Eutamias minimus operarius Merriam Least Chipmunk

_Specimens examined._--Total, 17: North Rim above Morfield Canyon, MV 7856/507; Morfield Canyon, 7600 ft. (obtained on Nov. 4, 1957), 75976; Middle Well in Prater Canyon, 7500 ft, MV 7855/507; Prater Canyon, 7600 ft., MVZ 74414; Park Point, 8525 ft., 69267-69270; 1/4 mi. S, 3/4 mi. W Park Point, 8300 ft., 69271-69272; Sect. 27, head of east fork of Navajo Canyon, 7900 ft., 69273; Far View Ruins, 7700 ft., 69274-69275, and two uncatalogued specimens in preservative; 3 mi. N Rock Springs, 8200 ft., 69276-69277.

Five of the fourteen specimens of known s.e.x are females, all of which were taken in August and September, and none of which is recorded as having contained embryos. The skulls of the eight August-taken specimens also suggest that young are born in late spring or early summer: the largest skull had well-worn teeth that might indicate an age of more than one year; four others had complete adult dent.i.tions that were barely worn; and three had not yet acquired complete adult dent.i.tions.

The records of _E. minimus_, like those of _Spermophilus lateralis_, indicate greatest abundance in the higher parts of the Mesa Verde and in areas of predominantly brushy vegetation.

Eutamias quadrivittatus hopiensis Merriam Colorado Chipmunk

_Specimens examined._--Total, 13: Prater Canyon, 7600 ft., MV 7838/507; Lower Well, Prater Canyon, 69278; Park Headquarters, MV 7889/507; near the old Park Well, 7300 ft., 5468 in Univ. of Colorado collection; Utility Area, 5469 and 5470 in Univ. of Colorado collection; Spruce Tree House, 4352-4355 in Denver Museum; Mesa Verde, 25 mi. [by road] SW Mancos, 149080-149081 USNM; Square Tower House, 7000 ft., 5467 in Univ. of Colorado collection.

Although both species occur in some of the same areas, _E. q. hopiensis_ is more abundant than is _E. minimus_ in stands of pinyon and juniper, along cliffs, and at low elevations. (A specimen of _hopiensis_, MV 7849/507, from 3 mi. S of the Park boundary where the 6000 foot contour line cross the Mancos River is indicative of the occurrence at low elevations.)

Th.o.m.omys bottae aureus J.A. Allen Botta's Pocket Gopher

_Specimens examined._--Total, 35: Prater Canyon, 7600 ft., 74408-74410 MVZ; Upper Well, Prater Canyon, 7575 ft., 69279; 1/4 mi. N Middle Well, Prater Canyon, 7500 ft., 69280; Middle Well, Prater Canyon, 7500 ft., 69281-69285, 75977; Morfield Canyon, 7600 ft., 75978; 3/4 mi. S, 1-3/4 mi. W Park Point, 8000 ft, 69286-69288; 1-1/4 mi. S, 1-3/4 mi. W Park Point, 8000 ft., 69289; 1-1/2 mi. S, 2 mi. W Park Point, 8075 ft., 69290; Sect. 27, head of east fork Navajo Canyon, 7900 ft, 69291-69292; 1/2 mi. N Far View Ruins, 7825 ft, 69293; Far View Ruins, 7700 ft., 69294, MV 7852/507, 7853/507; 3 mi. N Rock Springs, 8200 ft., 69295-69298; 2-1/2 mi. N, 1/2 mi. W Rock Springs, 8100 ft., 69299-69301; 2 mi.

N, 1/4 mi. W Rock Springs, 69302-69303; 1 mi. NNW Rock Springs, 69304; 1/2 mi. NNW Rock Springs, 69305; Mesa Verde, northern end, 8100 ft., 149087 USNM.

The pocket gophers of the Mesa Verde and vicinity are of one species, _Th.o.m.omys bottae_. The distribution and variation of this species in Colorado have been studied recently by Youngman (1958) who referred all specimens from the Mesa Verde to _T. b. aureus_. He noted that some specimens have dark diffuse dorsal stripes that are wide in specimens from the Mancos River Valley. The generally darker color of the specimens from the Mancos Valley as compared with that of specimens from on the Mesa was noticed in the field, and is another example of the local variability of pocket gophers. The nine specimens listed by Youngman (1958:372) as from "Mesa Verde National Park," Mancos River, 6200 ft., are not here listed among "specimens examined" because possibly some, or all, of the nine were trapped on the east side of the River and therefore outside the Park. None was, however, farther than 30 yards east of the Park.

