Brodifacoum in bait stations for managing Columbian ground squirrels

S Barllf Steven F

632«6 Brodl facouD) in

3232 bait stations for

A3tr83— 2manaainj4 Columbian
ground squirrels

MONTANA STATE LIBRARY

S 632.63232 A3tr 83-2 c. 1 Baril
Brodifacoum in bait stations for managin

3 0864 00061685 7

Montana Department of Agriculture
Environmental Management Division
Technical Services Bureau
Helena, Montana 59620

STATE DOCUMENTS COLLECTION

MAR 1 0 1989

MONTANA STATE LIBRARY
1515 E- 6th AVE.
HELENA, MONTANA 59620

Brodifacoum in Bait Stations for Managing Columbian Ground Squirrels

By

Steven F. Baril and Daniel Sullivan

Abstract

A pelleted grain bait containing 0.005% brodifacoum was placed in bait
stations to determine effectiveness for managing Columbian ground
squirrels. Ground squirrel activity, measured by direct visual counts,
was reduced an average 96.84% (range 95.43-100.00) on 3 treated plots.

A total of 11.55 kg (25.46 lb) of bait was consumed by squirrels in
treating 3.87 hectares (9.56 acres) of pasture or hayland.

Technical Report 83-2

December 1983

INTRODUCTION

Brodifacoum, a second generation anticoagulant, may be an effective
toxicant for managing ground squirrels. The for the Richardson

ground squirrel (Spermophilus richardsoni) was determined to be 0.130
mg/kg (95% CL = 0.062 - 0.188 mg/kg) (Baril and Pallister 1981), and is
probably similarly low for the Columbian ground squirrel (S .
columbianus) (Matschke, personal communication) . Matschke ^ ^ (1983)
showed that a 50 ppm bait applied once by hand (1 tbsp/burrow) reduced
populations of Richardson ground squirrels by 96.8 ± 1.5%.

Currently, the efficacy of brodifacoum has not been field tested on the
Columbian ground squirrel in Montana. This study was designed to test
the efficacy of a commercially available bait containing 50 ppm
brodifacoum (Talon-G TM Rodenticide) applied in bait stations to manage
Columbian ground squirrels. Sullivan (1982) demonstrated that bait
stations containing anticoagulant baits could be used to effectively
manage this species, and that bait stations might reduce the hazard of
primary poisoning of nontarget animals that can be associated with
baits scattered on the ground.

METHODS AND MATERIALS
Plot Selection

Study plots were selected near Helena, Montana (Lewis and Clark County)

1 •

in April, 1983. This intermontane valley is at the eastern edge of the
geographic range of Columbian ground squirrels.

Six plots, 3 controls and 3 test plots, were established on pasture and
hay fields that were colonized by Columbian ground squirrels (Table 1) .
The distance between the two most distant plots was about 26.2
kilometers (16.25 miles) , and plots were selected in proximal pairs
consisting of one control and one test. Orginally, an attempt was
made to select noncrop areas on which to conduct the test. This was
abandoned because of difficulty in finding colonized areas not cropped
or pastured and where vegetation height permitted visual observation of
squirrels .

An additional criterion for plot selection was that squirrel colonies ^
covered 1 to 5 acres and be confined by natural buffers or noncolonized
areas. Colonies of this size provided a sufficient number of squirrels
to monitor efficacy and natural buffers helped to preclude reinvasion
of treated areas by dispersing squirrels. In all cases, plots included
the entire colonized area.

Bait Stations

Because of immediate availability, two styles of bait stations were
employed, a commercially available station (Eco-bait TM) and one
constructed from government surplus 50 caliber metal ammunition boxes.

2

Table 1. Description of study plots.

Plot No.

Treatment

o ■ 3.

Size

(Hectares /acres)

Vegetation and
Agricultural Use

1

Test

1.06/2.63

Grass; pasture/hay

2

Control

0.47/1.16

Grass; pasture/hay

3

Test

1.03/2.54

Grass and alfalfa;
pasture/hay ^

4

Control

1.05/2.60

Sage and grass border,
alfalfa; pasture/hay ^

5

Test

1.78/4.39

Grass; pasture

6

Control

1.66/4.10

Grass and alfalfa;
pasture/hay

Includes gross treated area.

^ Pasture/hay denotes areas used as pasture during fall, winter and
spring with 1 or 2 hay crops harvested during the summer.

Both were tested by Sullivan (1982) against the Columbian ground ^

squirrel and no differences in effectiveness were detected. Both
styles allow access by squirrels to a protected source of bait.
Descriptions of each were presented by Sullivan (1982) . The stations
were used in an equal ratio on all plots, 3 Eco-bait stations to 1 ammo
box.

