NATIONAL PARK SERVICE Fauna of the National Parks of the United States No. 4 Ecology of the Coyote in the Yellowstone

Figure 46— Tracks show how coyote followed along a snow bank, bordering a road opened up by the snow plow,
hunting mice which come out in the road from under the snow.
Swan Lake Flat, April 14, 1938.

FIELD MOUSE

THE FIELD MOUSE, Microtus sp., along with pocket gopher, is the staple food item in the coyote diet from April to November. During the winter months some field mice are eaten but they represent a minor item in the diet over most of the winter range, where other foods are more available. This was especially true in the winter of 1937—38 when the snow was crusted and there was much carrion. When snow conditions are favorable and mice abundant, the coyote can subsist quite well on mice even though a foot or more of snow covers the ground. A total of 3,044 or 33.9 percent of the 8,969 food items were field mice. They occurred in 2,155 droppings.

Accurate measurements of field mouse population could not be made, but from general observations it was apparent that the populations were high in the spring of 1937 when the study was begun, and remained high during the winter of 1937—38. There seemed to be, from general observations, some slight reduction in the population by the beginning of 1938. In Jackson Hole, immediately south of the park, great numbers of field mice were found dead during the winter, and they became relatively scarce by the summer of 1938.

During the spring, summer, and fall, coyotes spend much time in the meadows hunting mice and pocket gophers. This occupation seems to be successful on the last spring snow, for coyotes often hunted there when bare areas were available. To illustrate the mousing behavior a few descriptions from my notes are given.

On May 14, 1938, at Willow Park I watched from 9:45 a. m. until 11:30 a. m. two coyotes hunting over a broad expanse of snow which was more than a foot deep in places. Much of Willow Park was bare of snow but these coyotes confined their hunting to the snow field. A single coyote was hunting mice near me when I started watching, but in a few minutes a second coyote came out on the snow from the woods opposite. By 10 o'clock the near coyote which I had been watching had moved to the far side and the other had come quite near where I was hiding. Between 10 and 11:30 o'clock the latter was seen to capture and swallow 11 animals, all of which appeared to be field mice. For about 10 minutes of the period the coyote was out of my view so it may have captured one or two others, and undoubtedly captured one or more mice during the 15 minutes it hunted, when I was watching the other coyote. In places the snow seemed to be crusted, for in pouncing, the coyotes occasionally were not able to break through the surface. Usually the mouse was not captured on the first pounce, but only after further quick strikes with the paws, three or four of which were sometimes made after the original major pounce. Sometimes the coyote would dig and paw for a minute or two before catching the mouse. The closer the coyote approaches to the point of capture, the more agitated it becomes, as is indicated by vigorous tail-wagging. Several times increased excitement on the part of the coyote was followed immediately by the capture. Once or twice a coyote was seen to cover 10 or 15 yards in four or five jumps before pouncing. Once, one of them ran about 15 yards and picked up a field mouse which was on the surface. In one place a second mouse was caught by further digging in the snow. The first pounce probably destroys the runways thus closing off ready avenues of escape and allows the coyote to pounce more accurately a second or third time. The coyote catching the 11 mice pounced without success about 30 times. These misses were in the snow, but in grassy areas misses were also frequent. The coyote was a male and seemed to be an adult. Once one of the coyotes stood at attention ready to spring for 5 minutes and then walked off without following through. Seven of the mice were caught in an area not more than 100 yards across. Both coyotes hunted throughout the period that I watched and were lost to view when they moved into the woods. At 4 o'clock a coyote was again hunting on the snow in the same locality.

Figure 47— Typical attitudes of a coyote catching a mouse.
Sketch from life by O. J. Murie.

As a coyote approaches a spot stealthily, it places each foot on the ground slowly and only gradually letting down its full weight. Sometimes it watches and listens with one forefoot poised in the air. Frequently a mouse is scented or heard while the coyote is trotting. It will then come to a stop, walk stealthily a few steps and poise for the spring. Standing with all four feet held slightly together, nose pointed at the spot, and ears cocked sharply, its body sways back a perceptible amount. Many times before actually leaping the coyote assumes a tense position only to relax and wait for the right moment. Generally the coyote springs high in the air and drops on its prey, hitting it with the front feet. The forelegs are held straight and braced to take the jar as it strikes. When the victim is caught beneath a mat of grass, the coyote must carefully paw aside the grass to get its prey.