In the Park, pocket gophers occur both on mesa tops and in canyons. Most of the localities listed above and others at which mounds were seen are areas of disturbance such as the old burn on Wetherill Mesa, the rights of way for roads, the river valley, and the grazed floor of Prater Canyon. Little evidence of pocket gophers was found on unusually rocky slopes, steep slopes, or in stands of pinyon and juniper or in relatively pure stands of oak-brush. In addition to workability of the soil, the presence of herbaceous plants, many of them weedy annuals, is probably the most important factor governing the success of pocket gophers in a local area. No female was recorded to have contained embryos, but two had enlarged uteri or placental scars. This fact and the capture of nine half-grown individuals indicate breeding prior to late August when most specimens were trapped.

Dipodomys ordii longipes (Merriam) Ord's Kangaroo Rat

Kangaroo rats have been seen crossing the highway in the Park less than one mile from the Park entrance by Jean Pinkley.

Castor canadensis concisor Warren and Hall Beaver

In 1935 Quaintance and White spent June 16 to June 20 in the Mancos River Bottoms at the mouth of Weber Canyon, looking for sign of fresh beaver work. They found none. Annual Animal Census Reports include the following information based on patrols along the Mancos River at the east boundary of the Park: 1937--estimate 4 beaver present, 1938--8, 1941--numerous bank burrows, 1942--uncommon, 1944--uncommon, 1945--most concentrated at southeast corner, 1946--runs and two small dams seen (flood had washed out larger dams), 1947--only in 1-1/2 miles north of boundary with Ute Reservation, 1949--two separate colonies (each with dams and one with a large house), 1950--none, owing to drouth and diversion of water upstream completely drying the river at times, 1951--none, 1953--present, 1955--present. On the Mancos River, 6200 ft., in late August, 1956, sign of beaver was abundant, numerous trees had been cut but none within a week, and a bank den was found on the west side of the river extending back 50 feet from the stream and caved in at three places. In 1959 dens were still present.

Reithrodontomys megalotis aztecus J.A. Allen Western Harvest Mouse

_Specimens examined._--Total, 38: North end Mesa Verde National Park, 7000 ft, 75984-75986; Park Point, 8525 ft., 69316-69317; Far View Ruins, 7700 ft, 69318-69319, 79220, MV 7897/507, and 23 uncatalogued specimens in preservative; 3 mi. N Rock Springs, 8200 ft., 69320-69321; 2 mi. NNW Rock Springs, 7900 ft., 69322-69323; 1 mi. NNW Rock Springs, 7600 ft., 69324; 1/2 mi. NNW Rock Springs, 7500 ft., 69325.

The specimen listed last (69325) was an adult male recovered from the stomach of a small (snout-vent length 334 mm., wt. 26.0 gms.) _Crotalus viridus_ that was trapped in a Museum Special mouse-trap on a rocky slope mostly barren of vegetation. The availability of samples taken in August (by Anderson in 1956), in September (by Shepherd in 1958), and in November (by Alcorn in 1957) makes the following comparison of age and reproductive condition possible. The sample from November includes some specimens from outside the Park as follows: 1 mi. W Mancos, Colorado, 75979-75983, and 2 mi. N La Plata [not shown on Fig. 2], San Juan County, New Mexico, some 18 miles southeast of the Park, 75987-76000.

The data shown in Figure 3 indicate that females are pregnant at least from in August into November. A smaller percentage of females was pregnant in November than in August or September. The fact that all females more than 130 mm. long were pregnant in September suggests an autumnal peak in breeding activity. A change in the ratio of small individuals (less than 130 mm. in length) to large individuals (130 mm.

or more in length) is indicative of a sustained breeding period throughout the time shown. In August the ratio was 1 to 2.3, in September the ratio was 1 to 1.2, and the ratio was 1 to 0.7 in November. The western harvest mouse is found usually in gra.s.sy areas.

Peromyscus boylii rowleyi (J.A. Allen) Brush Mouse

_Specimens examined._--Total, 14: North end Mesa Verde National Park, 7000 ft., 76002-76003; Far View House, 7700 ft., MV 7851/507, 7854/507; Far View Point, 5 uncatalogued specimens in preservative; 1/2 mi. N Spruce Tree Lodge, 34742; 25 mi. [by road] SW Mancos, 149094 and 149096 USNM; Oak Tree Ruin, 6700 ft., MV 7870/507; and Cliff Palace, 6800 ft., MV 7864/507.

The specimens were taken in August, September, and November. One adult female trapped on September 10, 1958, had six embryos.

