Number 86      June 24, 1920

OCCASIONAL PAPERS OF THE MUSEUM OF
ZOOLOGY

UNIVERSITY OF MICHIGAN

Ann Arbor, Michigan    Published by the University

THE MAMMALS OF WARREN WOODS,
BERRIEN COUNTY, MICHIGAN

By Lee Raymond Dice

Few detailed studies of the mammal associations of the forests
of the United States have been made. But if we are ever to
know, for our different species of mammals, the natural environments
under which their evolution and differentiation occurred,
we must study and describe their habitats and habitat limitations
before all the native areas in the country have been altered
by the activities of mankind. As a contribution to this subject
the following paper is presented.

The Warren Woods are a state preserve under the Edward K.
Warren Foundation. They are located in Berrien County, Michigan,
about three miles north of Three Oaks. The preserve consists
of about two hundred acres, of which somewhat less than
half is in clearing and the remainder mostly covered by forest,
much of it still nearly in its primitive condition.

Fig. 1.—Sketch map of Warren Woods Preserve. The distribution of
the various mammal habitats is indicated.

The topography is nearly level, though the area is cut by a
number of ravines draining to the Galien River, which flows
[Pg 2]
[Pg 3]
through the preserve. Along the river and in its bends there
are moderate-sized flood-plains.

On the flood-plains a few small buttonbush swamps occur;
and along the margins of the river a few freshly formed mud
bars have not yet become forested; but most of the flood-plains
are covered by heavy forest. The higher ground, except that
in the clearing, is covered by heavy beech-maple forest.

Several types of habitats are represented in the clearing: in
a few of the cleared ravines a thick growth of sedges and iris
occurs; on the higher ground small areas are dominated by rushes,
other areas by sedges, while the greater part is covered by grass.
In parts of the clearing blackberries and other shrubs have grown
up to form thickets, and in many places, especially along the edges
of ravines, second-growth trees of oak, maple, or beech grow
in the thickets or form small groves.

The mammal habitats found on the preserve may be listed
as follows:

It is unfortunate that all of the area in clearing and about
half of the forested area on the preserve has been and is being
heavily pastured by cattle and horses. The presence of stock
has changed the native conditions so much that, so far as interpreting
the primitive mammal associations is concerned, little
dependence can be placed on studies made in that portion of the
preserve. The grass and herbage is extensively eaten off, and
[Pg 4]
many of the shrubs and young trees eaten or badly mutilated.
Under the pastured forest little underbrush or herbage remains,
and the conditions are very poor for small mammals.

In all the forest, in the unpastured as well as in the pastured
part, a number of trees have been cut out in former years, and
although no trees are now being cut down, all the trees and branches
which fall are being cut up for firewood. This results in there
being few logs and little dead brush on the ground, and removes
a favorite place for small mammal nests and runways, as well as
largely eliminating as mammal food the insects and larvae which
are dependent on decaying wood. However, with the exception
of the removal of the logs and of a few trees, that part of the
forest to the north of the river is still in practically its native
condition, and it shows no evidence of ever having been pastured.
It is thus a splendid place for the study of the native faunal
conditions.

Although the whole area in the preserve is small there are
other areas of adjoining forest along the Galien River, so that,
for the smaller mammals at least, the results of the study and
trapping should indicate the primitive habitat distribution. The
relative abundance of the different species, however, is probably
much changed by the decrease throughout the whole region of the
carnivores, which have been much hunted and trapped by man.

The mammals of Warren Woods were intensively studied by
the author from July 3 to September 3 in the summer of 1919. A
camp was maintained near the edge of the Woods, and by trapping
and hunting every effort was made to determine the mammal
fauna of the preserve. From one hundred to two hundred traps
for small mammals and about twenty traps for the larger species
were kept constantly in operation. A small amount of trapping
was done in the fields and along the roadsides immediately adjoining
the Woods and a few records were secured from the camp
house.