Ranger Lee Coleman told me that in Pelican Meadows where the snow lies deep in winter he has frequently found coyotes hunting mice over areas which the buffalo have partly cleared in feeding. In these meadows where there are sometimes more than 200 buffalo wintering, this symbiotic relationship may be quite important at times to the few coyotes staying there.

On Swan Lake Flat on April 14, 1938, coyotes had been traveling the sides of the road along the snow bank made by the rotary plow. The fresh snow showed that many mice had come out of the base of the drifts onto the road. I saw a deer mouse which for several yards was unable to find a retreat in the snow. The coyotes had quickly learned of this mousing opportunity and had been there hunting.

Figure 48— Sagebrush killed by mouse birdling: part of a patch of 600 square yards
in which it was estimated one-fourth of the sage had been killed.
Across Lamar River from Buffalo Ranch, June 7, 1938.

POCKET GOPHER

Pocket gophers (Thomomys fuscus fuscus) made up 1,939 or 21.6 percent of the food items. Remains were found in 1,407 droppings. These rodents are active in winter under the snow and are occasionally taken at this time. Predators know the habits of the pocket gopher and have learned to wait for them to reappear above the ground when they are active at the open holes. Once a great grey owl at Yellowstone Lake was observed watching a spot for several minutes and then pouncing on a gopher when it appeared. In summer the coyotes readily capture them by waiting at the tunnel entrances where they are digging or coming out into the open to forage. From the results of the droppings analyses it appears that Thomomys are about as readily captured as field mice.

The coyote is probably one of the chief checks, due to predation, on the pocket gopher population. It is difficult to say how effective this control may be, or how beneficial in a wild region. In a mountain area, any harmful effects of moderately numerous pocket gophers may possibly be balanced by beneficial effects. Importance of the pocket gopher as a factor in erosion is not known. From general observations their activities in this respect seem to be beneficial as well as rather harmful. In some areas where sagebrush is an important deer and antelope winter food, and at the same time in a precarious condition from overbrowsing, coyote predation on pocket gophers may be highly valuable, for these rodents were found to cause considerable local though probably temporary damage to sagebrush during the winter months in places by cutting off branches and twigs. Some bushes were pruned down to within a few inches of the ground. Over most of the park, however, pruning of sagebrush by pocket gophers is probably not very harmful, and the animals are now absent or scarce over the heart of the critical antelope and deer winter ranges in the Gardiner region. Whether coyotes have had much to do with this local scarcity of gophers is not known but possibly the pocket gopher does not care especially for this Upper Sonoran habitat, particularly in its present overbrowsed condition.

Figure 49— Sagebrush trimmed by pocket gophers in winter. Mice and pocket gophers
are the leading items in the coyote diet for a large part of the year.
Blacktail Deer Creek, May 15, 1938.

In the coyote-pocket gopher relationship we find an apparent blending of harmful, beneficial, and neutral influences not readily segregated or measured. The rodent consumes a certain amount of forage, but also does service in soil building, furnishes an important food supply for raptores and carnivores, and acts as a buffer species. On the other hand, the coyote makes inroads on this natural food supply but does not exhaust it, and among birds and mammals certainly is the greatest single factor in keeping the pocket gopher population within safe bounds.

The relative frequency with which pocket gopher and field mouse occurs in the droppings depends upon the locality and possibly on the time of year. In localities where pocket gopher habitat is scarce the percentage of gopher remains is low, but where pocket gopher and field mouse habitats are both present the pocket gopher may occur as frequently or more frequently in the droppings than the field mouse. In droppings gathered at Swan Lake where mice are specially plentiful in the marsh and heavy sedge around the lake and pocket gophers occur on the surrounding slopes, remains of 122 pocket gophers and 256 field mice were identified.