Peromyscus crinitus auripectus (J.A. Allen) Canyon Mouse

_Specimens examined._--Total, 3: Mesa Verde [Spruce Tree Cliff Ruins], 149095 USNM; Balcony House, MV 7865/507, 7866/507.

Peromyscus maniculatus rufinus (Merriam) Deer Mouse

_Specimens examined._--Total, 396: North end Mesa Verde National Park, 7000 ft., 76004-76100; Prater Canyon, 7600 ft., 76101-76144, MV 7839/507, 7840/507; Upper Well, Prater Canyon, 7575 ft., 69328-69329; Morfield Canyon, 7600 ft., 76145-76184; Park Point, 8525 ft., 69330-69342, 69344-69360; 1-1/2 mi. E Waters Cabin, 6400 ft. (labels on some specimens read "West Bank Mancos River, Northeast side Mesa Verde National Park"), 69361-69376, 76185-76204; Sect. 27, head of east fork Navajo Canyon, 7900 ft., 69377-69380, 69422-69426; 3 mi. N Rock Springs, 8200 ft., 69403-69410; 2 mi. NNW Rock Springs, 7900 ft., 69411-69412; 1 mi.

NNW Rock Springs, 7600 ft., 69413-69418; 1/2 mi. NNW Rock Springs, 7500 ft., 69419-69421; Far View Ruins, 7700 ft., 69386-69402; Far View Point, 76530-76531, 79221 and 90 uncatalogued specimens in preservative; Mancos River, 6200 ft., 69382-69385; back of Park Museum, 6930 ft., MV 7857/507; Mesa Verde, 25 mi. [by road] SW Mancos, 149093 USNM; Cornfield, MV 7878/507.

The most abundant mammal is the ubiquitous deer mouse. Series of specimens taken in August (by Anderson in 1956), in September (by Shepherd in 1958 and 1959), and in November (by Alcorn in 1957) make possible the following comparisons of age, reproductive conditions, and molts.

The specimens obtained in August and November were placed in five categories according to age (as judged by wear on the teeth). These categories correspond in general to those used by Hoffmeister (1951:1) in studies of _Peromyscus truei_. From his descriptions I judge that wear in _Peromyscus maniculatus_ differs from wear in _Peromyscus truei_ in that the last upper molar is not worn smooth before appreciable wear appears on the first two molars, and the lingual and l.a.b.i.al cusps wear more nearly concurrently. The five categories differ as follows: category 1, last upper molar in process of erupting, showing no wear; category 2, some wear apparent on all teeth, but most cusps little worn; category 3, greater wear on all teeth, lingual cusps becoming rounded or flattened; category 4, lingual cusps worn smooth, l.a.b.i.al cusps show considerable wear; category 5, all cusps worn smooth. The condition of the pelage was noted for each prepared skin. Hoffmeister (_op. cit._: 4) summarized changes in pelage that he observed in _Peromyscus truei_, and he summarized earlier work by Collins with _Peromyscus maniculatus_. In _P. maniculatus_ a grayish juvenal pelage is replaced by a postjuvenal pelage in which the hairs are longer and have longer, pale, terminal or subterminal bands giving a paler and more buffy or ochraceous hue to the dorsal pelage. The postjuvenal pelage is replaced by an adult pelage that is either brighter or, in some cases, is not distinguishable with certainty from the postjuvenal pelage. Not only is the juvenal pelage distinguishable from the postjuvenal pelage, but the sequence of ingrowth of postjuvenal pelage follows a regular pattern that is usually different from that of subsequent molts. The loss of juvenal hair is less readily observed than the ingrowth of new postjuvenal hair on account of the greater time required for the growth of any individual hair than for the sudden loss of a hair.

