Mice and voles prefer spruce seeds : study in Maine suggests an explanation for the predominance of balsam fir seedlings in regeneration

Historic, Archive Document

Do not assume content reflects current
scientific knowledge, policies, or practices

U. S. D?CT. OF AGRICULTURE

LIBRARY

OCT 1 6 1961

by

Herschel G. Abbott
Arthur C. Hart

/yZ£±2-3.

(Ufa

im, mm m t c^rfnt $w

Mice and Voles
Prefer
Spruce
Seeds

SRR1AL RECORDS

Study in Maine suggests
an explanation for the
predominance of balsam fir
seedlings in regeneration

STATION PAPER NO 153 • NORTHEASTERN FOREST EXPERIMENT STATION • 1960
FOREST SERVICE • U S DEPARTMENT OF AGRICULTURE - - UPPER DARBY, PA

RALPH W. MARQUIS, DIRECTOR

/

About the Authors

HERSCHEL G. ABBOTT is Assistant Professor in the Depart-
ment of Forestry and Wildlife Management, University of
Massachusetts, Amherst. He has been engaged in teaching and
research in forest protection and forest regeneration since join-
ing the University faculty in 1953. His former work dealt with
forest management in Maine. He holds a Bachelor's degree in
forestry from the University of Maine and Master's degrees
from Harvard University.

ARTHUR C. HART is a Research Forester at the Northeastern
Forest Experiment Station's research center at Brewer, Maine.
Except for wartime service with other agencies, he has been
engaged in forest research in Maine since joining the staff of
the Northeastern Station in 1938. Most of his research work
has dealt with management of the spruce- fir forest type. He
holds a Bachelor's degree in forestry from the University of
Connecticut and a Master's degree from Yale University
School of Forestry.

Photo credits: photos on cover and page l

BY COURTESY OF U. S. BIOLOGICAL SURVEY.

by Herschel G. Abbott and Arthur C. Hart

Mice and Voles
Prefer
Spruce
Seeds

WHEN spruce-fir stands in the Northeast are cut, balsam
fir seedlings often predominate in the regeneration that
follows. Most landowners would prefer to have the spruce; but
they do not get it, and they wonder why.

Recently-completed studies lead us to believe that the reason
may lie in the feeding habits of mice and other small mammals:
they definitely reject the balsam fir seeds, but eat the spruce seeds.
This would keep down the proportion of spruce in the new stand
of seedlings.

Our attention was drawn to the small seed-eating mammals
several years ago, in connection with research into the silviculture
and management of spruce-fir stands on the Penobscot Experi-
mental Forest near Bangor, Maine. In 1953 a census was made
on the Forest in an attempt to learn something of the population
density of such animals. The mouse and vole population was
found to average nearly 12 animals per acre, and the predominant
species was the red-backed vole.1 (Voles are commonly called

1Grisez, Ted J. Results of a mouse census in a spruce-fir-hemlock stand in Maine.
Northeast. Forest Expt. Sta. Forest Res. Note 28. 2 pp., 1954.

1

mice, though they differ from true mice in several morphological
characteristics.)

Recent laboratory studies of tree seed consumption have estab-
lished the relative preferences of mice and voles for seeds of red
spruce, white spruce, and balsam fir.2 These tests have shown
that these small mammals reject balsam fir seed in favor of spruce
seed.

In the fall of 1959 a study was begun on the Penobscot Forest
in an attempt to determine the preferences of these small seed-
eating mammals under field conditions. This is a report on that
study.

The Study
Areas

Two 1-acre areas about 20 chains apart were used for the
study. Both areas were in mature softwood stands on well-
drained, stony silt loam soils. About half of the stand volume
was eastern hemlock. The remainder was red and white spruce,
northern white-cedar, balsam fir, red maple, and paper birch.

Area 1 was located in a stand that had been selectively cut in
1954, removing 26 percent of the original volume. Area 2 was
in a $tand selectively cut in 1955, removing 19 percent of the
original volume. The logging operations left small openings in
the canopy; and scattered debris consisting of bark, limbs, and
tops from the cut trees remained on the ground. A logging road
and a bucking yard bordered the southern edge of Area 1. A log-
ging road also passed diagonally through the northwestern por-
tion of Area 2, and another passed within half a chain of the
southern border. Both areas contained excellent cover for small
mammals.

There was no seed crop of balsam fir on either area in 1959,
and only a very light crop of spruce seed.

2Abbott, Herschel G. Tree seed preferences of mice and voles in the Northeast.
Manuscript accepted for publication by JOURNAL OF FORESTRY.