Between Swan Lake Flats and Norris near the road where there is but little pocket gopher habitat, the coyotes hunt mainly in the marsh along the creeks. Droppings gathered here contained remains of 82 pocket gophers and 352 field mice. Where good field mouse habitat predominates and pocket gophers are relatively scarce, gopher-mouse occurrence in the droppings is as follows: Gibbon Meadows and Elk Park, 39—362; Madison Junction, 3—53; Old Faithful, 17—222.

In other localities where pocket gopher habitat occurs extensively along with field mouse habitat, occurrence of the two animals in the droppings does not vary widely. Data from such localities follows:

During July, August, September, and October the relative proportion of pocket gophers to field mice in the droppings is higher than earlier or later in the season. This high pocket gopher incidence probably coincides with a period of greater surface activity. The occurrence of gophers and field mice in the droppings from April to November is as follows:

As mentioned before, only the majority, of the droppings gathered in a month were actually deposited that month, but the dates as a whole are probably accurate enough to show the general trend. Availability of the prey species largely determines the extent to which it is eaten.

SNOWSHOE HARE

Remains of snowshoe hare (Lepus bairdi bairdi) were found in 305 droppings, making up 3.44 percent of the items. Many of the hares had been eaten in winter while in the white pelage. Where hares are abundant, coyotes are able to hunt them systematically and subsist on them alone. Coyotes spending the winter in the interior of the park probably feed extensively on them.

The snowshoe hare is widely distributed but not abundant over most of the park. In a few places it seemed to be locally abundant; this was true in an area near Old Faithful. The sweeping cycles of abundance and scarcity do not seem to occur with any regularity in Yellowstone although there is, of course, some variation in numbers from year to year. In an old diary kept by the soldiers at Sylvan Pass in 1903, 15 or 20 hares were frequently reported seen in a day so that hares at that time must have been quite plentiful. Ranger Elliott (1931) reported hares more abundant at Yellowstone Lake than in several preceding years.

GOLDEN-MANTLED MARMOT

Remains of marmot (Marmota flaviventris nosophora) occurred in 120 droppings. The marmot is a natural coyote food and in some localities makes up an important part of the coyote diet. On the high ranges occupied by bighorn in Teton National Forest, where marmots are plentiful, they form the main item in the coyote diet. In Yellowstone, although marmots are plentiful in rocky areas, over large areas separating the typical marmot habitats they are scarce.

MUSKRAT

Remains of muskrat (Ondatra zibethica osoyoosensis) were found in 98 droppings. Muskrats are not very numerous but are generally distributed along the water courses and ponds. During the fall and spring they are especially vulnerable to coyote attack when they wander out over the snow. Should a muskrat be discovered by a coyote when journeying on land, its chance of escape, are, of course, slight. Journeys of more than 100 yards on the ice were noted. Once a fresh coyote track was seen which crossed the track of a wandering muskrat but, fortunately for the latter, they had not met. Coyotes have been found to investigate a network of tunnels along a stream hut it seems probable that muskrats are generally captured accidentally, for it would hardly be profitable for the coyote to spend a great deal of time hunting them. Coyote pressure on muskrats does not appear to be great.

GROUND SQUIRREL

Remains of ground squirrel (Citellus armatus) were found in only 46 droppings. The low incidence of ground squirrel in the diet is probably due to scarcity of these rodents over most of the park. It is possible that the ground squirrel population has been held in check by the coyotes. Ground squirrels make their appearance in the spring before the snow disappears, burrowing through the snow to reach the surface. Some ground squirrels appearing on the snow had traveled as far as 40 yards to feed on vegetation; at such times the ground squirrels would be highly vulnerable to predation. Tracks in the snow were noted several times showing that coyotes had chased ground squirrels.

On May 14, 1937, I saw a ground squirrel climb to the top of a sagebrush and peer after a coyote which had passed close to its hole, and this watchfulness was observed on two other occasions. Where the ground squirrel is plentiful on the coyote range it probably is an important food item.

