ArticlePDF Available

Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in Sagebrush Ecosystem


Abstract and Figures

Conifer encroachment in sagebrush ecosystems reduces habitat heterogeneity, niche space, and resource availability, all of which negatively affect many wildlife populations. Sagebrush restoration is recommended as a management action to mitigate conifer encroachment and restore wildlife across millions of hectares in the Great Basin. Despite this recommendation, the effects of conifer encroachment and sagebrush restoration are unknown for most wildlife species. Small nonvolant mammal communities include keystone species, consumers and prey; facilitate energy flow and ecological function; and provide important ecological goods and services. We assessed causal relationships between conifer encroachment and sagebrush restoration (conifer removal and seeding native plants) on small mammal communities over 11 yr using a Before-After-Control–Impact design. Sagebrush habitat supported an additional small mammal species, twice the biomass, and nearly three times higher densities than conifer-encroached habitat. Sagebrush restoration increased shrub cover, decreased tree cover, and density but failed to increase native herbaceous plant density. Restoration caused a large increase in the non-native, invasive annual cheatgrass (Bromus tectorum L.). Counter to prediction, small mammal diversity did not increase in response to sagebrush restoration, but restoration maintained small mammal density in the face of ongoing conifer encroachment. Piñon mice (Peromyscus truei), woodland specialists with highest densities in conifer-encroached habitat, were negatively affected by sagebrush restoration. Increasing cheatgrass due to sagebrush restoration may not negatively impact small mammal diversity, provided cheatgrass density and cover do not progress to a monoculture and native vegetation is maintained. The consequences of conifer encroachment, a long-term, slow-acting impact, far outweigh the impacts of sagebrush restoration, a short-term, high-intensity impact, on small mammal diversity. Given the ecological importance of small mammals, maintenance of small mammal density is a desirable outcome for sagebrush restoration.
Content may be subject to copyright.
Effects of Sagebrush Restoration and Conifer Encroachment on Small
Mammal Diversity in Sagebrush Ecosystem
Bryan T. Hamilton
,Beverly L. Roeder
,Margaret A. Horner
Great Basin National Park, Science and Resource Management Division, Baker, NV 89311, USA
Brigham Young University, Department of Biology, Provo, UT 84602, USA.
abstractarticle info
Article history:
Received 24 October 2017
Received in revised form 10 July 2018
Accepted 10 August 2018
Available online xxxx
Key Words:
Bromus tectorum
Great Basin National Park
spatially explicit capture recapture (SECR)
Conifer encroachment in sagebrush ecosystems reduces habitatheterogeneity, niche space,and resource avail-
ability,all of which negatively affect manywildlife populations. Sagebrush restoration is recommended asa man-
agement action to mitigate conifer encroachment and restore wildlife across millions of hectares in the Great
Basin. Despite this recommendation, the effects of conifer encroachment and sagebrush restoration areunknown
for most wildlife species. Small nonvolant mammal communities include keystone species, consumers and prey;
facilitate energy ow and ecological function; and provide important ecological goods and services. We assessed
causal relationships between conifer encroachment and sagebrush restoration (conifer removal and seeding na-
tive plants) on small mammal communities over 11 yr using a Before-After-ControlImpact design. Sagebrush
habitat supported an additional small mammal species, twice the biomass, and nearly three times higher densi-
ties than conifer-encroached habitat. Sagebrush restoration increased shrub cover, decreased tree cover, and
density but failed to increase native herbaceous plant density. Restoration caused a large increase in the non-na-
tive, invasive annual cheatgrass (Bromus tectorum L.). Counter to prediction, small mammal diversity did not in-
crease in response to sagebrush restoration, but restoration maintained small mammal density in the face of
ongoing conifer encroachment. Piñon mice (Peromyscus truei), woodland specialists with highest densities in co-
nifer-encroached habitat, were negatively affected by sagebrush restoration. Increasing cheatgrass due to sage-
brush restoration may not negatively impact small mammal diversity, provided cheatgrass density and cover
do not progress to a monoculture and native vegetation is maintained. The consequences of conifer encroach-
ment, a long-term, slow-acting impact,far outweigh the impacts of sagebrush restoration, a short-term, high-in-
tensity impact, on small mammal diversity. Given the ecological importance of smallmammals, maintenance of
small mammal density is a desirable outcome for sagebrush restoration.
Publishedby Elsevier Inc. on behalfof The Society for Range Management. Thisis an open access article underthe
CC BY-NC-ND license (
Small, nonvolant mammals (hereafter small mammals) play critical
ecological roles in sagebrush ecosystems. Seed caching enhances germi-
nation of plants such as bitterbrush (Purshia tridentata (Pursh) DC.;
Hormay, 1943; Young and Clements, 2002); Mormon tea (Ephedra
