetail numbers in Big Bend. An average whitetail density between 1972 and 1974 of 29 deer per square mile would yield 580 whitetails with a biomass of 1,930 pounds (868 kg) on the 20 square mile (52 km2) exclusive whitetail range.
Although the mean densities may be high, other Texas areas support higher whitetail numbers. Teer et al. (1965) estimated 92 whitetails per square mile between 1954 and 1961 in the Llano Basin and 44 whitetails per square mile during the same period in surrounding areas. Whitetail density on the King Ranch, Texas, of 38 per square mile (Beasom 1974) and of 35 per square mile on the Rio Grande Plain of South Texas (Harwell and Kierce 1972) are more similar to Big Bend's whitetail densities.
Estimates of mule deer around the foothills of the Chisos are higher than in similar areas. Wood et al. (1970) reported 13 deer per square mile for the Fort Sinton mule deer herd in New Mexico, and Truett (1972) found 13 and 11 desert mule deer per square mile on two mountain ranges in southeastern Arizona.
Both mule and white-tailed deer numbers appear to be stable in Big Bend National Park. The apparent population increase in 1972 (Tables 10 and 11) may be related to improved range conditions provided by high rainfall in 1970, 1971, and 1972 (Fig. 6). The importance of environmental factors to the breeding potential of white-tailed deer has been documented by Morton and Cheatum (1946), Ransom (1967), Roseberry and Klimstra (1970), Verme (1969), and Klein (1970). Both deer species may have responded to range conditions enhanced by rainfall.
As the topography ascends from the Rio Grande across the lowlands and desert, washes give way to canyons, and small hills and mesas gradually build up to the Chisos Mountains. Changes in vegetation are as contrasting as variations in topography. In less than 5 miles (8 km) harsh desert environments are replaced by montane woodland associations. Sparse desert vegetation is replaced by dense stands of mesic types and thick stands of pinyon pine, juniper and oak are in the higher elevations of the Chisos Mountains (Fig. 11 vs. Fig. 12).
Quantitative measures of whitetail habitat were from several locales. Areas in SotolGrassland Formations include Green Gulch, Panther Spring, and Pine Canyon, while the Basin, Boot Spring, and the South Rim (Fig. 2) represent the Woodland Associations.
Changes in vegetation associated with elevation, topography, and climate are seen with increasing altitude along Green Gulch as Chihuahuan Desert plants are replaced by woodland species. The lower canyon at 4,000 feet (1,200 m) is represented with mesquite, acacia, basketgrass, sotol, lecheguilla, and mimosa. As the elevation increases, sotol is still common but juniper (J. monosperma) becomes more abundant and scattered clumps of pines and oaks are common.
To the west along Green Gulch at 5,000 feet (1,525 m) elevation is Moss Well where oak is dense and follows the waterways out into the desert (Fig. 13). This area is the lower limit of exclusive whitetail range on the south side of the Chisos.
Panther Spring provides more water than other desert springs in white-tail habitat and the area is the northern extension of whitetail range in the Chisos. Water from the spring runs down the narrow canyon, much of which is rock floored and walled, before breaking out into the desert (Fig. 14).
Pine Canyon is similar to Green Gulch in that vegetational changes are seen as desert shrub plants are gradually replaced by woodlands. Prior to entrance into the narrow canyon, dense stands of sotol are abundant with scattered clumps of oak, juniper, pine, and sumac.
The Basin varies in elevation from 4,500 feet (1,373 m) at the Window to 5,500 feet (1,678 m) behind the lodge (Fig. 2). Abrupt and rolling hills dissected with deep washes draining out through the Window into Oak Creek are characteristic (Fig. 15). The lower areas contain dense stands of whitebrush (Aloysia wrightii) and acacia (Acacia constricta), but as elevation increases pines, oak, and juniper again become more important (Appendixes, I, II, and III).
Boot Spring vegetation was sampled at 6,300 feet (1,922 m) at the top of Boot Canyon (Fig. 2). Trees form a very dense canopy and grasses and succulents are scattered.
Areas sampled on the South Rim varied from 6,500 feet (1,983 m) to 7,500 feet (2,288 m). Large rolling hills with washes draining into Boot Canyon are common. Grasses, forbs, and woody and succulent plants are scattered and not as important as the dense associations of oak, pine, and juniper.
The Shrub Desert Formation contains a LecheguillaCreosotebushCactus Association and although it generally extends to 3,500 feet (1,068 m) it does invade mountainous areas where erosion and disturbances have eliminated grass cover, and where south slopes are exposed to higher temperatures. Creosotebush, lecheguilla, ocotillo (Fouquieria splendens), cactus, fluffgrass (Tridens pulchellus), yucca (Yucca torreyi), candelilla (Euphorbia antisyphilitica), and leatherstem (Jatropha spp.) are characteristic plants (Wauer 1971). Silverleaf, mariola, and guayacan are common, but grass is sparse.
A clear boundary does not exist between the Shrub Desert and SotolGrassland Formations. As the open xeric flats approach the Chisos, more grasses, succulents, and brushy vegetation are supported, typical of the Chihuahuan Desert uplands. Sotol forms a belt around the Chisos up to 5,000 feet (1,525 m) and is referred to as the Foothill Life Belt (Denyes 1956:301). Flats gradually increase to rolling hills butting up against the Chisos, where grasses are more abundant and there is a mixture of plants from both shrub desert and foothills.
The flats leading into Pine Canyon are representative of the SotolGrassland Association. Grasses and forbs having over 5% relative frequency of occurrence include sideoats grama (Bouteloua curtipendula), hairy grama (B. hirsuta), needle grama (B. aristidoides), spreading fleabane, bluestem (Andropogon spp.), threeawn (Aristida spp.), bristlegrass (Setaria spp.), and triadia grass (Tridens spp.). Important woody and succulent plants with at least 5% relative frequency of occurrence are mimosa, basketgrass, cactus, sotol, goldeneye, acacia, and evergreen sumac (Appendixes I, II, and III).
As indicated by Wauer (1971), vegetation of the washes and canyons within the SotolGrassland Formation is a mixture of plants found in associations both higher and lower in elevation. In Pine Canyon washes the most dominant woody vegetation is oak, which is associated with the Woodland Formations. Again, no clear-cut line separates this association from those below and there is an intergradation of plants.
Relatively heavy growths of sotol, lecheguilla, yucca, mimosa, basketgrass, snakeweed, agave, brickellia (Brickellia spp.), whitebrush, and numerous cacti are seen on the foothills and uneven outwash of the Chisos. Waterways support a variety of vegetation, depending on the association they are within, but normally include the more dense stands of mesic plants.
Shrubbery becomes denser in the higher foothills and lower mountain slopes are covered with sumac, ash, mountain mahogany (Cercocarpus spp.), with occasional junipers, pines, and oaks. Of the oaks, emory oak (Quercus emoryi), graves oak (Q. gravesii), and gray oak (Q. grisea) are more common.
Vegetational composition of higher elevations is quite simple. Various grasses such as muhly (Muhlenbergia spp.), grama, pinyon ricegrass (Piptochaetium fimbriatum), and forbs (Senecio millelobatus) provide ground cover. Basketgrass, pricklypear, and mountain mahogany provide much of the scattered woody and succulent layer and pines, oaks, and junipers form dense stands of trees. North slopes, in proportion to the degree of slope, become more woody and less grassy, while some south slopes are almost void of trees. Grass is dominant on lower layers, with sumac, sage (Salvia spp.), lecheguilla, and basketgrass making up the woody and succulent layer.