Figure 50— Ground squirrel coming forth in early spring. This animal crossed 40 yards of snow to feed at the bare area.
At such times ground squirrels are especially exposed to coyote attack.
Jackson Hole, Wyoming, April 1937.

WHITE-TAILED JACKRABBIT

Remains of jackrabbit (Lepus townsendi campanius) were found in 37 droppings. This animal occurs only on the north side of the park and is not abundant, although tracks can always be found on its range. Since it does not occur in the interior of the park, there was no chance for it to be represented in about 3,500 of the droppings collected, so the incidence is lower than it would be if all the droppings were collected in jackrabbit habitat. George Bird Grinnell (Ludlow 1876), reporting on a trip made to the park in 1875, states: "Where all the coyotes and grey wolves have been killed or driven off, the hares exist in great numbers; but where the former are abundant the latter are seldom seen."

Because jackrabbits feed extensively on many plants, such as fringed sage brush and various kinds of yellowbrush, which are important big game plants, the coyote's predation on jackrabbits tends to be beneficial from the standpoint of preservation of the over-utilized range.

Figure 51— The jackrabbit is often an important food item in localities where it is abundant,
but in Yellowstone it is of minor importance.
Mammoth 1935.

PORCUPINE

Remains of porcupine (Erethizon e. epixanthum) were found in 35 droppings. Full-sized quills which became sharp on drying were frequently found in the droppings. Wherever porcupine occur they tend to appear regularly in the coyote diet. O. J. Murie (1935) found porcupine remains in 78 of 714 droppings gathered in Teton National Forest just south of Yellowstone Park. The observations of several men are given, showing that the coyote can kill a porcupine and probably does so frequently. Porcupine quills were found under the hides of many of the coyotes shot at the time that coyote control was practiced in the park. Near the South Entrance on March 19, 1937, Ranger Verde Watson (Yellowstone Nature Notes, March 1938, p. 15—16) found a porcupine that had been killed by two coyotes. They had worried it while it traveled to the river, where it had escaped, only to be again attacked and killed when it came to shore about 200 yards downstream. Now that other predators are rare in Yellowstone National Park the coyote probably serves as a useful check on the species.

DEER MOUSE

Remains of deer mouse (Peromyscus maniculatus osgoodi) were found in only 34 droppings. Considering the abundance of this mouse in the park, one would expect to find more of their remains in the droppings. O. J. Murie in his Jackson Hole coyote studies also found its incidence in the diet to be low, only eight deer mice occurring in 714 droppings and 64 stomachs. In some fox studies on the George Reserve in Michigan (Adolph Murie, 1934) I found that the deer mouse likewise made up but a small item in the fox diet. The deer mouse, although active above ground at night, is not so easily secured as the field mouse. It scurries from cover to cover, while the field mouse often feeds and travels in runways where it is easily captured by the coyote. It furnishes food for other animals equipped to feed readily upon it. Santee and Granfield (1939, p. 3—9) for instance, in California found the saw whet owl, Cryptoglaux acadica feeding almost entirely on the deer mouse, Peromyscus sp.

PINE SQUIRREL

Pine squirrel (Sciurus hudsonicus ventorum) was found in 25 droppings. These squirrels are abundant in the pine woods, but as one would expect, it appears that they are captured only incidentally by the coyote.

BEAVER

Remains of beaver (Castor canadensis missouriensis) were found in 17 droppings. On June 29, 1937, along the upper reaches of the Gardiner River, the carcass of a beaver was found which had been eaten mainly by birds but had also been visited by a coyote. The beavers are generally distributed over the park. Where colonies have eaten out their food supply there are many abandoned dams, and in some places meadows have replaced the beaver ponds. In some streams the older deserted dams and stumps and stubble of the beaver harvest are found near the mouth, and as one approaches the head of the stream the signs become more and more recent until one finally arrives at the occupied ponds. It appears that food supply is the real control of the beaver population. In some ponds whose shores have been denuded of most of their beaver food for several years, a few beaver continue to subsist. The fact that they are able to persist under unfavorable conditions existing at these denuded ponds also suggests that coyote pressure on them is not unduly severe. The beaver pond-beaver meadow cycles do not seem to have been interrupted in Yellowstone by a long continued abundance of coyotes.