[Pg 5]
The work was supported by the Michigan Geological and
Biological Survey. Dr. Alexander G. Ruthven directed the work
and gave much assistance in securing the needed collecting equipment.
Much help was given by George R. Fox, Director of the
Warren Foundation, not only in getting to and from the camp
at the Woods and in numerous other courtesies, but also in information
about the Woods and about the mammals there. The
plant identifications were made by Mr. C. Billington.

The figures following the specific names in the lists of mammals
from each habitat indicate the number of individuals trapped,
shot, or seen and positively identified in that habitat.

Natural Habitats

Aquatic habitat:

Mink and muskrat are reported by residents to occur in the
Galien River in and near Warren Woods, but I was unable to
secure any specimens though traps were set for them; neither
did I see any signs of their presence.

Buttonbush-swamp habitat:

A large buttonbush swamp occurs in the preserve just south
of the Galien River, but around this the native trees have been
cut away and over its accessible edges it has been heavily pastured,
so that it is not at all in its natural conditions, and it was
not trapped. In the unpastured flood-plain north of the river
there is another buttonbush swamp of several hundred yards
extent. This latter area was the one selected for study.

In this typical swamp the buttonbush, Cephalanthus occidentalis,
is practically the only plant present. It thickly covers
the area with its tangled branches, which grow to heights of four
[Pg 6]
to ten feet. The water had drained away in August, leaving the
ground bare, though still wet and soft. Under the cover of the
buttonbush there is no herbage whatever, and upon the ground
there are only a few decaying logs and a few small sticks, which
often carry a light growth of moss.

Around the edges of this swamp there is a narrow belt of
thick herbage, closely encroached upon by the typical forest of
the flood-plain.

Fifty mouse traps set in this habitat took eight northern
white-footed mice and two house mice the first night, August 5.

Shore habitat:

Along the shores of the Galien River a narrow strip of bare
ground was exposed in July and August. The ground of this strip
is mostly mud, but in a few places it is sand or gravel. Usually
the habitat is narrow, but in some places it is five to ten feet wide.

Tracks of raccoon were frequent on the shore along the river.
From a few mouse traps set on the bare mud shore one northern
white-footed mouse was taken August 4 beside a drift log.

Mud-bar herbage habitat:

A few small recently formed mud bars occur along the Galien
River in its outer bends. These bars have not yet had time to
become forested, and on their outer edges nearest the river there
is usually no vegetation, this part being included in the shore
habitat. On their older portions next the forest of the flood-plains
occurs a thick growth of herbs, several annual grasses, and
rarely a willow, Salix sp., or a seedling tree of white elm, Ulmus
americana, cottonwood, Populus deltoides, maple, Acer rubrum or
saccharinum, and others of the typical flood-plain species. The
vegetation during August is very thick, and reaches a height of
[Pg 7]
four to six feet. The soil is either mud or in a few places fine
sand, and the ground is quite moist.

In this habitat four northern white-footed mice were trapped
August 3–4.

Flood-plain forest habitat:

There are considerable areas of flood-plain along the Galien
River, and except for the recently formed mud bars the flood-plains
are heavily covered with a mixed forest in which the linden,
Tilia americana, white elm, Ulmus americana, and sycamore,
Platanus occidentalis, are conspicuous species. Under this forest
there are a few small trees, but there is very little underbrush.
The herbage also is sparse and, though in a few places there is a
considerable growth of ferns, grasses, and sedges, and of other
herbs, there are also many bare areas.

One of the significant features of the flood-plains, so far as
the mammals are concerned, is the flooding to which these areas
are subjected during the spring high-water. At that time the
flood-plain for a number of days or weeks may be covered with
several feet of water.

During the period between July 29 and August 28 a total of
one hundred and seventy mouse traps set in the flood-plain forests
along the Galien River took for the first nights’ trapping, twenty-two
northern white-footed mice and one house mouse. Short-tailed
shrews, more white-footed mice, pine voles, and a jumping mouse
[Pg 8]
were secured on nights after the first. Larger traps took during
the whole period one raccoon, one New York weasel, and two
woodchucks. Several other woodchucks and a number of red
squirrels were seen. Ridges of moles were numerous, but no
specimens were secured. Tooth marks on an old, partly fallen
linden were probably made, perhaps a number of years ago, by a
porcupine.