Several microhabitats on the north slopes in the higher canyons support plant species not found elsewhere. Upper Pine Canyon and Boot Spring contain Arizona cypress, Douglas fir (Pseudotsuga menziesii), ponderosa pine (Pinus ponderosa), oaks, and maple (Acer grandidentatum) because of the cooler, more moist north slopes of the Chisos Mountains.
Only the grosser characteristics are presented in the foregoing description, but it emphasizes both the change in vegetation as altitude increases and the variable habitat of the Carmen deer. Although their habitat is diverse, woody cover is found throughout the habitat.
Based on deer pellet distribution, there were significantly (P<0.01) more whitetails in woodlands than in the lower sotol grasslands. The three mountain ranges examined outside the park that support whitetails were not as wooded as the Chisos (Appendixes IV, V, and VI), but two relationships were found. Carmen deer were associated with (1) dense cover and (2) free-standing water. Oak washes with free-flowing water on the north side of Chinati peak (Fig. 16) supported whitetails, whereas terrain lacking oaks and water was void of these deer. The only whitetails seen in the Christmas Mountains were associated with junipers, oaks, and limited water, and in the Rosillos Mountains whitetails were observed only close to water or dense vegetation. More detailed vegetative data for these three areas are presented in Appendixes IV, V, and VI.
Although the exclusive habitat of mule deer and whitetails are quite different (high canyons, bluffs, and slopes with dense vegetation vs. open, rolling, desert foothills and flats), the question arose as to what separated the two species in the areas of range overlap. Vegetation on the exclusive deer ranges bordering the overlap zones were compared by using a similarity index (Curtis 1959:83). The closer to 100% this index is, the closer are the areas in vegetational composition. High vegetational similarity was found in all areas measured. Important plants used in computing the index were those which contributed at least 5% to the deer's diet or were significant as cover. The similarity of important vegetation in exclusive mule deer and whitetail habitats along Green Gulch was 31%, and 95% and 75%, respectively, along Panther Ridge and the northwest slopes of Panther Ridge. The important point is that in these areas the two deer species are first separated, but the vegetation changes slightly rather than drastically. This indicates that topographic features probably play a major role in species separation. Those areas supporting Carmen deer outside the park are similar to the foothills of the Chisos and support both mule deer and whitetails. Numbers in these areas are not known, but mule deer outnumber whitetails. Mule deer range from the bottom to the tops of the mountains, but whitetails are restricted to densely covered areas with free-standing water. That whitetails prefer more dense vegetation than mule deer is not surprising and has been documented throughout their ranges (Kramer 1972). In Canada, whitetails are associated more with brushy and wooded river flats, coulees with aspens (Populus tremuloides), mixed forests (Soper 1964; Webb 1967), dense thickets of lower valleys, waterways, burned over areas (Cowan and Guiguet 1956), and willow (Salix spp.). Mule deer prefer more open forests, prairie, and badlands.
Similar preferences exist in Washington as whitetails occur more frequently in brushy river bottoms than mule deer, which are more common in open forests and other open habitat (Ingles 1965). In other western states whitetails are associated more closely with woody cover than mule deer (Queal and Hlavachick 1965; Hoffmann and Pattii 1968; Martinka 1968; Kamps 1969). In the southwestern states whitetails sometimes are associated with more open hillsides or flats (Ruhl 1956) but generally prefer dense vegetation afforded by mountains, while mule deer are common in open deserts, foothills, and chaparral types (Borell and Bryant 1942; Cowan 1956; Ruhl 1956; Swank 1958; Anthony 1972; Truett 1972).
Although there are significant differences in habitat preferences, either species can occupy many diverse habitats over their ranges. This suggests wider physiological tolerances than might be concluded from studies made in any one locale (Kramer 1972).
In areas where vegetation was sparse, whitetails sought washes for cover. Of 190 bedding sites examined, 119 (63%) were in washes with dense vegetation such as oak, pinyon pine, juniper, mountain mahogany, sumac, ash, Texas persimmon (Diospyros texana), and desert willow (Chilopsis linearis).
Bedding sites not in washes were associated with the heavier vegetation available, which included pinyon pine, evergreen sumac, oak, single-seeded juniper, mesquite, and cain cholla (Opuntia imbricata) (Fig. 17). Other vegetation which provided bedding cover included grass, snakeweed, mountain mahogany, lecheguilla, goldeneye, sotol, acacia, littleleaf sumac (Rhus microphylla), skunkbush sumac (Rhus trilobata), pricklypear, yucca, algritia (Mahonia trifoliolata), silverleaf, mimosa, eysenhardtia (Eyscnhardtia angustifolia), Mexican buckeye (Ungnadia speciosa), Texas persiminon mariola, and Mormon tea (Ephedra spp.).
The substrata of bedding sites varied from hard ground and rocks to plush, sandy-dead vegetation mixtures up to 6 inches (15 cm) in depth. Many sites appeared to have been used repeatedly and others showed little evidence of prolonged visits. Of bedding areas not in washes, only 4 were not associated with any type of cover, and only 10 were on level ground. Table 12 presents the degree of slope related to bedding areas. Bedding sites were distributed almost equally between north, east, and west exposures. South slopes were avoided; only 6% of the beds were found on this exposure. South slopes were drier and did not provide as much cover as other exposures.
TABLE 12. Slope of bedding sites of Carmen deer in Big Bend National Park.
Major forage classes in this study included browse, forbs, succulents, and grasses. Succulents, which included lecheguilla and cactus, were a major forage class because of the importance to whitetail diets and the high availability on the range. Table 13 lists the volumetric percentages of all forage species obtained from rumens and Table 14 presents food items observed being eaten by deer but not found in stomachs.
TABLE 14. Food items observed to be eaten by white-tailed deer but not found in rumen samples collected between June 1972 and April 1974 in Big Bend National Park.
The diet of free-ranging deer depends on forage availability of which a broad spectrum is available. From plant phenological data, 351 plants representing 74 families were described for the study area.
Overall, browse comprised 35.1% of the diet, followed by succulents (28.1%), forbs (14.3%), and grass (3.5%); 18.7% of the diet was undetermined (Table 13).
Lecheguilla and pricklypear (Fig. 18) were utilized most by whitetails, comprising 17% and 10.9% of the diet, respectively. Other species that made up more than 5% of the diet were acacia (7.4%), oak (6.7%), euphorbia (Euphorbia serrula) (6.6%), and evergreen sumac (5.9%) (Table 13).
Grass utilization was low throughout the year but highest (12%) during developmental stages. Overall, grass was consumed in small quantities by Carmen deer but seasonally may provide significant dietary requirements. Grass consumption was also at a peak during summer months for the Coues whitetail in Arizona (Anthony 1972).