COTTONTAIL RABBIT

Remains of cottontail rabbit (Sylvilagus nuttalli grangeri) were found in only eight droppings. The cottontail is absent from much of the park, and from the areas in which most of the coyote droppings were gathered. Only about 300 droppings were gathered in the district where cottontails are found. They are quite common along the Gardiner River below the mouth of Lava Creek, around Mammoth, and on the Gardiner and Game Ranch areas. Along the Yellowstone River a few tracks were noted as far up stream as Hellroaring Creek but this seems to be the limit of this Upper Sonoran form. It is interesting to find several Upper Sonoran species that have extended into the park along the Yellowstone River. Ranger Grimm said that many rabbits died around the elk trap at the Game Ranch during the winter of 1935—36, and that since that time they have not been nearly as abundant. For a stretch below the mouth of Lava Creek along the Gardiner River, rabbits were quite common. This area was also much frequented by coyotes attracted there by carrion as well as by the Mammoth dump. The rabbits were living in holes and in heavy tangles of brush. Near the Government Garden on February 6, 1938, a coyote was seen with a recently caught rabbit in his jaws. Between Mammoth and Gardiner two rabbits were found which had been killed by cars and were probably later found by coyotes, ravens, or magpies. The rabbit is present in fair numbers within its range and there is apparently no excessive predation on it by coyotes.

Figure 52— A cottontail rabbit was found in this spot on two successive days.
Several coyotes were feeding on an elk carcass at this time about 10 yards away.
Gardiner River, April 6, 1938.

MISCELLANEOUS SMALL MAMMALS

The jumping mouse (Zapus princeps) was found in only seven droppings. Although several species of chipmunks (Eutamias sp.) are represented in the park, and the animals are fairly common, their remains were found in only six droppings. Coyotes probably pick up chipmunks only as they chance to come upon them. These rodents are, no doubt, too alert and active to be profitably hunted.

As woodrats (Neotoma cinerea orolestes) usually live among the rocks they are not readily available to the coyote. Their remains were found in four droppings.

Remains of mink (Mustela vison energumenos) were found in three samples. It is my impression that, although widely distributed, mink are not abundant in the park.

Shrew remains (Sorex sp.) were found in two droppings. On September 4, 1937, at Gibbon Meadows a shrew carcass was picked up about 50 yards from where I saw a coyote pup. The shrew was fresh and still moist around the neck so it appeared that it had been caught and left uneaten by the coyote a few moments before I found it. O. J. Murie (1935) found no shrews in 714 droppings gathered in the mountains in the Jackson Hole region. This fact, together with the low incidence in the Yellowstone material, suggests that they are not relished. In many places where the coyotes hunt, shrews are active and available. In my study of foxes at the Edwin S. George Reserve near Pinckney, Mich., it was found that many shrews were captured and left uneaten. It is likely that this is also true of coyotes when food is abundant.

Remains of coney (Ochotona princeps ventorum) were found in but one dropping. This species is well protected from coyote attack by the rocks in which it lives.

Weasel remains (Mustela frenata) were found in only a single dropping. At Crevice Creek on February 17, 1938, signs indicated that the coyotes had been playing with a weasel carcass. The rear half was lying on the snow about 500 yards from the anterior portion. Apparently neither had been eaten.

It is likely that the remains of bat (Myotis sp.) found in one dropping represented carrion.

The one dropping noted containing house cat was found near Gardiner where cats occasionally stray into the park.

Flying squirrel (Glaucomys sabrinus bangsi) remains were found in one dropping. The species is not abundant, nor is it readily available to coyotes.

Remains of a very young marten (Martes caurina origenes) were found in one dropping. This may or may not have been carrion. Martens are too active to fall frequent prey to coyotes, although they occur in moderate numbers in the park.