Bait stations were placed on each plot at intervals of 36.6 to 45.7
meters (120 to 150 feet). Where visual observation of squirrel burrows
indicated population centers, the closer interval was used. Stations
were arranged so that the approximate maximum distance from any single
burrow on the plot to a bait station was 22.9 meters (75 feet). This
was a reasonable distance considering the findings of Alsager (1972)
who marked a numJDer of Richardson ground squirrels v;ithin 100 feet of
bait stations and, in an unspecified period of time, found that 77.7%
of the marked individuals visited the stations.

Bait stations were placed on plots 1 through 5 on April 13 and on plot
6 on April 14 , 1983 . Initially, all stations were prebaited v/ith
commercial horse feed made from whole, rolled oats and molasses.
Previous experience has shown this to be a palatable bait for Columbian
ground squirrels. Stations were checked at 2 day intervals and bait
replenished as necessary. When pretreatment visuals were complete
prebait was replaced with Talon-G, a grain-based pellet containing 50
ppm brodifacoum, in the stations on test plots. Using the same
criteria, bait in the stations on control plots was switched to
nontoxic pellets used as rabbit feed containing alfalfa and grain. ^

4

Bait consumption was measured by recording the amount of bait added to
stations minus that remaining at the completion of the study.

Efficacy Determination

Efficacy was determined by comparing pre and post treatment visual
counts (activity indices) of squirrels on the entire treatment area
following standard methods proposed by Fagerstone (1982). In a single
exception a small baited area of plot 1 was not included in visual
observations because of poor visibility. An advantageous location from
which to conduct counts was established for each plot and used
throughout the study.

Pretreatment counts were started when at least half of the prebaited
stations were being used by squirrels, and counts were then conducted
daily for 3 consecutive days, weather permitting, with the last count
being just prior to the introduction of toxic bait (test plots) or
nontoxic pellets (control plots) . Post treatment counts were started 5
days after the introduction of toxic bait or rabbit pellets and
conducted at 2 day intervals, weather permitting. Post-treatment
counts were conducted over a period of 22 days.

Each daily visual count on a single plot was conducted by carefully
counting all live squirrels on the counting area using 7 power
binoculars. This was repeated 3 times at 5 minute intervals and the
highest count of the 3 used for determining efficacy.

5

A paired t test was used to determine the statistical significance of ^
the change in pre and post treatment activity indices {A. I.) on control

plots. After it was determined that pre and post treatment activity
indices from control plots were not significantly different; efficacy,
or the percent change in activity post treatment, was determined for
treated plots using the following formula:

Percent Average Pretreatment _ Average Post Treatment

Reduction = A. I . A. I . x 100

In Activity Average Pretreatment A. I.

During bait station maintenance, plots were observed for the presence
of carcasses of ground squirrels or nontarget animals.

RESULTS

Bait Consumption

On plots 1 and 2 squirrels began to investigate stations soon after
placement and within 6 days over half of the stations were being used.
On plots 4 and 5 initial use began after 5 days and after 7 days over
half of the stations were used. On plots 3 and 6, initial use began
after 8 days and not until 13 days after placement were half of the
stations being used. After an initial period of accustomization ,
stations were readily utilized and commonly emptied of their contents
(2 cups in Eco-baits and 4 cups in ammo box stations) in a day. At the
time toxic bait was placed, nearly all of the stations were being used.

6

The net amount of toxic bait consumed by squirrels during the test was
11.55 kg (25.46 lb) or about 0.43 kg (0.95 lb) per station or 3.01
kg/hectare (2.55 Ib/ac). The net amount of nontoxic pelleted bait
consumed on control plots was 15.21 kg (33.53 lb) or 0.50 kg (1.75 lb)
per station or 4.78 kg/hectare (4.27 Ib/ac) . Table 2 shows the amount
of control bait and toxic bait consumed post treatment per plot.

The pelleted formulation of Talon-G was acceptable to squirrels; in
fact, no apparent changes in bait consumption or station use were
detected when prebait was replaced with Talon-G.

On plot 1 each station was replenished with approximately 200 grams of
bait 3 days and 6 days after introduction of toxic bait. Very little
additional bait consumption occurred after this. A single bait station
on each of plots 3 and 5 was refilled 5 days after introduction of
toxic bait; however, bait consumption had essentially ceased by this
time .

Control stations were used consistently by squirrels throughout the
post treatment period. However, there was some observational evidence
that the alfalfa and grain pellets used as control bait were less
acceptable than either the prebait or brodifacoum formulation.