Succulents were utilized more than other forage classes, except in late summer when rainfall was high, and were consumed more on the lower elevations of the range of Carmen deer. At higher elevations more free standing water was available and plants were more lush than at lower levels. Succulent plants provide a source of moisture and although nutrient content of these plants is important, the water in them may be more significant to the whitetail's ecology when free-standing water is not available. Deer collected near springs in October 1973 utilized succulent plants in less than 6% of their diet. The mean monthly rainfall for the preceding 3 months was high at 2.13 inches (5.4 cm), which provided free-standing water. In December of the same year, succulents comprised 70% of the diet when springs were drying up because of low precipitation during the preceding 3 months. Diets of deer collected prior to March rains in 1974 contained over 50% succulents. In January, February, and the first half of March less than 0.15 inches (O.38 cm) of rain was recorded. Occurrence of browse was at a peak in diets during the late summer when available moisture reduced the necessity for deer to obtain water by consuming succulents.
Importance of succulents to whitetails is evident from the analysis but they also may contribute to mule deerwhitetail range separation in Big Bend. Free-standing water in Big Bend, even in wet seasons, is limited. Impressions in rocks, cavities formed by leaves of succulent plants, or washes may hold water for several days, and even the most productive springs produce only a trickling stream. Whitetails in the park and other southwestern areas have a remarkable adaptability to hot, arid conditions. In New Mexico whitetails apparently can survive on moisture present in plants alone (Raught 1967). Water sources in whitetail habitat were all utilized even though some greatly increased vulnerability to predation.
Recent examinations (Knox et al. 1969) failed to show that Rocky Mountain mule deer were less dependent on water than whitetails. Kramer (1972) suggested that "differential distribution of the two species [mule and whitetail deer] in regard to moisture may be due to reasons other than the presence or absence of free water." Field evidence in this study may suggest otherwise.
Observations of white-tailed deer around water sources commonly included drinking activity, whereas observations of mule deer involving water consumption were the exception. Areas with free-standing water in the Chisos, Chinati, Rosillos, and Christmas mountains supported whitetails, whereas there was little sign of whitetails in areas without some free-standing water.
Anderson (1949) and Mearns (1907) often observed desert mule deer passing surface water during dry months without drinking and felt that free-standing water was not required because moisture was provided in lecheguilla.
Sex and Age Ratios
Whitetail sex and age classes are summarized in Table 15. Observer bias is probably the largest limiting factor in calculating accurate ratios since yearlings, does, and sometimes fawns may not be sexed accurately in the field (Leopold 1933:112; Teer et al. 1965). Also, adult males were more conspicuous than other sex and age classes due to their larger size and secondary sexual characteristics. Behavior of the deer, density of the vegetation [especially above 5,000 feet (1,525 m)], and nature of the terrain affected visibility.
TABLE 15. Sex and age classification of whitetails from field observations in Big Bend National Park between June 1972 and April 1974.
Only six newborn fawns were observed during the entire study. Observations of fawns increased in the months following birth as they increased in size and abandoned their seclusive habits. Of three gravid females examined, each had one fetus, and only one corpus lutea was present in ovaries which were examined. These data indicate a low incidence of twinning.
A total of 1,218 whitetails was recorded in 510 separate observations. The mean group size was 2.4 whitetails per group (Table 15). Small group size is common for whitetails in southwest Texas and southeastern Arizona. Anthony (1972) reported the mean group size of whitetails in southeastern Arizona to be 2.3 in the San Cayetano Mountains and 2.2 in the Dos Cabezas Mountains. Groups consisting of more than seven animals were uncommon in the Chisos Mountains.
White-tailed deer are sedentary and occupy the same home range sites throughout the year in the Chisos. Hahn and Taylor (1950) and Thomas et al. (1964) reported the home range of Llano Basin deer in Texas to have a radius of less than 1.5 miles (2.4 km). Adult male whitetails on the Welder Wildlife Refuge occupied up to 800 acres (324 ha) for their home range (Michael 1965). Home range size appears similar for whitetails in the Chisos Mountains.
The white-tailed deer is the least gregarious species of the genus Odocoileus and social groupings are often small (de Vos et al. 1967). However, several group associations of whitetails have been identified. The family group is the core of the social organization and consists of does and yearlings, or does, fawns, and yearlings (Canton 1877; Newsom 1926; Townsend and Smith 1933; Severinghaus and Cheatum 1956; Hawkins and Klimstra 1970). Frequency of family groups for Carmen deer were highest in the summer and constant through the remainder of the year (Table 16). The family groups may be matriarchal, with three generations present (Palmer 1951; Hawkins and Klimstra 1970).
Primary associations involve the sociobiology between the female and fawn or fawns of the year (Hawkins and Klimstra 1970). These bonds were highest following the fawning season but were maintained throughout the year. Montgomery (1959:54) examined the social behavior of whitetails and did not find any unusual disruptions in the primary associations during the breeding season. Observations collected in the present study agree with the above conclusion.
Male groups included one or more males without females or fawns. Hawkins and Klimstra (1970) found buck groups less common than family groups, but this association did not appear to be as stable as family groups. Although my observations show male groups to be more common than family groups (Table 16), observer bias may account for the difference because adult males were more conspicuous than other classes. Male groups were variable, but two to three males per group were most common. Thomas et al. (1965) reported small male groups of two to five animals following the rut on the Edwards Plateau of Texas, and Linsdale and Tomich (1953) found similar associations in mule deer. Male social organization is often hierarchical (Severinghaus and Cheatum 1956; Michael 1966). Hierarchical dominance of male groups was obvious during the breeding season in the Chisos Mountains, and the social group system apparently had been established prior to the rut. In the few instances of intraspecific encounters observed, the presence of a larger member was all that was required to establish dominance. An excellent review of ungulate aggressive encounters is provided by Brown (1971).
Yearling groups included those female yearlings that were observed singly or in groups without adults or fawns. Observations of yearlings were constant throughout the study. Brown (1971) believes this association interacts with a variety of other groups, but strong social bonds do not develop, resulting in transitory relationships.
Random associations consisted of various sex and age classes that grouped together out of some social attraction but did not develop bonds or individual recognition (Dasmann and Taber 1956). The higher incidence of random groupings (Table 16) during the breeding season may be in response to males joining family groups during the rut. Hawkins and Klimstra (1970) in their deer studies found this lasted only a few days and data collected during this examination also indicate these groupings to be more dominant during the breeding months.
Unfortunately, most observations were of disturbed deer, and true social interactions probably were masked. However, no data were obtained to indicate that group associations and social interactions of the Carmen deer differed from the whitetail sociobiology reported in other southwestern areas of the United States (Brown 1971; Hirth 1973; Atkinson 1975).
Although whitetails are generally considered to be crepuscular, individuals may be active during any hour. Three general daytime activity periods of deer have been reported for whitetails (Halloran 1943; Hahn 1949), mule deer (O. h. californicus) (Cronemiller and Bartholomew 1950), and black-tailed deer (O. h. columbianus) (Taber and Dasmann 1958): (1) dawn to midmorning feeding and movement; (2) midmorning to midafternoon bedding; and (3) movement and feeding from midafternoon or late afternoon until dark or longer (Montgomery 1963). Gladfelter (1966) and Howard (1969) found whitetails feeding prior to and past sunset, with three nocturnal bedding periods separated by two feeding periods. The fourth feeding began before sunrise and extended into the early daylight hours. Similar activity patterns were observed in Carmen deer (Table 17). Browsing was common in early morning (0500-0800) and late afternoon (1700-2000) hours, with periods of inactivity during the warmer hours (0900-1600). Deer bedded earlier during winter months in response to cooler temperature, which may be an energy conservation measure.