2

Procedure

In this study the animals were allowed free access to known
weights of balsam fir and white spruce seed exposed in specially
constructed two-compartment feeders during September and
October 1959. Six feeders were located in each study area. Before
seeds were exposed, a small-mammal census was taken by live-
trapping to determine species and numbers of animals present
on each area. At the completion of the study, a -final census was
made by snap-trapping the animals assumed to be responsible
for taking seeds from the feeders.

Small Mammal Census

The pre-feeding census was started on both areas on September
1, using Sherman live traps. Twenty- five traps were set out in a
grid pattern 40 feet apart on each area. Trap spacing followed
the recommendations of Burt3 and Quick4. Stakes, marked to
show trap line and trap number, were established at each trapping
point on the grids. One trap was then set at the most suitable
location within a 5 -foot radius of each stake. The traps, located
insofar as possible in the immediate vicinity of rocks, stumps,
fallen trees, slash, and other debris that offered cover, were baited
with a mixture of peanut butter, rolled oats, raisins, and bacon
grease. They were checked daily between 8 and 9 a.m.

Captured animals were identified by species, marked by clip-
ping the nail of the second toe of the right hind foot, and
released. Trapping was continued on successive nights until no
unmarked animals were taken. This required 5 nights on Area 1
and 11 nights on Area 2.

The initial small-mammal population on Area 1 consisted of
9 red-backed voles ( Clethrionomys gap peri) and one deer mouse
( Peromyscus maniculatus ) . The population on Area 2 consisted
of 19 red-backed voles, 2 deer mice, and 5 masked shrews ( Sorex
cinereus ) (table 1).

3Burt, William H. Territorial behavior and populations of some' small mammals
in southern Michigan. Univ. Mich. Mus. Zool. Misc. Pub. 45: 1-58, 1940.

4Quick, Horace. Small mammal populations on the University forest. Univ. Maine
Forestry Dept. Tech. Note 28, 2 pp., 1954.

3

Area 2 was trapped initially during the same 5-day period as
Area 1. With the exception of one station, a different lot of
Sherman live-traps were used on each area. One extra trap from
the lot used in Area 1 was set at Station 2-2 in Area 2. For some
undetermined reason, the only captures (3 animals) on Area 2
during the initial 5 -day period were made in this trap.

Table 1. — Small mammal census data: numbers of mice and

voles on study areas

Census

Area 1

Clethrionomys Peromyscus

Total

Pre- feeding
(live trapping)

9* 1

10

Post-feeding
(snap trapping)

20 12

32

Census

Area 2

Clethrionomys Peromyscus Sorex

Total

Pre-feeding
(live trapping)

19** 2 5***

26

Post-feeding
(snap trapping)

6 11 0

17

*One died in handling.
**Two died in live-traps
***A11 died in live-traps.

The two lots of traps were similar in size, shape, and con-
struction, yet the traps used on Area 2 were completely ineffective.
Therefore, at the end of the initial trapping period, all traps
were removed from both areas, and those traps previously set in
Area 1 were reset in Area 2. This second trapping period was
continued for six consecutive nights until no unmarked animals
were captured. This trapping was successful and produced the
additional 23 animals.

4

Seed Exposure

Each feeder consisted of a base, containing two seed trays, and
a- hinged top. Top and base were each 16 inches square and 3
inches deep. The seed trays, approximately 7 x 14 inches outside
measurement and 3 inches deep, were supported by two slats
nailed to the bottom of the base (fig. 1). Copper screening was
used for the bottoms of the seed trays. But since 18 x 14 mesh
would allow some small spruce seeds to pass through, the screen-
ing used on the trays for spruce seed was stretched diagonally
at time of installation to reduce the mesh size.

Figure 1. — Feeder in place, with top open to show trays
and seed. One tray contained balsam fir seed, the other
spruce seed. .

5

Figure 2. — Feeder in set position. The fy-inch opening
lets the small animals in but keeps larger animals and birds
out.

The feeder top was also covered with 18 x 14 mesh copper
screening; and a piece of roofing paper was nailed over the top
to keep the seed dry. Stops and hook fasteners were used at the
front of the feeders to maintain a %-inch opening between the
base and the top; this was to let the small animals in, but to
keep larger animals and birds out (fig. 2).

Cutting tests showed that 9 6 percent of the white spruce seed
and 36 percent of the balsam fir seed was sound. The spruce seed
tested 99 percent in purity, with 186,180 seeds per pound; the
fir was 95 percent pure, *\vith 69,180 seeds per pound. All seed
for the study was weighed and packaged in 1 -ounce units before
the feeding tests were begun.

The feeders were placed near the most productive trapping
stations on each area. They were put out in Area 1 on September
6 and in Area 2 on September 12. Each feeder was supplied with
1 ounce of white spruce seed and 1 ounce of balsam fir seed,
each in its respective tray.