Molt was observed in some individuals no longer having juvenal pelage; some new pelage was observed on the skins of seven mice collected in August. Each of these was in category 4 or 5 and probably had been born in the previous calendar year. These seven molting individuals make up nearly 17 per cent of 42 individuals that had completed the juvenal to postjuvenal molt. In November, 80 per cent of individuals (92 of 115) that had previously obtained their postjuvenal or adult pelage were molting. These mice were in age-categories 3, 4, and 5. Some of the individuals in category 3 were developing new hair beneath a relatively unworn bright pelage that I judge to be an adult pelage rather than a postjuvenal pelage. If this judgment be correct and if the relatively unworn dent.i.tion (category 3) means that these animals are young of the year, we must conclude that individuals born in early summer may molt from juvenal to postjuvenal, then to adult pelage, and finally in the autumn into another adult pelage. Other individuals, six in number and of categories 2 and 3, are simultaneously completing the juvenal to postjuvenal molt and beginning the postjuvenal to adult molt. The juvenal to postjuvenal molt begins, as has been described by various authors, along the lateral line and proceeds dorsally and ventrally and anteriorly and posteriorly, and the last patch to lose the gray juvenal color is the top of head and nape, or less frequently the rump. In some individuals a gray patch on the nape remained but emerging hair was not apparent; perhaps the molt had been halted just prior to completion. The progressing band of emerging hair is narrow in most specimens but in some up to one-fifth of the circ.u.mference of the body has hair at the same degree of emergence. Subsequent molts, both from postjuvenal to adult pelage and between adult pelages, are less regular in point, or points, of origin, width of progressing molt, and amount of surface molting at one time. Half or more of the dorsum is oftentimes involved in the same stage of molt at once. In some specimens the molt begins along the lateral line, and in others in several centers on the sides.

In some skins distinct lines of molt are visible without parting the hair, and in some others the molt is patchy in appearance. Growth of new hair is apparent at various times of the year as a result of injury such as that caused by bot fly larvae, cuts, scratches, or bites of other mice. Abrasion, wear, irritation by ectoparasites, and other kinds of injury to the skin may play a part in the development of a patchy molt.

Both breeding and molting are sources of considerable stress, and the delay of the peak of molting activity until November when breeding activity has decreased seems of benefit to the mice. A change in the ratio of young mice (categories 1, 2, and 3) to old mice (categories 4 and 5) between August and November was noted. In August, 29 per cent of the population is composed of old mice, and in November only 6 per cent.

This change results from birth of young as well as death of old mice, but may indicate that a mouse in November has less than one chance in ten of being alive the following November. Some females born early in the reproductive season breed in their first summer or autumn. For example, a female of category 2, taken on August 12, and probably in postjuvenal pelage, had placental scars. Undoubtedly the young of the year contribute to the breeding population, especially late in the season.

[Ill.u.s.tration: FIG. 3. Frequency distributions, according to size, of _Reithrodontomys megalotis_ and _Peromyscus maniculatus_ in three samples taken in August, September, and November. s.e.xes and pregnancy or nonpregnancy of females are indicated. See discussion in text.]

In Figure 3 the proportion of females bearing embryos in August, September, and November is shown. Of the females trapped in August, 11 of 32 that were more than 144 mm. in total length contained embryos; an additional 14 females were lactating or possessed placental scars or enlarged uteri. Therefore, approximately 80 per cent of the larger females were reproducing in August. In September two females were pregnant and an additional sixteen of the 44 females examined showed other evidence of reproduction; these eighteen females make up 41 per cent of those more than 144 mm. in total length. The only reproductive data available for November pertain to the presence or absence of embryos. No female was pregnant although 35 females more than 144 mm. in total length were examined. Some of the skins show prominent mammae indicative of recent nursing, and juveniles less than a month old were taken. The reproductive activity of deer mice on the Mesa Verde seems to be greatly reduced in autumn.

Peromyscus difficilis nasutus (J.A. Allen) Rock Mouse

_Specimen_: 1 mi. NNW Rock Springs, 7600 ft., 69413, a young individual completing the molt from juvenal to postjuvenal pelage.

Peromyscus truei truei (Shufeldt) Pinyon Mouse

_Specimens examined._--Total, 42: North end Mesa Verde National Park, 7000 ft., 76220-76232; Far View Ruins, 7700 ft., 69326-69327, 79222, and 8 uncatalogued specimens in preservative; Far View Point, 76532-76535; Far View House, 7700 ft., 74416 MVZ; 1/2 mi.

NNW Rock Springs, 7500 ft., 69429-69430; Rock Springs, 7400 ft., 69431-69435; Park Well, 7450 ft., 69428; Headquarters, MV 7882/507; back of Museum, MV 7879/507, 7880/507, 7881/507; Square Tower House, 6700 ft., 69438.

In August three females were pregnant or lactating, or both. None of seven adult females taken in November was pregnant.

Neotoma cinerea arizonae Merriam Bushy-tailed Wood Rat

_Specimen_: Head of Prater Canyon, MV 7873/507. Another, in the Denver Museum, from Spruce Tree House, was reported by Finley (1958:270).

_Neotoma cinerea_ prefers vertical crevices in high cliffs but occupies other areas.

Neotoma mexicana inopinata Goldman Mexican Wood Rat