Nocturnal observations were limited to the hours following darkness until midnight, with the majority observed between 2200 and 2300 hours. While spotlighting from September 1971 to April 1972 (Atkinson 1975; this study), 207 whitetails were observed. Forty-five percent were bedded and 12% were browsing. Twenty-five percent were standing and 18% of the observations involved uncertain activity of the deer. Undoubtedly, the spotlight disturbed many deer and those standing or undetermined may have been involved in other activities.
Undisturbed white-tailed deer were difficult to observe in the Chisos Mountains due to topography and alertness of the deer. Of all whitetails observed, 17.6% were not alerted to the presence of an observer. An additional 5.7% of the observations were of deer that were alerted but continued their activities. The remaining 76.6% of whitetail observations were terminated by the alerted animal leaving the area.
Reproductive Activity and the Fawning Season
Calculation of timing and duration of events in the sexual cycle of the Carmen deer are a composite of data collected from June 1972 to April 1974. Antlers began to grow in the latter half of April and the first half of May, during the primary development phase of reproductive activity as described by Robinson et al. (1965). Antler development continued through September when the bucks began shedding velvet. By the first week of October practically all antlers had a polished appearance. This stage of sexual activity is associated with spermatogenesis and has been described as the "full production stage" (Robinson et al. 1965). Testicles collected from three whitetails showed spermatogenesis was occurring during this time period and some males were sexually active in late September. Bucks were often together in October but showed little aggression toward each other and few observations of sexual relationships with females occurred until the latter part of November.
Late in November, sexual activity was more pronounced. Bucks would form "scrapes" (shallow depressions in the ground formed by pawing and into which they urinated), and increase their sparing and thrashing activity. At this time most bucks had swollen necks. Most sexual activity occurred from mid-December to mid-January. Breeding seasons for the Carmen deer peak approximately one month later than recorded for the northern whitetails (Wislocki 1942:645; Cheatum and Morton 1942), which can be expected due to latitude variations (Cheatum and Morton 1942). In southern Texas, Illige (1951) also found the peak of breeding to occur in December.
Pursuit of females was observed commonly in late December but the peak reproductive activity was between 2 and 12 January. Copulation was never observed, but bucks with their noses to the ground trailing does and bucks chasing females were common during this time period. Some sexual behavior was observed as late as 15 February.
Pregnant females were observed in June and July. Fawns less than 24 hours old were observed on six separate occasions: 7, 21, 22, 24, 25 July, and 1 August. A fetus was collected at 106 days of age, two road-killed fawns were aged, and a gravid female was examined in March. Assuming a 201-day gestation period (Severinghaus and Cheatum 1956), these individuals would have been conceived during the peak sexual activity period. Atkinson (1975) reported similar fawning periods in Big Bend during 1971.
Spotted fawns were common until the end of September, but by mid-October most had completely lost their spots. The latest observation of a spotted fawn occurred on 5 November 1973. The animal was extremely small and did not appear to be more than a week old. If this observation us interpreted correctly, conception dates for some whitetails may be as late as April.
Northern whitetails normally drop their antlers in late December or January following the rut (Wislocki 1942). Whitetail populations farther south generally have later breeding seasons which last longer, with antlers being retained longer. The Carmen deer retained antlers until early March. New growth reoccurred during the latter portion of April when the reproductive cycle began once more. Based on testicular examination, there was no spermatogenesis and sexual activity was quiescent during this "rest phase."
Birch (1957) states that competition occurs when animals of the same or different species utilize a common resource that is in short supply. If the resource is not in short supply, competition occurs when the species seeking that resource harm one another in the process by behavior that is detrimental to survival or reproduction. In this paper competition is as defined by Birch (1957), either through interference or exploitation. Exploitation is the utilization of a resource in short supply and interference operates when interactions between organisms affect survival or reproductive success of the species (Park 1954).
Population control, interspecific competition, and species isolation may all influence species diversity in natural communities (Miller 1967). Interspecific competition cannot be evaluated without an idea of basic niche intersection in areas where overlap occurs and does not occur. Unless it can be demonstrated that species distribution at any level is less than what would be expected from an unrestricted intersection of the overall niche, evidence for competition as a limiting factor is minimal (Miller 1967).
Coexistence is less likely as ecological requirements of two species become more similar and competition becomes more intense. The less similar the two species' requirements become, potential for competition decreases and chances of coexistence increase (Miller 1967).
Competitive displacement in theory is an all or nothing situation; one species exists and the other dies out since different species with identical ecological niches cannot coexist (Lack 1944; De Bach 1966). When species coexist indefinitely, they must have different niches and not be ecological homologues (De Bach 1966).
Competition for food between large ungulates has been examined by several researchers. Martinka (1968) precluded competitive interactions due to the abundance of commonly utilized resources of mule and white-tailed deer in Montana. Morris and Schwartz (1957) demonstrated high utilization of grass in the diets of mule deer and elk. Hill and Harris (1943) showed differences in foods of mule deer and whitetails. Competition was not evident between whitetails and mule deer due to minor use of the common resource by mule deer (Allen 1968). Competition was not evident between whitetails and elk due to light use of forage plants by elk although the same foods were being eaten (Allen 1968). Research by Thilenius and Hungerford (1967) revealed that signs of inadequate food supply were not evident in deercattle areas, indicating that the presence of cattle during the summer did not adversely affect the food supply of the deer.
Competition is not present in the above examples due to the light use of the common resource by one species or an abundance of the common resource. Only a few articles dealing with interspecies relationships conclude that competition has occurred and most of these involve small mammals (Raun and Wilks 1964; Sheppe 1967; Koplin and Hoffman 1968; Morris 1969; Cameron 1971; Grant 1971).
There is little agreement in the literature concerning competition between mule and white-tailed deer. Kramer (1973) stated that whitetails and Rocky Mountain mule deer have coexisted through their evolutionary history and do not competitively exclude each other from sympatric habitat. Kramer (1973) further suggested that optimum habitat for either deer species is void of the other. In a Montana study, Kamps (1969) concluded that forage competition does not exist between whitetails and mule deer but dual use of the range may be more efficient than utilization by a single species.
Anthony (1972) stated that desert mule deer competitively exclude whitetails in Arizona's San Cayetano Mountains and explained that the deer were competing actively for food, but it was a transient phenomenon. Anthony (1972) believed the mule deer would eventually exclude whitetails in this Arizona range.
Mule and white-tailed deer utilized similar foods in Big Bend but forage preferences did not appear to be a major separating factor between the deer habitats (Krausman 1978). The obvious mechanism separating the species appeared to be topography but behavioral interactions, or interference, could not be ruled out.
On sympatric range 1,180 deer were observed: 551 (47%) whitetails and 629 (53%) mule deer. When either species was able to observe or detect the other, the observation was classified as a "dual species interaction" in which 172 (15% of the total) deer were involved: 96 mule deer (8% of the total) and 76 whitetails (7% of the total).
