Grazing Effects on Deer Mice with Implications to Human Exposure to Sin Nombre Virus

Scientific Disciplines
Biological Sciences - Terrestrial
University of Montana
deer mice
peromyscus maniculatus
Sin Nombre virus
Montana Tech of the University of Montana
Volume 17, No. 1-4

Grazing effects on deer mice with implications 
To human exposure to sin nombre virus
Abigail J. Leary, University of Montana, Interdisciplinary Studies Graduate Program, Missoula, MT 
59812; Department of Biology, Montana Tech of the University of Montana, Butte, Montana 59701
Amy J. Kuenzi, Department of Biology, Montana Tech of the University of Montana,  
Butte, Montana 59701
Richard J. Douglass, Department of Biology, Montana Tech of the University of Montana,  
Butte, Montana 59701
We examined the effects of grazing on deer mouse (Peromyscus maniculatus) movements 
into buildings using passive integrated transponder (PIT) technology and small simulated 
buildings located on 0.6-ha treatment (grazing) and control (no grazing) plots. Twelve 
experimental 9-day trials were conducted over the course of the study. During these trials, 
mouse movements into buildings were monitored during three time periods (each 3 days in 
length). In the treatment plots these time periods corresponded to pre-grazing, grazing, and 
post grazing by horses. The number of individual deer mice entering buildings over time 
decreased in both the grazed and control plots during the 9 days of each experiment. The 
number of entrances per/individual among the pre-grazing, grazing and post grazing periods 
was different between control and treated plots for both males and females. The distribution 
of entrances/individual among the three periods differed between males and females in both 
grazed and control plots. The habitat modification caused by grazing appeared to reduce 
deer mouse activity (entrances/individual) in buildings but does not affect the number of 
mice entering buildings. Reducing vegetative cover by grazing or mowing may not affect the 
number of mice investigating small structures but grazing creates different activity patterns in 
the structures for neighboring deer mice. 
Key Words:  hantavirus, deer mouse, Sin Nombre Virus (SNV), Peromyscus maniculatus
rate of 35 percent (CDC). In many HPS 
Ecological and environmental changes 
cases, human exposure to SNV has been 
due to changing land use practices may 
linked to contact with infected rodents and/
or their excreta in and around buildings 
provide opportunities for increased 
(peridomestic settings) (Armstrong et al. 
transmission of infectious diseases 
1995). However, the ecology of deer mice in 
(Woolhouse and Gowtage-Sequeria 2005) 
peridomestic settings is not well understood 
by increasing human contact with zoonotic 
(Kuenzi and Douglass 2009, Kuenzi et al. 
hosts, their ectoparasites, and the diseases 
2001), and little data exist on what causes 
they carry. Rodents, such as deer mice 
deer mice to move from surrounding natural 
(Peromyscus maniculatus), are reservoir 
areas into peridomestic settings. 
hosts for many infectious diseases, including 
Modification of surrounding habitat 
hantaviruses (Morse 1995). Hantaviruses 
is one factor that may cause mice to enter 
are rodent-borne pathogens that can cause 
buildings. Livestock grazing is probably 
serious human illnesses. In the United 
the most common habitat modification in 
States, the deer mouse is the principal 
Montana. Small mammal populations and 
reservoir for Sin Nombre virus (SNV) 
communities can be directly and indirectly 
(family Bunyaviridae, genus Hantavirus
affected by grazing (Hayward et al. 1997). 
(Childs et al 1994, Nichol et al. 1993). SNV 
Trampling of burrows, compacting soil, 
causes Hantavirus pulmonary syndrome 
and competition for food resources are 
(HPS) a serious human illness with a fatality 
direct effects of grazing, while altering 