Throughout the woods there are a number of small ravines.
These ravines are forested with trees mostly of the flood-plain
type, and there is evidence in many of the ravines, at least in
their lower parts, that flooding occurs in the ravine bottom during
the spring.

Fifty mouse traps set August 26 in a large ravine north of the
county road took on the first day five northern white-footed
mice and one house mouse. Short-tailed shrews, more white-footed
mice, and pine voles were trapped on following days. Here
also red squirrels and a cottontail were seen, and tracks of raccoon
noted. Evidently the fauna is the same as that of the flood-plain,
with which it is here included.

Beech-maple forest habitat:

The climax forest of the region is dominated by the beech,
Fagus grandiflora, and the sugar maple, Acer saccharum. The
trees in this forest are very large, so that the forest crown is
high and the shade dense. Only a few small trees occur and
these are mostly young beeches and young sugar maples. The
[Pg 9]
underbrush varies much in height and denseness; mostly it is
quite open, so that one can easily walk through the forest, but
in a few places the growth is more dense. Common members
of the underbrush on the higher ground are the small beeches,
sugar maples, and the spice-bush, Benzoin aestivale; while on the
lower benches along the river seedlings of the paw-paw dominate
the undergrowth. The ground is heavily covered by decaying
leaves and a little dead brush and fallen branches, but nearly
all the logs have been removed. Early in the spring a thick
growth of herbs covers the ground, but by July the herbs are mostly
gone, only a few remaining, and there are many small bare areas
covered only by leaves. The soil under this forest seems to be
mainly clay; in spring or after heavy rains pools of water are
formed, and these remain for a long time.

Between July 21 and August 27 a total of two hundred and
eighty-five mouse traps set in the upland forest took on the first
nights one short-tailed shrew and thirty northern white-footed
mice. In addition to these species pine voles and a jumping
mouse were trapped on days after the first. One shrew was
caught alive August 30, as he was running about on the forest
floor at 7:30 A.M. A few tracks of raccoon were seen from time
to time on the road leading through the woods. A few fresh
burrows of woodchucks were noted at the edges of benches and
of ravines. A few red squirrels were seen at different times and
two collected. Fox squirrels were rare, being noted only a few
times; Mr. Norman A. Wood also saw these squirrels on two
occasions in May. One cottontail was shot, in the climax forest.
Mr. Wood collected a chipmunk in the climax forest on May 15,
1918, and saw another in the same habitat in May, 1919.

Aerial habitat:

Bats were seen on a few evenings, flying about over the climax
forest, and over the adjacent region, but they were extremely
rare, and efforts to shoot a specimen failed.[Pg 10]

Modified and Artificial Habitats

Second-growth forest and scrub habitat:

Small trees and brush have grown up along the edges of many
of the ravines in the cleared fields in and surrounding the preserve.
Many of the trees are oaks, but beech and hard maple also occur,
a few of them being relics from the original forest. Considerable
brush is present, formed by a large variety of species. A few
other small patches, especially in ravine bottoms and on flood-plains
have been allowed to grow up to brush and small trees. In
nearly every case these areas are heavily pastured.

The conditions here included in the second-growth forest and
scrub habitat are not homogeneous, but differ in each different
location where the habitat is found, tree and shrub species abundant
in one situation not being present in another. The habitat
is usually narrow in extent, being often confined to the width of
the steep ravine wall.

Owing to its poor development and uncertain characteristics
no intensive trapping was done in this type of habitat. A weasel
was trapped in open beech-maple-oak forest at the edge of a
cleared ravine, and a woodchuck was shot in the same type of
habitat. Another woodchuck and a skunk were trapped at
different times in low willow brush on the banks of the river just
north of the woods. Northern white-footed mice, Pennsylvania
voles, and a house mouse were trapped in thick oak brush and
trees alongside a road north of the woods. A red squirrel was
shot in second-growth oak and aspen woods in the north part
of the preserve, and they were seen in open woods along ravines.
[Pg 11]
Cottontails were noted a few times in blackberry thickets, in
brush in ravines, in clearings along the river, and in beech-maple-oak
forest along ravines.