Table 18 lists the species, their activity, distance apart, and type of interaction. Apparent disregard for each other was involved in 65% of the encounters and was the most common behavioral reaction observed. Mutual disregard was common for distances less than 30 feet (9 m). Disregard is as the word implies, and in interactions so classified no behavioral changes due to the presence of the opposite species were detected. Only four instances of active avoidance were observed; all involved an adult being dominant over submissive subadults. In two instances adult male whitetails were walking to a spring as two yearling mule deer were leaving on the same trail. In each instance the adults stopped and stared at the yearling, which avoided contact by departing from the trail and going around the adults in order to proceed. When the mule deer passed, the bucks discontinued their stare and continued walking.
Aggressive encounters and alerted scattering were each observed twice. One aggressive approach involved a bedded yearling and an adult female whitetail. The whitetails bedded 30 feet (9 m) apart. A mule deer doe approached the yearling and departed. The yearling walked to the whitetail doe. A second mule deer doe approached them both at a trot. At this time the adult whitetail got up and with the yearling retreated a short distance. The approaching mule deer then departed as the whitetails returned to their original bedding sites.
The second aggressive encounter involving yearlings (Table 18) included a whitetail female walking slightly to the right and in back of two immature mule deer. After lagging behind, she trotted toward the mule deer, at which time the male charged her with head low and ears back. The whitetail moved to avoid contact and then remained stationary. As the mule deer left, the whitetail stood for a moment and then ran out of sight.
Alarmed animals that had been alerted by observers ran into groups of the opposite species, resulting in deer running out of sight in several directions. This happened on two occasions when whitetails fled into groups of browsing mule deer.
Dual species interactions between whitetails and javelinas were observed on nine occasions. Two cases involved conspicuous watching and two involved disregard. Alerted scattering occurred five times. Table 19 lists species activity and interaction behavior. The instances of scattering occurred due to disturbances made by the javelina and not by the observer. The cases usually involved deer browsing or bedded in thickly vegetated areas and as the javelina groups moved down through the brush, the whitetails departed.
Due to the possibility of competition between mule deer and whitetails in overlap areas, it is of value to know the extent of dual species as sociations in these areas. Dice (1945) referred to coefficients of associations and proposed association and coincidence indexes to measure associations between two species. These have been utilized by McMillan (1953) to relate moose and elk associations on feeding grounds. Both authors (Dice 1945; McMillan 1953) explain the derivation of the indexes and provide information for their application.
The coefficient of association measures the difference between the number of times two species occur together and the number of times they are expected to occur together by chance. The association index measures the amount of association between one species, taken as the base for comparison, and a second species being compared. The third method, a coincidence index, has a value intermediate between the first two indexes and is a measure of the amount of association between both species compared.
Values of the latter two indexes range from 1.0, indicating association, to 0.0, which indicated failure to associate. The computed values relate to the proportional amount of association. Values equal to 1.0 in the coefficient of association would indicate occurrence the same as expected by chance, less than 1.0 would indicate occurrence of association less than expected by chance, and greater than 1.0 would indicate an amount of association greater than expected by chance.
Table 20 lists the measures of association and their values for mule deer and whitetail relationships. The coefficient of association indicates that the two species occur together approximately one-third as many times as expected by chance. Association indexes show that 12% of the observations in which whitetails occur they were associated with mule deer, and the association index of mule deer with whitetails reveals that 22% of the observations of mule deer were associated with whitetails. The tendency reveals limited association between the two as indicated by the low coincidence index.
TABLE 20. Measures of association between whitetails and mule deer in Big Bend National Park.
All of the index values were combined and subjected to the chi-square analysis. The calculated chi-square was 54.02. With one degree of freedom, this value is far above the 5% level of significance, which suggests that lack of association of whitetails with mule deer is due to factors other than chance errors in random sampling. One species or perhaps both actively but subtly avoid each other.
In working with similar species in the Southwest, Anthony (1972) concluded that interference was not of importance. From the behavioral interactions he observed, the infrequent occurrence of interactions and usual nonaggressive nature of both species were not very important in relationships between desert mule deer and whitetails. He did find mule deer to be dominant in all encounters with whitetails when dominance could be determined. Kramer (1973) found that in interspecific encounters between mule and white-tailed deer neither species could be considered socially dominant but rank order of sex and age determined the interspecific hierarchy.
Both authors felt interference did not occur or was not important, but while Kramer (1973) said separation by competitive exclusion was unlikely, Anthony concluded that mule deer would eventually outcompete whitetails in the San Cayetano Mountains in Arizona.
Data collected herein suggest that neither mule deer nor whitetails were socially dominant over the other. More avoidance occurred between species than within, deer occurred more with each other than with other species, and separation by competitive exclusion is unlikely. That the coexistence of whitetails and mule deer, being sympatric over much of North America, rests on habitat difference and preferences (Carter 1951; Martinka 1968; Kamps 1969; Kramer 1973) is also probable in Big Bend National Park.
Ninety-one deer deaths were recorded during the study: 42 mule deer and 49 whitetails. Mountain lions accounted for 34 deaths, vehicles hit and killed 30 deer, and the causes of 26 deaths were undetermined. One death was the result of a whitetail doe breaking her leg when tangled in a fence. Table 21 summarizes sex and age of deer mortality from accidents and undetermined causes. Mortality related to predation will be discussed later.
TABLE 21. Sex and age of deer killed by cars, undetermined causes, and fences in Big Bend National Park between June 1972 and April 1974.
Sexing and aging vehicle-killed animals was often difficult since only portions of the carcass were found. Scavengers, such as coyotes and turkey vultures (Cathartes aura), and high temperatures causing decomposition could reduce an intact carcass to a scattered assemblage of bones in a short period of time. Often, only clues remained as to sex and age if the bodies were not found within a few days. On two separate occasions I received reports of deer killed around 0700 hours and was able to arrive on the scene within 2 hours. Both times all that was found was the backbone and portions of the rear legs. A fawn kill was reported to me one evening, and although I was at the site within 15 minutes, all that remained was a blood spot and part of the hide.
Only two whitetails were reported killed by cars: a 14-month-old female and a 5-month-old fawn of undetermined sex. An adult female died as a result of a broken leg caused when she became tangled in the wire fence surrounding the sewage lagoon located in the Basin. This is one of the few fences in the park. Other fences recently have been installed along the pack trail leading to Laguna Meadow to prevent horses and people from cutting across switchbacks. The effect of fences on the deer will be minimal, however, occasional deaths due to the fences should be expected.
Cause of death of the remaining 24 whitetails was not determined and any reason provided would be speculation only. However, the four dead fawns found were not killed by predators. Three were intact and a fourth consisted of an unchewed skeleton. Fawns have a higher mortality rate than other age classes (Swank 1958:50).
Two adult females died from causes other than predation. Their carcasses were found intact in bedding sites. Five of the remaining carcasses may have been the result of predation, but evidence was too inconclusive to classify them as known predator-killed deer.
Robinette et al. (1954) reported that sick and debilitated deer move downhill where they die. This behavioral characteristic appeared to be operative with the Carmen deer since 17 carcasses were found at the bottom of washes and canyons. Sick animals at the bottom of a wash have little strength with which to move.
The majority of dead mule deer found were killed by vehicles. Only four deaths were unexplained. The higher incidence of vehicle-killed mule deer than whitetails is easily explained. There are over 80 miles (129 km) of paved and well traveled roads in mule deer habitat and less than 3 miles (5 km) of paved roads in whitetail habitat. Although mule deer were hit by vehicles during all seasons of the year, 62% of the subadults were killed during the fawning and breeding seasons. During these times yearling bonds with adults may break down, resulting in more wandering by yearlings.