The seed supply was replenished daily or as needed, the
amount depending on the quantity of seed consumed during the

6

preceding 24 hours. A record was maintained o£ the amount of
seed supplied to each feeder. At intervals, the residue was
removed and analyzed for uneaten seeds to determine the actual
amounts consumed or removed. Feeding was continued on both
areas until October 11, when the supply of spruce seed available
for the study was exhausted.

Census Alter Seed Exposure

On October 12, immediately after termination of the seed-
exposure period, a final small mammal census was taken. Three
snap traps were set at each of the 25 trapping stations on each
area. At each station, one trap was baited with a mixture of
peanut butter and rolled oats, a second trap with bacon grease,
and the third with raisins. Trapping was conducted for three
consecutive nights, and daily records of results were kept
(table 1).

In Area 1, 20 red-backed voles and 12 deer mice were caught,
while on Area 2 only 6 voles and 11 deer mice were captured.
All 49 animals were caught in the first two nights of trapping,
and 43 of them were caught during the first night.

An item of no obvious significance in the experiment, but of
general biological interest, was the sex ratios among these
animals. For the deer mice, the ratio of males to females was
1 to 1.4 on Area 1, and 1 to 0.4 on Area 2. For the votes, the
ratio was 1 to 9 on Area 1, and all females on Area 2.

lount of Seed
Consumed

The small-mammal population on the two areas consumed
25.75 pounds (412.05 ounces) of white spruce seed and 1.01
pounds (16.10 ounces) of balsam fir seed (table 2). This indi-
cates a preference ratio of 25.5 to 1 in favor of spruce over fir,
based upon the weight of commercially-clean seed.

7

Table 2. — Quantity of commercially clean seed exposed and
consumed or removed by small mammals , in ounces

Item

Area

1

Area 2

Total

Spruce

Fir

Spruce

Fir

Spruce

Fir

Seed exposed

259.00

12.00

156.00

12.00

415.00 24.00

Seed recovered

in residue

1.51

3.57

1.44

4.33

2.95

7.90

Seed actually
consumed or
removed

257.49

8.43

154.56

7.67

412.05

16.10

However, this ratio is somewhat unrealistic because of the
difference between the percentages of sound seed in the two seed
lots. When the consumption figures are adjusted to represent
the weight of only sound seed consumed, it is found that 24.47
pounds of white spruce seed and 0.35 pound of balsam fir seed
were consumed. The preference ratio of spruce to fir is thus
increased to 70 to 1.

The relative consumption of different kinds of seeds may be
greatly influenced also by their availability. In this study, no
balsam fir seed was eaten or removed as long as white spruce
seed was available in the same feeder. The fir seed went un-
touched on both areas from the date of first exposure until the
weekend of September 26 and 27. On this weekend, the feeders
were not tended and the supply of spruce seed was exhausted.
Apparently it was only under these conditions that the mice were
interested in the fir seed.

Since there was no natural seed crop of balsam fir seed and
only a very light crop of spruce seed on the study areas, seed
availability, for all practical purposes, was limited to the seed
offered in the feeders.

On September 30, consumption of spruce seed began to show
a marked increase (fig. 3), so the amount of seed exposed in
each feeder was increased from 1 ounce to 2 ounces. It was at
this time that the second replenishment of balsam fir seed was

8

OUNCES OF SEED CONSUMED

made — 1 ounce in each feeder. The following day the supply of
white spruce seed was increased to 3 ounces. Even this quantity
was not adequate to satisfy the demands of the mice; soy on
October 2, 4 ounces of seed were placed in each feeder. From
this time on, 4 ounces were added as needed until October 11,
when our supply of spruce seed was exhausted.

In comparing seed consumption between areas, we found little
difference in the amounts of fir seed consumed (8.43 ounces on
Area 1 versus 7.67 ounces on Area 2).

Consumption of spruce seed, however, was much greater on
Area 1 than on Area 2: 257.49 ounces versus 154.56 ounces. This
difference probably was due to the greater number of mice on
Area 1. On the basis of the post-feeding census (table 1), 32
animals were feeding on Area 1, as compared to 17 on Area 2,
during the latter part of the test period when the sharply accel-
erated seed consumption occurred. Consumption of spruce seed

5 9 13 17 21 25 29 3 7

SEPTEMBER % OCTOBER

Figure 3. — Cumulative consumption of white spruce and
balsam fir seed.

9

per animal for the entire test period — again on the basis of the
post-feeding census — averaged 8.04 ounces on Area 1 and 9-09
ounces on Area 2. This difference possibly is related to the dif-
fering proportions of Clethrionomys and Peromyscus in the two
populations (table 1).