Efficacy

From May 9 until May 13, visual monitoring was interrupted by a major
spring storm which resulted in 6-8 inches of snow on the plots. Also

7

Table 2 Net amount of nontoxic and toxic bait consumed post

treatment, and average pretreatment activity index per plot.

Plot

Bait

Toxic

Consumption (q)
Nontoxic

Average
Pre- treatment
Activity Index

1

7275

-

27.33

2

-

9791

10.33

3

1547

-

12.33

4

-

1534

5.67

5

2731

-

15.33

6

—

3887

14.00

8

on May 2 and 6 visual observations were not conducted because of cool,
wet weather and poor squirrel activity. Results of visual monitoring
are presented in Appendix 1.

The combined average activity index for control plots was slightly
lower pretreatment than post treatment (Table 3) , although not
significantly (paired t =2.56, d.f=2,p greater than 0.1). On treated
plots, the average activity index declined significantly by the first
post treatment visual counts 5 to 6 days post treatment. Activity
indices on treated plots declined to zero by 17 to 18 days post
treatment but by 22 days post treatment 4 squirrels were observed on
plot 1, and 1 on plot 3.

Total numbers of ground squirrels counted on treated plots were lowest
on day 12 post treatment and subsequent counts. Consequently, the 4
counts conducted on each treated plot after 12 days post treatment were
compared with pretreatment counts to determine efficacy (Table 4).

Six days after stations were filled with toxic pellets, 4 moribund
ground squirrels were present above ground on plot 1, and 8 days post
treatment an additional 2 carcasses and 1 moribund squirrel were
observed. Five days post treatment, 2 ground squirrel carcasses were
observed on plot 3. No ground squirrel carcasses were found on control
plots, and no carcasses of other animals v/ere observed on any of the
plots. Magpies (Pica pica) were observed feeding on squirrel carcasses
on plots 1 and 3.

9

Table 3

Pre and post treatment Columbian ground squirrels activity
indices and average indices for control plots and plots
treated with 0.005% brodifacoum bait.

Treated

Plots

Control

Plots

Days Pre/Post

Plot

Number

Plot

Number

Treatment

1

3

5

Ave .

2

4

6

Ave .

-2,3

28

11

13

17.33

10

3

12

8.33

-1

25

14

15

18.00

10

8

14

10.67

0^

29

12

18

19.67

11

6

16

11.00

+5-6

5

8

13

8.67

12

7

15

11.33

+ 8-10

6

5

5

5.33

17

6

14

12.33

+12-15

1

0

0

0.33

16

9

9

11.33

+17-18

0

0

0

0.00

21

7

8

12.00

+ 20

0

1

0

0.33

20

5

14

13.00

+ 22

4

1

0

1.67

15

5

15

11.67

Day 0 corresponds to the day that toxic pellets or nontoxic pellets
were placed in bait stations on treated and control plots,
respectively.

10

r

r

r

Table 4 Efficacy of 0.005% brodifacoum bait placed in bait stations for control of Columbian
ground squirrels.

Average Activity Indices

Treated Plots Control Plots

1

3

5

Ave .

4

6

Ave

Pretreatment

27.33

12.33

15.33

18.33

10.33

5.67

14.00

10.00

Post treatment

1.25

0.50

0.00

0.58

18.67

5.67

12.33

12.22

Percent Change ^

-95.43

-95.94

-100.00

-96.84

+80.74

0.00

-11.93

22.20

a

+ or

indicates increase or decrease in activity, respectively

DISCUSSION

There is some indication that springtime activity was still increasing
on plot 2 at the time pretreatment visual counts commenced on April 22.
The first ground squirrels emerging from hibernation near the study
area were observed on March 19 in a flat hay meadow. Shaw (1925a)
indicated that emergence is sequential, adult males appearing first,
followed by adult females about 7 to 15 days later. He indicated that
yearlings might emerge even later. In a local area, emergence can vary
as much as 10 days depending upon slope (Shaw 1925b) , squirrels on
north facing slopes emerging later. From this information it v/ould
appear that springtime activity across the area of the present study
would have been sufficiently underway. However, this may not have been
the case because, as shown in table 4, activity on plot 2 underwent a
dramatic increase. Plots 1 and 2 were at a slightly higher elevation
than other plots and on a decidedly north facing slope which may have
caused emergence to be delayed compared to the other plots.

Dispersal of yearling squirrels is another source of recruitment;
however, it is not thought to be the cause of the increase.