On two occasions deersnake encounters were observed in Big Bend. Taber and Dasmann (1957) listed rattlesnake bites as a cause of deer deaths, but this type of mortality would be difficult to document under natural conditions. The first encounter involved an adult female whitetail and a Western diamondback rattlesnake (Crotalus atrox). The deer was walking along a ridgetop and was alerted by the snake. She approached the snake rapidly, struck it with her front hooves, and killed it. The second encounter occurred when several whitetails ran out of Panther Canyon. A nearly dead Texas Lyre snake (Trimophodon vilkinsoni) was found in the tracks of the fleeing deer. As the deer departed, one may have accidentally stepped on the snake but it also may have been crushed deliberately. Although the above observations indicate aggressive behavior of deer toward snakes, it is unlikely that snake bites result in, or contribute to, mortality of deer in Big Bend.
Parasites, Disease, and Deer Condition
Forty-eight deer were examined for external parasites: 19 whitetails and 29 mule deer. Fawn, yearling, and adult age classes were represented. Most deer examined were free of external parasites. Twelve whitetails (63%) did not have parasites, six (32%) had one or more winter ticks (Dermacentor albipictus), and a deer nose bot larva (Cephenemyia spp.) was located in one animal (5%). Of the 19 mule deer examined, 18 (62%) were free of external parasites, 9 (31%) had winter ticks (one in this category also had a nose bot larva), 1 (3%) had a spinose ear tick (Otobius megnini), and 1 (3%) was infested with screwworm (Cochliomyia hominivorax).
Winter ticks were the most common for both deer species and were found in and on ears, the neck, shoulder, udder, anus, head, and between the hooves. Only one deer was infested heavily. The shoulders of an adult male mule deer were covered with this parasite. Seventy ticks were removed from a 15.5-square-inch (100-cm2) area representing the infestation. Although the deer was killed by a car, the heavy infestation may have altered his behavior, contributing to a lack of awareness. Excessive ticks may result in death, especially on poor range (Krull 1969:457). Evidence collected, however, does not indicate that the winter tick is detrimental to deer in Big Bend.
Nose bot larva infest the nasal sinuses of deer and are commonly referred to as head- or nose-maggots or bots; most deer are infested to some degree. Bots may be more abundant in deer in Big Bend than indicated by visual examination.
Spinose ear ticks may be a source of irritation causing ear cankers, nervous and digestive disturbances, a lowering of animal condition, and decreased milk flow in lactating animals (Krull 1969:436). Low incidence of occurrence in examined deer indicates that they are of little importance to deer health in Big Bend.
Screwworms, true parasites that attack living animals having fresh lesions on which larvae must feed, have been a problem in Texas and the Southwest (Teer et al. 1965; Krull 1969:350-351), but only one case of screwworm infestation was observed during this study period. This parasite attacked the top of an adult male mule deer's head and the flesh was consumed to the skull.
Although this was the only case of screwworms observed, it may play an important role during other periods when favorable climatic factors enhance the success of screwworms. Teer et al. (1965) attributed low deer productivity in 1955 and 1957 to screwworms in the Llano Basin of Texas, and ranchers stated that the navel of every newborn calf was infested with screwworms within a few hours of birth if the calf was not found and preventive measures taken during the same period.
External parasites did not appear to affect deer adversely in Big Bend, although given certain conditions, the effects of parasites, especially screwworms, could be substantial.
Internal Parasites and Deer Condition
Of all tissues examined, only three mule deer samples suggested parasitic infection but none was found. Internal parasites played a minor role in deer condition, especially for the whitetails, during this study.
Inflammation of the myocardium (myocarditis) was found in two mule deer and one whitetail. In each case, the condition was subacute and minimal. Although the etiology of the myocarditis was not determined, minimal lesions in the hearts of four mule deer may be of the same or similar etiology.
Significant lesions were found in only two mule deer: one with myocarditis, the other with biliary proliferation and phyelonephritis, inflammation of both the lining of the pelvis and parendyma of the kidney. In the latter case, the animal was collected in a weakened condition. There was a proliferation of bile ducts in the liver, and the presence of hepatocytes suggests a prior toxic hepatitis. The biliary proliferation was probably secondary to infections from toxin and the phyelonephritis.
Omental, kidney, heart fat, and marrow from 16 whitetails and 20 mule deer were examined visually. The results are presented in Table 22.
Laboratory tests, field examinations, and observations of live deer indicated that both mule and white-tailed deer were not affected adversely by parasites or disease and were in good physical condition from 1972 to 1974. Atkinson (1973 pers. comm.) claimed deer were also in good physical condition during 1971.
Predators and Deer
Interest in predation recently has provided a body of knowledge which has enhanced the understanding of predatorprey relationships. Hornocker (1970) and Seidensticker et al. (1973) have published accounts of mountain lion population mechanics in the Idaho Primitive Area, and Knowlton (1964), White (1967), and Beasom (1974) have discussed the effects of coyotes and bobcats on deer populations. That some carnivores are very effective in preying on ungulate young has been demonstrated with coyotes and antelope in Texas (Jones 1949) and Arizona (Arrington and Edwards 1951), and with coyotes, bobcats, and whitetails in south Texas (Knowlton 1964; White 1967; Beasom 1974).
Controversy over predators in the Big Bend region of Texas is legion, with bias against carnivores. Many ranchers and hunters shoot predators on sight, and protected areas are limited. When Big Bend was established as a national park, the area was called a protected breeding ground for "livestock-killing vermin." National Park Service personnel adopted the common philosophy that predators controlled the deer. Murie (1954:120) found several lion-killed whitetails during his visit to Big Bend and felt that the lions were keeping the deer under control. Wauer (1973:93) recently has proposed the same "balance of nature."
Predatorprey relationships are being studied, but little information is available on lions in the Southwest. Atkinson (1975) reported on predatorprey relationships in Big Bend during 1971, and during the time of the present study, Roy McBride of Sul Ross State University investigated mountain lion movements in the Big Bend region. Examinations of lion populations are taking place in other southwestern states, but more data are needed, especially in undisturbed areas. Big Bend National Park is such a large, undisturbed area where predators are protected. The following data are presented to better understand the relationships between carnivores and prey in natural habitats.
The ungulates in Big Bend are affected by lions, coyotes, and bobcats. Although lions are listed as uncommon for the park, they are fairly numerous in the Chisos Mountains and adjoining mountains and hills. As mentioned earlier, numerous lions were killed in the Chisos during the 1920s and 1930s. This suggests that the population has been augmented by individuals coming in from Mexico and the surrounding ranges in Texas. Although the National Park Service records report as many as 40 lions in the park in 1952 and 1953 (Table 1), more widely used figures are between 6 and 12 lions at any one time (Wauer 1973; McBride 1974 pers. comm.). To date, an accurate census has not been made and sufficient data have not been collected to estimate the number of lions in Big Bend. Visitor sightings of lions in the park between 1952 and 1974 have ranged from no observations to a high of 61 observations in 1953 (Anon. 1945-present). In 1956, a government trapper reported killing 40 lions in the Rosillos Mountains in a 14-day period. The reliability of these data is questionable although they indicate an abundance of lions. During the 1973 hunting season, at least seven lions were killed by hunters in the Christmas Mountains, and others may have been killed but not reported.