Cleared-ravine sedge habitat:

A large ravine south of the river had been cleared of trees
evidently several years previously, and it has now grown up
mostly to sedges, grass, and iris in its more moist parts. Along
the little brook which flows through the ravine there is a fringe
of willows, and among the sedges a number of shrubs and small
trees occur, mostly thorns and a few young sycamores and black
walnuts.

Fifty mouse traps set in this habitat took on the first day,
August 16, three northern white-footed mice and three Pennsylvania
voles. More voles and a short-tailed shrew were taken
on later days.

Cleared-upland rush habitat:

In the shallow, poorly drained depressions of the cleared
upland the vegetation is dominated by rushes, which grow in
clumps and form a thick growth, reaching a height of about
one meter as a maximum. At the edges of the habitat and in
places not thickly covered by the rushes a growth of sedges, grasses,
and moss covers the ground; but under the thickest growth of
rushes the ground is bare and is evidently covered by water
during a part of the year. On this upland one small pond surrounded
by rushes did not dry up until late in August. In a
[Pg 12]
few places small blackberry thickets occur in the areas of rushes
and dominate all other plants.

Fifty traps set in this habitat, on August 8, took on the first
night one prairie white-footed mouse and one Pennsylvania vole;
the prairie white-footed mouse was taken just at the edge of
the growth of rushes. Other Pennsylvania voles as well as a
lemming-vole, a prairie vole, and several young ground squirrels
were secured on later nights by the same trap-line. A skunk
den was situated in a blackberry thicket in the midst of the largest
patch of rushes.

Cleared-upland sedge habitat:

Sedges are dominant over a part of the cleared upland, occurring
on the moist gentle slopes which are too well drained for rushes,
but not in any numbers on the drier and higher parts of the upland.
With the sedges there are a few grasses, and the ground is sometimes
covered by a moss, but the sedges are by far the most abundant
plant.

Fifty mouse traps set in this habitat took on the first night,
August 15, one prairie white-footed mouse and one prairie vole.
Other prairie voles and white-footed mice were taken on later
nights, as well as one Pennsylvania vole and one house mouse.
Burrows of the thirteen-striped ground squirrel were numerous
in the sedges.

Cleared-upland blue-grass habitat:

The most widespread habitat of the cleared upland on the
south part of the preserve is the blue-grass habitat. In this
[Pg 13]
habitat the blue-grass, Poa, is the dominant plant, growing to a
height of usually not over 0.5 meter. With the blue-grass are
associated a few thistles, yarrow, and several other herbs. During
the whole period of my stay in the region, July and August, the
habitat was very dry, and the grass and herbs had mostly dried
up. This habitat and all the other habitats of the clearing were
being heavily pastured by stock.

Trap-lines totaling one hundred mouse traps, set on August 6
and August 18, took on the first nights two prairie white-footed
mice, one house mouse, and one young thirteen-striped ground
squirrel. More white-footed mice were taken on later nights.
Many ground squirrels were taken in larger traps at the
mouths of their burrows; most of these were young of the year,
and all of them were very fat in preparation for their approaching
hibernation.

Cultivated field habitat:

Cultivated fields occur throughout the region about the woods,
but the only ones in which trapping was done were wheat stubble-fields.
After the wheat was cut off these stubble-fields grew up
rather thickly to weeds. On the ground there were a number
of fallen heads of wheat and some shelled-out grain, furnishing
abundant food for mice.

Seventy mouse traps set in a wheat stubble-field just north
of Warren Woods, caught on the first nights, August 13 and August
29, sixteen prairie white-footed mice. Other white-footed mice
and two house mice were taken on later nights. Several cottontails
were seen in this field, and a few mole ridges were noted.
A prairie mole was taken by Clifford Reid in a grassy patch at
[Pg 14]
the edge of a garden. In another wheat field the Helming boys
picked up a prairie vole.

Orchard habitat:

No especial study of the mammal life of the orchard was made,
and no intensive trapping was done in the habitat. However,
the orchard cannot be included in any of the other habitats recognized
in the region.