In comparing seed consumption on the two areas, a minor
detail to note is that the feeding period was started 6 days earlier
on Area 1 than on Area 2. However, this had little influence on
the total results, as only 3 ounces of spruce seed were consumed
on Area 1 during the initial 6-day period.

Discussion

The fact that balsam fir seed was consumed in this study only
when spruce seed was not available supports previous observations
by Abbott that Clethrionomys and Peromyscus prefer spruce seed
over balsam fir seed. The sharp increase in seed consumption
beginning about October 1, or 3 to 4 weeks after seed was first
placed in the feeders, was of particular interest. Such an increase
in seed consumption might come about in two ways: (1) from
an increase in number of animals present, or (2) from increased
activity by animals already present in removing seed for storage.

The authors believe that the increased consumption in this case
primarily reflects increased seed removal for storage. Considerable
mouse activity was noted at this time, and it was not unusual to
observe mice visiting the feeders during daylight hours. The best
evidence, however, was in the fact that the amount of seed-coat
residue left in the feeders became progressively smaller in several
feeders during this period. A similar observation on the storage
of white pine seed, based upon seed coat residues left in feeders,
has been made by Abbott.5 He found that white-footed mice
( Peromyscus leucopus) and red-backed voles stored approximately
50 percent of the seeds removed.

5Abbott, Herschel G. White pine seed consumption by small mammals. JOUR.
FORESTRY 59: 197-201. 1961.

10

The seed-storage theory is further supported by the fact that
the patterns of increased consumption on the two areas in early
October were similar, whereas, according to the census data
(table 1), changes in the mouse-vole populations during the
test period were markedly dissimilar (increasing from 10 to 32
on Area 1, decreasing from 21 to 17 on Area 2). The increased
consumption thus seems clearly to reflect a change in the activities
of the animals that is unrelated to population changes. The
logical assumption is that the animals rather abruptly entered a
phase of seed-storage activity.

The cumulative seed consumption shown in figure 3 and the
consumption totals in table 2 provide a quantitative picture of
spruce seed losses of a magnitude that might substantially in-
fluence spruce regeneration. The data show that a mixed popu-
lation of 15 to 30 red-backed voles and deer mice per acre
possesses a consumption potential for spruce seed far exceeding
the amount normally produced in natural seed crops.6 Our study
was terminated 'during the peak of storage activities by the
animals, so we have no measure of the amount of seed they
could have taken. However, in a 5 -week period they had already
taken some 13 pounds of spruce seed per acre, almost 10 pounds
of which they took in less than 2 weeks of intensive storage
activity.

Naturally disseminated seed on and in the forest floor un-
doubtedly would not be found and removed as rapidly as the
freely accessible seed in the feeders. However, the capacity for
consuming much more seed than natural crops provide, coupled
with the marked preference shown for spruce seed, suggests that
the animals may be largely responsible for the low ratio of spruce
to fir seedlings commonly found under spruce-fir stands in the
Northeast.

6The spruce seedfall per acre in 8 softwood stands on the Penobscot Experimental
Forest in 1953, when both red and white spruce trees produced heavy cone crops,
ranged from 145,000 to 1,529,000 seeds per acre, with an average of 818,000.
Assuming an average of 190,000 seeds per pound, these figures convert to 0.75 to 8
pounds, averaging 4.3 pounds per acre on the 8 areas sampled.

11

Summary

A study conducted on the Penobscot Experimental Forest near

Bangor, Maine, during September and October 1959, dealing

with the relative consumption of white spruce and balsam fir

seed by small mammals showed that:

• The animals present (red-backed voles and deer mice) ex-
hibited a definite preference for spruce seed.

• They ate balsam fir seed only when spruce seed was not
available.

• Large quantities of spruce seed apparently were stored in
addition to the amount eaten.

• The daily rate of seed consumption was not constant but
showed a rapid increase starting about October 1, 3 weeks
after the study was begun. This trend is believed to be asso-
ciated with seed storage.

• Approximately 26 pounds of spruce seed were eaten or stored
in a 5 -week period by the mice and voles on two 1-acre plots.

• The number of animals presumably responsible for destroying
this quantity of seed was 49 (26 red-backed voles and 23
deer mice) .

• Spruce seed consumption per animal for the 5 -week period
averaged 8 to 9 ounces.

12

Acknowledgment

The spruce seed used in this
study was contributed by the Maine
Forest Service through Forest Com-
missioner Austin Wilkins. Dr.
Horace Quick and Dr. Sanford
S chemnitz of the University of
Maine School of Forestry assisted
in gathering the small -mammal
population data.