vegetation structure that then influences 
life traps (8 x 9 x 23 cm, H.B. Sherman 
habitat selection by small mammals is an 
Trap Co., Tallahassee, Florida, USA), with 
indirect effect.  Significant changes in the 
10-m spacing between traps, were set with 
nutritional dynamics and physical structure 
the experimental buildings in the center 
of vegetation have been caused by grazing 
of the grid, and checked each day for 3 
by both bison and cattle (Damhoureyeh 
consecutive days. Traps were baited with 
and Hartnett 1997) and presumably horses. 
peanut butter and oatmeal and contained 
Such alterations could potentially cause deer 
polyester bedding. All captured animals 
mice to leave their normal habitat and enter 
were transported to a central location for 
peridomestic settings. 
processing. Species, body mass, sex, age, 
In an effort to improve our 
and reproductive condition of captured 
understanding of the relationship between 
animals were recorded. Deer mice were ear-
anthropogenic environmental changes and 
tagged with monel #1005-1 tags (National 
human exposure to SNV, we examined the 
Band and Tag Co., Newport, Kentucky). PIT 
movement of deer mice into peridomestic 
tags (Passive integrated transponder tags, 
settings in response to habitat modification 
BioMark, Inc; Boise, Idaho) were adhered 
(grazing). We hypothesized that grazing in 
directly to the skin between the shoulders 
areas around buildings would increase deer 
of each deer mouse. A small patch of fur 
mice movement into those buildings. To test 
was shaved away to secure the PIT tag 
this hypothesis we monitored deer mouse 
closer to the animal.  The adhesive was then 
movements into simulated buildings before, 
used to coat over the PIT tag and glue the 
during, and after grazing by horses in the 
surrounding fur over the tag to help with 
surrounding habitat. We used horses to graze 
tag retention. Based on modifications to 
the surrounding habitat because they were 
methods used in previous work, (Kuenzi 
easier to control and move than cattle.  The 
et al 2005) we assumed pit tag retention 
results from this study are necessary for the 
was nearly 100 percent for the duration of 
development of recommendations to help 
each trial. A hand held reader was used to 
reduce the risk of human exposure to rodent 
verify that the PIT tags functioned after 
borne diseases. 
attachment. PIT tag numbers were then 
recorded and individuals were released at 
the point of capture.  
Study Site 
Experimental Design
This study was conducted near Gregson, 
The effects of habitat modification (by 
Silver Bow County, Montana. The dominant 
grazing) on mouse entries into buildings 
vegetation at the study site consisted of 
and availability of food resources (rolled 
antelope bitterbrush (Purshia tridentata), 
oats) within buildings were examined using 
spotted knapweed (Centaurea maculosa), 
a series of 12 experimental trials. For each 
cheatgrass (Bromus tectorum), and big 
trial, one treatment (grazing) plot and one 
sagebrush (Artemisia tridentata).
control (no grazing) plot (0.6 ha) were used. 
Two small simulated buildings were placed 
Rodent trapping And Processing
in the center of each plot. Each simulated 
We conducted 12 experimental trials (9 
building was 2.44 m x 1.22 m x 1.22 m 
days each) from April through June 2005; 
with a 5-cm diameter opening in one end. 
October through November 2005; February 
Our simulated buildings do not represent 
2006; June through August 2007; and May 
all features that actual outbuildings would 
through July 2008. Deer mice were trapped 
present to mice; however, we feel the initial 
and marked 3 days prior to each trial.  In 
response by mice to simulated buildings 
each plot we attempted to saturate the 
would be similar to that of actual buildings. 
plot with traps set in grids containing (see 
Further experimentation would be required 
experimental design below), five rows of 
to determine what would allow deer mice 
25 traps (Sherman non-folding, aluminum 
to establish residency in buildings. One 
Grazing Effects on Deer Mice with Implications to Human Exposure to Sin Nombre Virus       31