Recruitment was not observed to any appreciable extent on other plots
and probably occurred at a later date than the observed population
increases on plot 2 (Betts 1973) .

Activity indices dropped to zero on all treated plots beginning about
15 days after stations were baited with brodifacoum pellets. Beginning
about 20 to 22 days post treatment, a few active squirrels were

12

observed on plots 1 and 3. It is suspected that these squirrels had
immigrated to the plot and were not survivors of the treatment. Later
visual observations were conducted at a time when dispersal of
yearlings is thought to occur, during the lactation period of adult
females (Betts 1973). It may well be that the average 96.84%
population reduction on treated areas is a slightly conservative
estimate .

An initial period of about 7 to 14 days was required before significant
use of stations by squirrels began to occur. This corresponds with
other studies where an initial period of accustomization was noted
(Clark 1978, Sauer 1976, Sullivan 1982) and may be due to a general
reluctance of v/ildlife to approach new or novel objects in their
environments. After their initial shyness of the bait stations,
squirrels in this study readily consumed oats used as prebait and grain
based pellets containing brodifacoum.

This study demonstrated that 0.005% brodifacoum bait (Talon-G) can be
effectively used in bait stations for managing Columbian ground
squirrels, is apparently readily accepted into the diet after a period
of prebaiting, and can be effective early in the spring shortly after
the squirrels emerge from hibernation.

13

9

LITERATURE CITED

Alsager, D.E. 1972. Experimental population suppression of

Richardson's ground squirrels (Spermophilus richardsonii) in
Alberta. Proc. Fifth Vertebrate Pest Conference, Fresno, CA
93-100.

Baril, S. and Pallister, G. 1981. Laboratory determinations of
acute oral and efficacy of brodifacoum on the

Richardson ground squirrel, Spermophilus richardsonii .
Montana Dept. Agriculture, Tech. Rep. 81-1, Helena, MT. 15
pp.

Betts, B.J. 1973. The adaptiveness of the social organization of

a population of Columbian ground squirrels (Spermophilus
columbianus) . PhD. Dissertation, U. of Montana. 235 pp.

Clark, D.O,

1978. Control of ground squirrels in California

using anticoagulant treated baits. Proc. Eighth Vertebrate
Pest Conference, Sacramento, CA 98-111.

Fagerstone, K.A. 1982. Comparison of direct counts and

mark-recapture as techniques for censusing ground squirrels.
Test Methods for Vertebrate Pest Control and Management

Materials, ASTM STP , D.E. Kaukeinen, Ed., Am. Soc.

Test, and Materials. (in press).

Matschke, G.H., Fagerstone, K.A. , Baril, S., and Balaskiewicz ,

R.V7. 1983 . Population reduction of Richardson's ground
squirrels using a brodifacoum bait. U.S. Fish VJildlife
Service, Denver Wildlife Res. Center, Job Completion Rep.,
Denver, CO 35 pp.

Sauer, W.C.

Shaw, W.T,

1976. Control of the Oregon ground squirrel
(Spermophilus beldingi oregonus) . Proc. Seventh Vertebrate
Pest Conference, Monterey, CA. 99-109.

1925a. Duration of the aestivation and hibernation

of the Columbian ground squirrel (Citellus columbianus) and
sex relation to the same. Ecology VI: 75-81.

Sullivan, D. 1982. Bait stations as a means of rodenticide

presentation to control Columbian ground squirrels. Montana
Dept. Agriculture, Tech. Rep. 82-3, Helena, MT. 26 pp.

Shaw, VJ. T.

1925b. The seasonal differences of north and south

slopes in controlling the activities of the Columbian ground
squirrel. Ecology VI: 157-162

14

Appendix 1 Pretreatment and post treatment activity indices a

and dates conducted to determine the efficacy of
0.005% brodifacoum for managing Columbian ground
squirrels .

Treated Plots Control Plots

Date

1

3

5

2

4

6

4-22

28

10

4-24

25

10

4-25

2 9^

11^

4-27

11

13

3

12

4-28

14

15

8

14

4-29

12^

is’^

6^

16^

5-1

5

12

5-3

8

15

5-4

8

13

7

15

5-5

4

19

5-7

1

16

5-8

5

5

6

14

5-13

0

21

5-14

0

0

9

9

5-15

0

20

5-16

0

0

7

8

5-17

4

15

5-19

1

0

5

14

5-21

1

0

5

15

15

Each index is the highest count of 3 consecutive counts taken on that
date and location.

Following
or nontoxic
treated and

the completion of these indices, 0.005% brodifacoum pellets
rabbit food pellets were placed in the bait stations on
control plots respectively.

16