Data on coyote and bobcat numbers in the park also are scanty. Bobcats are less abundant than coyotes. Coyotes are common and have been estimated at 400 for the entire park (Anon. 1944-73). Without more accurate data on predator numbers, it will be difficult to evaluate the ungulate population dynamics in relation to predation.
Mortality of Predators
No deaths of bobcats were discovered during this study, and the few coyote deaths reported were caused by vehicles. Very few cases of natural mortality of lions have been reported since the park's conception. In 1967, a yearling was drowned in Ernst Tinaja (Anon. 1945-present), and only one natural death was recorded during this study. Cause of death was not determined but tooth wear was excessive, indicating old age.
Frequency of Lion Kills
Turkey vultures often indicated the presence of a kill, but most were discovered during routine field activities. Only when signs were conclusive were kills attributed to lions. Signs included scats (Fig. 19), scrapes (Fig. 20), covered carcasses (Fig. 21), removed stomachs, tracks, and drag marks. Most scrapes were found in washes and at wash junctions. Others were made in dead sotol stumps and other plant litter. Of 10 scrapes measured, the average radius was 6 inches (15 cm). Most kills were found in rough canyons and washes and were covered under vegetation such as evergreen sumac, mesquite, pines, juniper, whitebrush, and oak branches. Although vegetation normally was used to cover the remains of prey, rocky outcrops forming small caves were sometimes used to cache remains of prey.
Lions normally do not consume viscera of their prey, and on many occasions the entire stomach had been removed and buried 6-15 feet (2-5 m) away from the rest of the carcass.
Kills were found throughout the study, but killing frequency could not be determined. From June to October 1973, four lion kills were found at monthly intervals in Panther Canyon, but this indicates only lion activity since all the kills in the area were not located, and the number of lions involved was not known. Seidensticker et al. (1973) demonstrated that although over a given year kills by resident lions are made randomly in their home ranges, over the years there are areas where kills are made more frequently.
Panther Canyon provided an advantage to lions when pursuing whitetails. Deer use the spring at the canyon's bottom for water, but sheer rock slopes and the canyon floor restrict rapid movement. For a deer to slip and fall when alarmed in the area is common.
Hornocker (1970) found that lions in the Idaho Primitive Area completely utilized each kill, unless disturbed, which maximizes expended energy in hunting and killing and minimizes potential danger in attacking large prey animals. Generally, the same can be applied to lions in Big Bend when warm weather does not cause rapid decomposition of killed prey. On warm days microtemperatures are often lower than ambient temperatures, and canyons are often cooler. A kill found in the hot month of June in a canyon was eaten completely, whereas other kills made on exposed areas were consumed only partially. This kill, an adult female whitetail, was covered after a portion of the shoulder had been eaten and the viscera removed. The carcass was moved 75 yards (69 m) down the canyon from point of attack. On the following night, it was moved 17 yards (16 m) and the heart and left shoulder were eaten. A third meal was obtained the next night when the right shoulder and flank, front leg, neck, and head were consumed. The deer had been moved 50 yards (46 m) down the wash and left uncovered. Remaining flesh was dark red and not rancid. Similar feeding patterns have been reported by Robinette et al. (1959). Movement of the carcass after each feeding may have been in response to disturbance caused by me.
On another occasion an adult male whitetail had been killed and entirely consumed within 12 hours except for the stomach, intestines, and larger bones. A female with two kittens was reported in this area and may have been responsible.
Other lions ate only shoulder portions of their kills, covered them, and failed to return. This type of behavior has also been reported by McBride (1973 pers. comm.). Warmer southwestern climates may increase this type of feeding activity due to decomposition, but would reduce efficient resource utilization. On one occasion three adult female mule deer had been killed in the same area during October. In all cases the stomach was buried away from the carcass, and after portions of the shoulder were consumed, the remainder was covered with available vegetation where the bodies decomposed.
Lions in the Idaho Primitive Area killed deer every 10-14 days. Other mammals supplemented the summer diet thereby reducing ungulate kills (Hornocker 1970). Evidence of predation on small ungulates is difficult to obtain since they are often consumed entirely (Young and Goldman 1946:126; Hornocker 1970). During this study the kill remains of a doe and fawn were discovered. A small portion of hide and hair was all that remained of the fawn.
Robinette et al. (1959) suggested that lions kill a deer per week in the winter in Utah and southwestern researchers (McBride 1973 pers. comm.) claim lions kill a deer every 7-10 days. McBride (1973 pers. comm.) felt that it was not uncommon for more than one kill to be made during a week; his comments are supported by limited observations reported by Young and Goldman (1946), suggesting that adult lions have killed about every 3 days in the Southwest.
Table 23 is a summary of all lion kills found during the study. One of the javelina known to be killed by a lion was an adult with malformed hooves which affected locomotion. Although only two kills of javelina were found, they did constitute part of the lion's diet. Hornocker (1970) suggested that the tight-knot group behavior of bighorn sheep (Ovis canadensis) in Idaho allowed them to cope well with lions. Since javelina are very social, a similar mechanism may reduce their vulnerability to predation. Also, the physical structure of javelinas may assist in survival against lions. Their thick layer of bristly hair and hide, short neck, and thick skull provide protection. Canines of the javelina are well developed and can severely injure an attacker. Energy expenditures of capturing prime adult javelina are probably too great to justify the effort from the standpoint of predators. One piglion encounter was noted however. An adult javelina in poor physical condition was sacrificed and examined. Claw and puncture marks were on each side of the shoulders, as if attacked from behind, and the rear leg was broken. A lion was responsible for the injury, but the important point is that the pig was not consumed, so the lion's efforts were wasted. Similar encounters with prime adult javelinas may influence lion selection against them.
TABLE 23. Sex and age of whitetails, mule deer, and javelinas known to have been killed by mountain lions in Big Bend National Park between June 1972 and April 1974.
Nineteen whitetails of known ages were killed by lions. Twelve (63%) were 2-years-old or less, five (26%) were between 3 and 7.5 years, and two (11%) were over 8 years. As found by Hornocker (1970), the majority of deer killed were young or old animals (74%). More males 2 years or older were killed than females of the same age class (5 vs. 2). The higher incidence of males killed by lions (Table 23) may be due to their wandering nature which brings them in contact with lions, and reduced alertness and physical condition during the rut, as suggested by Robinette et al. (1959) and Hornocker (1970).
Food Habits of Predators
Lions: Studies of food habits of cougars in the western United States have been published by Hibben (1939), in Utah and Nevada by Robinette et al. (1959), in Idaho by Hornocker (1970), and in British Columbia by Spalding and Lesowski (1971) (Table 24). There is no published study of food habits of cougars in Texas' Big Bend region.
Table 25 lists all food items found in lion scats. Deer comprised over 70% of the diet throughout the year, and as much as 90% in the late summer of 1972. Next in importance was javelina, followed by a number of less important items. Together, deer and javelina made up 85% of the lion's diet (Fig. 22).
TABLE 25. Food items in 161 mountain lion scats from Big Bend National Park.