A northern white-footed mouse was caught alive by the
Helming boys in an orchard on July 16. A vole was seen to cross
a road in a small orchard on July 15; a trap was set and two
Pennsylvania voles secured, one a young of the year.

Edificarian habitat:
[1]

From time to time a few traps were set in the old farmhouse
and in the barn and other outbuildings of our camp. In these
buildings northern white-footed mice, Norway rats, and house
mice were taken. Several cottontails were seen in and under the
old barn. The Helming boys trapped a short-tailed shrew and
also house mice and a white-footed mouse in the basement of
their house.

Records of the Number of Embryos

All the females taken were examined for embryos, and the
results are here tabulated by species and dates. The term subadult
is used to indicate an individual which has reached adult
size, but which is still immature as shown by the pelage, unworn
[Pg 15]
condition of the teeth, and weakness of the skull sutures. For
each individual the age is first stated, next the number of embryos
if any, and last the length of the embryos in millimeters measured
as they lie rolled in the fetal membranes. For embryos too
small to measure with field equipment the term small is applied.

These records indicate the breeding period of the different
species and the number of young.

[Pg 17]

New State Record of the Prairie Vole

The specimens of prairie vole, Microtus ochrogaster, listed
above from the sedges and rushes of cleared upland in the Warren
Woods preserve and from a nearby wheat field constitute the
first authentic record of the species from Michigan. The prairie
vole had previously been erroneously reported to be abundant in
Washtenaw County by Covert
[2], who evidently mistook the Pennsylvania
vole for this species.

Other Mammals of this Region

The following notes on mammals formerly or at present living
in the region but not found by me in Warren Woods are based
mostly on information furnished by George R. Fox, Director of
the Edward K. Warren Foundation, and by William Schmidt, a
trapper and hunter.

Didelphis virginiana virginiana. Virginia opossum. Three
were seen and one captured in Three Oaks in the winter of 1919–20,
and another was found in a granary on the outskirts of the town.
The previous winter one was shot just southwest of town, and
another was seen in a ravine north of Harbert. One was killed
in Warren Woods some time ago.

Condylura cristata. Star-nosed mole. Mr. Hans captured one
four miles south of Three Oaks, near the Indiana line in 1919,
and brought the skin to Mr. Fox. One was taken by Dr. Parker
at Lakeside in June, 1919.

Ursus americanus americanus. Eastern black bear. Formerly
occurred in the region.

Canis lycaon. Timber wolf. About eleven years ago three
were killed from a pack of nine wolves at Lakeside. About 1910
a pack of eight were seen at various times during the winter on
[Pg 18]
the marsh between Three Oaks and Galien. The following winter
the same pack, or another of the same number, was seen southwest
of Three Oaks, and four of the pack were killed just over
the Indiana line.

Vulpes fulva. Eastern red fox. Common in the region. One
was killed in December, 1919, after having been chased from the
Warren Woods by dogs. Another was seen in January, 1920,
half a mile north of Three Oaks. In preceding winters a number
have been killed south of Sawyer where they evidently come from
the sand dunes.

Lutra canadensis canadensis. Canada otter. Formerly
occurred along the streams, but are now apparently all gone.

Felis couguar. Cougar. One was killed a few miles northeast
of Three Oaks many years ago. Another was killed near New
Buffalo in the early days.

Lynx ruffus ruffus. Bobcat. One was killed a few years
ago near the Galien River a few miles northeast of Three Oaks.

Erethizon dorsatum dorsatum. Canada porcupine. Found here
in the early days.

Sciurus carolinensis leucotis. Northeastern gray squirrel.
Black squirrels were found in numbers in the early days.

Glaucomys volans volans. Eastern flying squirrel. Said to
be common. A large number were seen by Mr. Schmidt on
one old stub in Warren Woods some years ago. Several were
seen in 1918 by Mr. Fox in second-growth oak woods in a little
ravine about one and one-fourth mile west of Warren Woods. One
was caught in Three Oaks in the fall of 1919.

Cervus canadensis canadensis. American elk. One horn from
a large swamp in Berrien County is in the Chamberlain Memorial
Museum of Three Oaks.