building in each plot contained ~ 1 kg 
and the home range size of deer mice.  
of rolled oats. An electrical fence was 
Therefore, this experiment was replicated by 
constructed around treatment pastures to 
performing multiple (N = 12) experimental 
control horses.
trials. The location of treatment and control 
We equipped each building with a 
plots were randomized among a set of 
passive integrated transponder (PIT) tag 
pastures within a single ranch. Accordingly, 
transceiver (Model 2001F, Biomark Inc, 
the 12 experimental trials were conducted 
Boise, Idaho) linked to an antenna located 
over a relatively long period (Apr 2005 to 
around the building opening. The antenna 
Jul 2008) to enhance the independence of 
detected pit-tagged deer mice that entered/
observations, i.e., reduce the probability that 
exited the building. The pit tag number, 
individual mice would become habituated to 
date and time of the entrance/exit were 
the experiment.
recorded in the transceiver. Transceivers in 
each building were activated in the evening 
and turned off every morning to conserve 
Of the 174 deer mice fitted with PIT 
electricity. Each transceiver was powered 
tags, 69 (39.7%) entered buildings at least 
by a 12-volt deep cycle battery charged by 
once. Out of these 69 individuals, only 
a solar panel. Transceivers were retrieved 
four individuals entered buildings on all 9 
from the field in the morning and data from 
days of a given trial. Across all trials and 
the previous night were downloaded onto 
plots, 63 of the 69 individual mice entered 
a desktop computer. Movement into these 
buildings during the pre-grazing period, 
buildings was monitored for 9 nights during 
42 during grazing period and 32 during the 
each experimental trial. 
post grazing period. Some individual mice 
Buildings in both the treatment and 
were recorded during multiple periods and 
control plots were monitored 3 days prior to 
were included in the number of individuals/
introducing the horses (pre-grazing). Both 
plots were monitored for 3 days with the 
Across all trials in the treatment plots, 
horses in the treatment plot (grazing) and 
we recorded a total of 26 individual mice 
then monitored again for 3 days after the 
entering buildings with 24 individuals 
horses were moved out (post grazing). Six 
entering during the pre-grazing period, 18 
horses were used to graze each experimental 
during the grazing period, and 14 during 
plot.    The horses basically removed all 
the post grazing period (Fig. 1). In the 
herbaceous vegetation (to within 2 cm of the 
control plots across all trials, 43 individuals 
soil) within the three days they were present 
entered buildings with 39 mice entering the 
in the treatment plots. All plots had an entire 
buildings during the pre-grazing period, 24 
winter and growing season between trials. 
during the grazing period, and 18 during the 
Data Analysis
post grazing period. Regardless of treatment, 
To determine if habitat modification by 
the number of mice entering the buildings 
grazing would affect deer mouse entrances 
was greatest during the first 3 days of the 
into our simulated buildings, we used Chi-
trial (pre-grazing period) and smallest during 
square analyses (Zar 1984) to compare the 
the last 3 days of each trial (post grazing 
numbers of individuals entering buildings 
period; Fig. 1). The number of individuals 
(a numeric response) and the number of 
entering buildings among periods was 
entrances/individual (a behavioral response) 
similar between control and treated pastures 
during the pre-grazing, grazing and post 
(χ2 = 0.017, P > 0.05) (Fig. 1) with the 
grazing periods between the control and 
number of individuals entering buildings 
grazed pastures summed across all trials. A 
declining in both control and treatment 
P-value of 0.05 was used for all analyses.
plots over the 9 days of a trial. In control 
It was logistically difficult to spatially 
plots 18 individual deer mice accounted 
replicate this experiment due to the large 
for 87 percent of movements into buildings 
area necessary, a limited supply of horses, 
whereas nine individuals accounted for 70 
32          Leary et al.

Figure 1. The number of deer mice entering simulated buildings in grazed versus control 
(non-grazed) plots in Southwestern Montana from 2005 through 2008.
percent of the movements into buildings.
sample sizes were insufficient for seasonal 
The behavioral response (number of 
entrances/individual) was significantly 
The presence of food had no effect 
different between control and grazed plots 
on entrances/individual in the grazed plots 
for both females (χ2 = 7.68, P < 0.05) and 
(food vs. no food, χ2 = 1.68 P > 0.05). 
males (χ2 = 95.68, P < 0.001; Fig. 2). The 
However in the control, entrances/individual 
number of female entrances into buildings 
increased continuously in the building with 
in grazed plots decreased over 9 days but 
food and remained fairly constant in the 
remained fairly constant in the control plots. 
building with no food over the nine days  
Male entrances also decreased in the grazed 
(χ2 = 17.67, P < 0. 001; Fig. 4).
plots but increased in the control plots from 
one period to the next.
Males and females responded 
Research on the effects of livestock 
differently in terms of numbers of entrances 
grazing on deer mouse ecology has yielded 
per individual through the duration of a trial 
various results for different vegetation 
in both the grazed plots (χ2 = 12.38, P < 0.05) 
types (Douglass and Frisina 1993, Clary 
and the control plots (χ2 = 26.36, < 0.001; 
and Medin 1992, Medin and Clary 1989, 
Fig. 3). However, overall responses differed 
Oldemeyer and Allen-Johnson 1988). In this 
in the grazed plots compared to the control 
study we found no difference in the number 
plots. Both female and male entrances 
of individuals entering buildings in grazed 
decreased over the nine days in the grazed 
versus ungrazed plots. Our movement data 
plots with males decreasing more than 
are consistent with Oldemeyer and Allen-
females. In the control plots, male entrances/
Johnson (1988), who found little or no 
individual increased during the nine days 
difference in deer mice abundance between 
while female entrances remained relatively 
grazed and ungrazed sites with dominant 
constant. Most animals in the experiments 
vegetation types of spotted knapweed, 
were adults thus sample sizes were 
antelope bitterbrush, and cheatgrass. 
insufficient to test age related responses and 
Grazing Effects on Deer Mice with Implications to Human Exposure to Sin Nombre Virus       33