Although grass content was high in several scats, the high frequency of occurrence of vegetation was probably due to accidental consumption.
As in other studies (Dixon 1925; Hibben 1939; Young and Goldman 1946; Robinette et al. 1959; Hornocker 1970; Spalding and Lesowski 1971), deer (or deer and elk as reported by Hornocker) furnished lions with more food than all other prey species combined. Schwartz (Robinette et al. 1959) collected a small sample of scats in Washington and found varying hare (Lepus americanus) to be more abundant than deer and Atkinson (1975) reported deer abundance in scats to be less than 20% during 1971 in Big Bend National Park. Small samples in both cases may not have represented preference of cougars.
Hornocker's (1970) data indicated complete prey utilization in Idaho, and Seidensticker et al. (1973) documented lions feeding only once on a kill and then moving on if disturbed. As mentioned earlier, similar instances were recorded in this study. If lions feed on a kill and then move long distances as suggested by McBride (1974 pers. comm.), fecal deposition would not necessarily be in the same locale as a kill. Lion movement in Big Bend is longitudinal from mountain range to mountain range, and a lion in the park one day may be 25 miles (40 km) or more removed the following day (McBride 1973 pers. comm; Scudday 1975 pers. comm.). Of particular interest is the fact that no domestic livestock remains were found in any scat even though livestock is consumed outside the park. Most scats were collected deep in the park's interior and remains may have been deposited prior to a lion's arrival in the collection areas.
Coyotes and bobcats: The importance of lagomorphs to bobcats and coyotes has been documented (Young 1958:70-77; Knowlton 1964; Bailey 1972). Lagomorphs were the prey most commonly consumed for both bobcats and coyotes during this study (Tables 26 and 27).
TABLE 26. Food items in 128 bobcat scats from Big Bend National Park.
TABLE 27. Food items in 245 coyote scats from Big Bend National Park.
Fifty-eight percent of the coyotes' diet consisted of equal parts of rabbits and vegetation (Table 27). Texas persimmon was the most commonly utilized plant. The fruits of this plant were ripe in late May and consumed through December. Their importance to diets of coyotes probably reduced the impact of coyote predation on small mammals which made up 17% of the diet. Deer made up 15% of the diet, and javelina, 7%.
Various small mammals (Table 26) comprised 28% of the food items found in 128 bobcat scats, and the third most abundant food was deer, making up 22% of the diet. Vegetation and javelina were of minor importance to bobcats.
Similar small mammals were eaten by bobcats and coyotes during this study and the higher incidence in bobcat scats may be correlated to the increased use of Texas persimmon by coyotes. Ecological and behavioral differences permitted these predators to use the same area and food sources in different ways in Idaho (Bailey 1972), and similar mechanisms may be operative in Big Bend.
Effects of Predators on Deer
That the diet of coyotes and bobcats consisted mainly of small mammals, while mountain lions consumed primarily larger ungulates, emphasizes the fact that relative prey size is a factor in predatorprey relationships as suggested by Hornocker (1970). Although Bailey (1972) found no evidence of bobcat predation on deer, the killing of deer by bobcats has been reported (Young 1928, 1958; Marston 1942; Dill 1947). Hamilton and Hunter (1939), Pollack and Sheldon (1951), Rollings (1945), and Knowlton (1964) found deer to comprise 20-30% of the bobcat's winter diet, much of which may be carrion. Data collected in this examination approximate this figure, although rabbits comprise the staple food (Fig. 23). Lack of snow and easy access to rodents may lessen the importance of deer to bobcats. Knowlton (1964) found that bobcats were not a source of attrition on deer in south Texas and although data in this study are inconclusive, I do not feel that bobcats are detrimental to deer populations in Big Bend even though they contribute to mortality.
Circumstantial evidence that juvenile mortality among North American deer is high and represents one of the major factors determining population density, especially in unhunted populations, has been compiled by Taber and Dasmann (1957:238-240), Brown (1961:61), Knowlton (1964), and Cook et al. (1971). The latter researchers concluded that coyote predation on fawns was a major factor in the stability of the dense and healthy whitetail population on the Welder Wildlife Refuge, Sinton, Texas. Knowlton (1964) showed that on the same area the amount of deer eaten by coyotes rose sharply during the fawning season in June. Diets of coyotes at this time consisted of more than 75% deer, mainly fawns. Increases of deer in the diets of coyote during the fawning period were also reported by Salwasser (1974) in central California. Figure 24 shows an increase of deer in coyote scats during the 1973 fawning period, but age of deer consumed is undetermined.
Coyotes and bobcats are responsible for fawn deaths in Big Bend but the impact on the population is probably minimal. Of coyote scats containing deer remains, only 11% had hooves or hair that could be identified conclusively as fawns.
As mentioned earlier, coyotes could reduce a road-killed animal to a pile of bones in less than 2 hours. Most of the deer in coyote scats may have been road-kills or other carrion. It is well known that the coyote is an opportunistic feeder that will eat the most easily obtainable food.
Insects and fruit are of ecological importance to both the coyote and prey (Knowlton 1964). Consumption of these items may relieve predatory pressure on prey species during periods of greater vulnerability. Since Texas persimmon was consumed more than any other plant material by coyotes, it may play an important ecological role to deer in areas of coyote occurrence. Coyotes were observed, but uncommon, above 5,000 feet (1,525 m) and probably exert a greater influence on mule deer than on whitetails.
The total number of lions preying on the Carmen deer must be considered in order to assess the impact of lion predation on whitetails, but population dynamics of lions in the Chisos is unknown. The park supposedly supports 812 lions at any one time (McBride 1973 pers. comm.; Wauer 1973), but this is an opinion not based on quantitative data. With an average adult lion's home range of about 13 square miles (34 km2) (Hornocker 1970), it is realistic to assume that at least three resident lions inhabit part of the Chisos in their home ranges which radiate out to the lower hills and desert. Assuming that the adults supported two yearlings which ate the same amount as one adult, and that transient lions were equivalent to one resident adult present at all times, five adult lions would be exerting predatory pressure on the Carmen deer throughout the year. Historically, large lion numbers have been removed from the Chisos, suggesting a large number of transients.
The average live weight of Carmen deer is 67 pounds (30 kg). If large predators consume 70% of their prey as suggested by Schaller (1967) and Hornocker (1970), 47 pounds (21 kg) would be consumed from each deer kill. Assuming a 5 pound (2.3 kg) meat requirement per lion per day (Hornocker 1970), each lion would require at least one Carmen deer every 10 days. Five adult lions exerting predatory pressure on deer throughout the year would remove 183 deer or 8,601 pounds (3,905 kg). If proportions of mule deer to whitetails in lion kills are accurate, then it can be speculated that 126 whitetails are removed from the population by lions each year.
Limitations are obvious in this discussion: Predator numbers and dynamics are poorly understood, all kills were not located, killing frequency was based on studies in other areas, and the percentage of whitetails in the diet of lions was estimated roughly. Data collected provide a starting point for future investigation. Since this is the first examination on the subject in Big Bend National Park, only a base has been established. Additional information on numbers, movements, food habits, habitats, and general behavior patterns of coyote, bobcat, and mountain lion in the park will allow better understanding and evaluation of predatorprey relationships in unexploited areas of the Southwest.
Last Updated: 08-Oct-2008