Odocoileus virginianus borealis. Northern white-tailed deer.
Formerly abundant in the region. Several molars were picked
up in Warren Woods.

[Pg 19]

Fossil Mammals of Berrien County

The records of fossil mammals from Berrien County here
published have been collected by George R. Fox, who has also
kindly loaned from the collections under his care several specimens
for identification. Dr. E. C. Case assisted with the identification
of the mammoth teeth.

Mammut americanum. Mastodon. (1) About 1897 the teeth
of a mastodon were dug up by a dredge within the village limits
of Eau Claire. Their disposition is unknown.

(2) At Snow, in section 36, Lake Township, a tooth was found.
This came into the possession of Frank Striker of Buchanan.

(3) A portion of a tusk and part of a skull were found on the
Beebe Farm near Baroda. These were sent to Washington, D.C.

(4) In the excavation of a ditch through a small marsh lying
near Bakerstown in Section 3, of Bertrand Township, the dredge
uncovered in the distance of between two and three miles bones,
teeth, and other evidences of six mastodons. Of these remains
the most important was a nearly complete skull with teeth in
place and disintegrated parts of the tusks. The skull was secured
by Dr. E. H. Crane, who restored parts, the tusks he did not
attempt to restore. This skull is now on exhibition at the Ward
Museum, Rochester, New York.

The skull lay about seven feet deep. It was under a bed of
matted oak brush on top of which lay a huge stone slab, estimated
to weigh two or three tons. Above was a layer of silt, then gravelly
clay; above were more silt beds, three in number. W. Hillis
Smith, who helped Dr. Crane secure the skull, furnished the above
information.

(5) On the Avery marsh, two miles east of Three Oaks, a
badly decayed mastodon skull and several teeth were excavated
in 1884. Dr. Bonine, Sr., of Niles, assisted at the excavation.
The bones and teeth were in the muck about twenty inches below
the surface. Other teeth, making seven in all, were found near
[Pg 20]
the same place. Some of the teeth from this locality are on exhibition
at the Chamberlain Memorial Museum of Three Oaks.

In addition to the above there are several indefinite records
of mastodon remains from the county.

Elephas columbi. Columbian mammoth. (1) One-half of
the tooth of a mammoth was found in the spring of 1917 on the
Beeson and Holden farm in Section 6, NE. ¼, Township of Galien,
by D. H. Beeson while cultivating corn. Two weeks later the
other half of the tooth was found. The specimen is now in the
Chamberlain Memorial Museum of Three Oaks.

The specimen is a well-worn lower third molar having twenty
ridge-plates, but some of the ridge-plates have been worn out and
lost. Seven and a half ridge-plates are included in a 100 mm. line.
The greatest length of the tooth is 280 mm. and its greatest
breadth 95 mm.

(2) A complete set of mammoth teeth with some portions of
the bones was found about the year 1900 on a farm two miles
southeast of Three Oaks owned by E. K. Warren. They were
discovered while digging post holes. The specimens are on exhibition
in the Chamberlain Memorial Museum.

One of the teeth sent us for identification proves to be a partly
worn lower third molar having twenty-four ridge-plates. There
are seven and a half ridge-plates in 100 mm. The greatest length
of the tooth is 350 mm. and its greatest breadth 95 mm.

(3) There is another record of a mammoth tooth which was
found at an unknown location in Berrien County. It was at one
time owned by Mr. Smith, who gave it to Dr. Crane. Its present
whereabouts is unknown.

[Pg 21]
[Pg 22]
[Pg 23]

PLATE I

A mud bar beside the Galien River in Warren Woods. A growth of mud-bar
herbs adjoins the flood-plain forest on the left. August 29, 1919.

Flood-plain forest in Warren Woods. There are few shrubs, but a
considerable amount of herbage is present. August 29, 1919.

[Pg 24]
[Pg 25]

PLATE II

Buttonbush swamp in Warren Woods. The swamp is surrounded
by flood-plain forest. August 29, 1919.

Climax beech-maple forest on the higher ground in Warren Woods.
August 29, 1919.