Figure 2. The number of entrances/individual mice into simulated buildings located in control 
(non-grazed) versus grazed plots in Southwest Montana from 2005 through 2008.
In our study, the number of mice 
However, we detected activity to differ 
entering simulated buildings declined over 
between control and grazed plots for both 
the 9-day period of each experimental trial. 
males and female deer mice, as represented 
This reduction in numbers over time was 
by number of entrances into buildings per 
most evident in the control plots of the 
individual. Activity decreased over the 9-day 
study in the absence grazing, suggesting 
duration of experimental trials in the grazed 
that the longer mice had access to buildings, 
plots but in the control plots it remained 
the less the mice were “inclined” to enter 
fairly constant for females and increased 
them. Fewer mice may have entered over 
for males (Fig. 3). In grazed plots deer 
the duration of the study possibly because 
mice may possibly have been using a new 
after initial investigations, they found the 
resource, e.g., seed in horse manure, instead 
simulated buildings unsuitable or they may 
of entering buildings or the mice hesitated 
have left the area entirely.  
to move across open areas to access the 
34          Leary et al.

Figure 3. Response of female versus male deer mice to grazing expressed as the number 
of  entrances/individual mice into simulated buildings located in control (non-grazed) 
versus grazed plots in Southwest Montana from 2005 through 2008
Figure 4.  Effects of food in simulated buildings on the distribution of entrances/
individual deer mouse in non-grazed pastures in Southwest Montana from 2005 through 
2008. Animals responded to food and non-food buildings equally in grazed pastures.
Grazing Effects on Deer Mice with Implications to Human Exposure to Sin Nombre Virus       35

buildings. The increase in activity by males 
health measures to protect humans from 
in the control plots may have been due to 
SNV.  More research, perhaps with more 
the presence of food in one of the buildings. 
severe and expansive habitat modification 
In a previous study, presence of food 
than was created by a few days of grazing 
significantly increased activity represented 
and more complex buildings is necessary to 
by entrances into buildings (Kuenzi and 
clarify the influences of habitat modification 
Douglass 2009). However, the effect of food 
on deer mice entering buildings. However, 
only occurred in the control plots and not 
our preliminary results suggested that 
in grazed plots. Why male activity would 
grazing may reduce activity of male mice in 
increase and female activity did not is 
unclear.  Perhaps males could shift activity 
to the new food resource more easily than 
females that may have had litters in nests 
Special thanks are due to Hank, Maggie, 
some distance from the buildings.
Abby, Emily, and Jay Peterson and Susan 
  Previous research has shown that 
LaRue for unlimited access to their property 
adult males are the most likely to become 
as well as the use and transportation of their 
infected with SNV (Douglass et al. 2007) as 
horses. Financial support was provided by 
well as to disperse (Lonner et al. 2008). If 
the National Institute for Health (NIH) grant 
males are more likely to enter or are more 
# P20RR16455-05 from the INBRE – BRIN 
active in peridomestic systems upon sylvan 
program. Amy Skypala initially got this 
disturbance (grazing in this instance) than 
project off the ground and provided valuable 
females, grazing could increase chances 
information. Brent Lonner, Kevin Hughes, 
for human exposure to SNV.  However, 
Dean Waltee, Arlene Alvarado, Justine 
grazing did not increase the number of either 
Wilson, Krista Clark, Tessa Spear, Karoun 
males or females entering buildings but did 
Bagamian, Flavia Mazzini, Stephanie 
modify the behavioral response (number of 
Torelli, McKenna Leary, and Tom Horne 
entrances/individual). The males increased 
provided valuable assistance both in the field 
activity in buildings located in the control 
and out.  
plots during the duration of the trials but 
decreased activity into buildings located 
Literature cited
in the grazed plots. This may indicate that 
grazing in small areas around buildings may 
Armstrong, L. R., S. R Zaki, M. J. Goldoft, 
reduce human exposure to SNV in buildings. 
R. L. Todd, A. S. Khan, R. F. Khabbaz, 
The simulated buildings do not replicate 
T. G. Ksiazek, and C. J. Peters. 1995. 
all resources presented to deer mice in 
Hantavirus pulmonary syndrome 
normal houses, barns and sheds. Normal 
associated with entering or cleaning 
buildings are much more complex, larger 
rarely used, rodent-infested structures. 
and provide places for nesting, hiding and 
Journal of Infectious Diseases 172: 1166. 
caching food. The resources provided by 
Childs, J. E., T. G. Ksiazek, C. F. 
normal buildings would affect how deer 
Spiropoulou, J. W. Krebs, S. Morzunov, 
mice used the buildings once the outside 
G. O. Maupin, K. L. Gage, P. E. Rollin, 
environment was modified differently than 
J. Sarisky, and R. E. Enscore. 1994. 
our simulated buildings did. However, the 
Serologic and genetic identification of 
simulated buildings do provide insight into 
Peromyscus maniculatus as the primary 
the initial response (entering) of deer mice to 
rodent reservoir for a new hantavirus in 
habitat modification surrounding buildings.  
the southwestern United States. Journal 
Information on factors that draw 
of Infectious Diseases 169:1271-1280.
mice into buildings is necessary to better 
Clary, W. P. and D. E. Medin. 1992. 
understand the relationship between 
Vegetation, breeding bird, and small 
peridomestic and sylvan settings. This 
mammal biomass in two high-elevation 
understanding is necessary to design proper 
sagebrush riparian habitats. Pp. 100-110 
36          Leary et al.

in Clary W. P., E. D. McArthur, D. 
Medin, D. E. and W. P. Clary. 1989. Small 
Bedunah, and C. L. Wambolt, compilers., 
mammal populations in a grazed and 
Proceedings of the symposium on 
ungrazed riparian habitat in Nevada.  
ecology and management of riparian 
USDA Forest Service Research Paper 
shrub communities; Sun Valley, ID.
INT-413, Intermountain Research 
Damhoureyeh, S. A. and D. C. Hartnett. 
Station, Ogden, UT.
1997. Effects of bison and cattle on 
Morse, S. S. 1995. Factors in the emergence 
growth, reproduction, and abundances 
of infectious diseases. Emerging 
of five tall grass prairie forbs. American 
Infectious Diseases 1:7-15.
Journal of Botany 84:1719-1728.
Nichol, S. T., C. F. Spiropoulou, S. 
Douglass, R. J., C. H. Calisher, K. D. 
Morzunov, P. E. Rollin, T. G. Ksiazek, H. 
Wagoner, and J. N. Mills. 2007. Sin 
Feldmann, A. Sanchez, J. Childs, S. Zaki, 
Nombre virus infection of deer mice 
C. J. Peters. 1993. Genetic identification 
in Montana: Characteristics of newly 
of a hantavirus associated with an 
infected mice, incidence and temporal 
outbreak of acute respiratory illness. 
pattern of infection. Journal of Wildlife 
Science 262:914-917.
Diseases 43:12-22.
Oldemeyer, J. L. and L. R. Allen-Johnson. 
Douglass, R. J. and M. R. Frisina. 1993. 
1988. Cattle grazing and small mammals 
Mice and management on the Mount 
on the Sheldon National Wildlife Refuge, 
Haggin Wildlife Management Area. 
Nevada. Pp. 391-398 in Szaro R. C., 
Rangelands 15:8-15.
K. E. Severson , D. R. Patton, editors, 
Hayward, B., E. J. Heske, and C. W. Painter. 
Management of amphibians, reptiles, 
1997. Effects of livestock grazing on small 
and small mammals in North America. 
mammals at a desert cienage. Journal of 
General Technical Report RM-166, Fort 
Wildlife Management 61:123-129.
Collins, CO.
Kuenzi, A. J. and R. J. Douglass. 2009. An 
Waltee, D., B. N. Lonner, A. J. Kuenzi, and 
experimental test of factors attracting 
R. J. Douglass 2009.  Seasonal dispersal 
deer mice into buildings. Intermountain 
patterns of sylvan deer mice (Peromyscus 
Journal of Science 15:27-31.
maniculatus) within Montana Range  
Lands. Journal of Wildlife Diseases 
Kuenzi, A. J., M. M. Zumbrun, and K. 
Hughes. 2005. Ear tags versus passive 
integrated transponder (PIT) tags for 
 Woolhouse, M. E. J. and S. Gowtage-
effectively marking deer mice. Intermountain 
Sequeria. 2005. Host range and emerging 
Journal of Sciences 11:66-70.
and reemerging pathogens. Emerging 
Infectious Diseases 11:1842-1847. 
Kuenzi, A. J., R. J. Douglass, D. White, 
Jr, C. W. Bond, and J. N. Mills. 2001. 
Zar, J. H.  1984. Biostatistical analysis. 2nd 
Antibody to Sin Nombre virus in rodents 
edition. Prentice-Hall, Englewood Cliffs, 
associated with peridomestic habitats in 
NJ. 718pp.
west central Montana. American Journal 
of Tropical Medicine and Hygiene 
Lonner, B. N., R. J. Douglass, A. J. Kuenzi, 
Received 16 May 2011
and K. Hughes. 2008. Seroprevalence 
Accepted 21 December 2011
against Sin Nombre virus in resident and 
dispersing deer mice. Vector-Borne and 
Zoonotic Diseases 8:433-441.  
Grazing Effects on Deer Mice with Implications to Human Exposure to Sin Nombre Virus       37