viridis Coville; Everett et al., 1978; Hollander et al., 2010); and Indian
rice grass (Achnatherum hymenoides [Roem. & Schult.] Barkworth;
McAdoo et al., 1983). Burrowing by small mammals aerates soils
(Huntly and Inouye, 1988), cycles nutrients (Sirotnak and Huntly,
2000), and maintains early seral state plant communities (Kitchen and
Jorgensen, 1999). As the prey base for many predators, small mammals
are an important trophic link in food webs (Bekoff, 1977; Glaudas et al.,
2008). Small mammals also scatter hoard pine nuts, juniper berries, and
cheatgrass (Bromus tectorum L.) seed, resultingin the establishmentand
dispersal of conifers and invasive annual grasses (Chambers et al., 1999;
Young and Clements, 2009). Although plant germination is enhanced by
scatter hoarding, small mammal herbivory and larder hoarding can also
result in signicant mortality of seeds and newly established plants
(Clements and Young, 1996), both decreasing the establishment of de-
sirable native plants and increasing the prevalence of conifers and
cheatgrass. Habitat alteration can disrupt the ecological roles of small
mammals in sagebrush ecosystems.
In the past 130 yr, late-successional conifer woodlands have in-
creased 10-fold in the Great Basin (Miller and Tausch, 2001). Conifer
encroachmentdescribes a successional process of increasing conifer
cover and density in sagebrush ecosystems. Historically, conifer en-
croachment was regulated by periodic natural disturbances, most
Rangeland Ecology & Management xxx (xxxx) xxxxxx
This work wasfunded in part by the National Fire Programand Southern NevadaPub-
lic Lands Management Act Eastern Nevada Landscape Restoration Program.
Correspondence: Bryan T. Hamilton, 100 Great Basin National Park, Baker, NV 89311,
E-mail address: (B.T. Hamilton).
RAMA-00336; No of Pages 10
1550-7424/Published by ElsevierInc. on behalf of The Society forRange Management. Thisis an open access article underthe CC BY-NC-ND license(
Contents lists available at ScienceDirect
Rangeland Ecology & Management
journal homepage:
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
importantly high-intensity re (Miller et al., 2005; Tausch et al., 2009).
In recent decades, human-induced factors of re exclusion (Gruell et al.,
1994; Keane et al., 2002), increased atmospheric carbon dioxide
concentration, increased winter precipitation, warmer temperatures
(Rapp, 2004), and selective herbivory by livestock (Miller et al., 1994)
have interacted to increase the rate and scale of conifer encroachment
across the Great Basin.
In conifer dominated woodlands, the majority of plant biomass is se-
questered as unpalatable cellulose or lignin, which is unavailable to
most animals as food. Pine nuts and juniper berries are high in energy
and protein, but conifer mast is produced in erratic and unpredictable
resource pulses (White et al., 1999; Felicetti et al., 2003). Shrub, grass,
and forb production show less interannual variation than conifer mast,
produce more palatable seeds and forage, and support higher insect di-
versity than woodlands, providing a more reliable food source to wild-
life than conifer mast (Miller, 2008; McIver and Macke, 2014). Overall
conifer encroachment in sagebrush ecosystems reduces habitat hetero-
geneity, niche space, and resource availability, negatively affecting
many wildlife populations, such as sage grouse, pygmy rabbits, and
mule deer (Miller et al., 2005; Hanser and Knick, 2011; Baruch-Mordo
et al., 2013; Woods et al., 2013).
Conifer removal is the primary restoration tool in conifer-
encroached, sagebrush ecosystems. Great Basin coniferous woodlands
are dominated by two species: singleleaf piñon pine (Pinus monophylla
Torr. & Frém.) and Utah juniper (Juniperus osteosperma [Torr.] Little).
Methods of conifer removal include chaining in high-density conifer
stands, lop and scatter of low-density conifers, mastication using ma-
chinery, prescribed re, and hand cutting with chainsaws (Bombaci
and Pejchar, 2016). To increase shrubs and herbaceous plants, conifer
removal projects often incorporate seeding of native shrubs and herba-
ceous plants into management actions (Weltz et al., 2014).
Despite their critical role in ecosystem function, the effects of conifer
encroachment on small mammals have received minimal attention rel-
ative to other wildlife species. In a comparison of recent and historic
small mammal communities, Rickart et al. (2008) attributed shifts in
species composition to increasing conifer woodlands. Changes in
species composition included a decrease in sagebrush specialists,
Great Basin pocket mice (Perognathus mollpilosus) and least chipmunks
(Tamius minimus), and an increase in woodland specialists, piñon mice
(Peromyscus truei) and cliff chipmunks (Tamias dorsalis). Sagebrush res-
torationand conifer removal were suggested as a means to restore sage-
brush-dependent small mammal diversity in conifer-encroached
ecosystems (Rickart et al., 2008). Similarly, Rodhouse et al. (2010)
noted the potential for sagebrush restoration and conifer removal to
negatively impact woodland-associated small mammal species, such
as piñon mice and cliff chipmunks.
We evaluated the relationships between conifer encroachment,
sagebrushrestoration, and small mammal diversity in a sagebrush eco-
system over 11 yr. We hypothesized that conifer encroachment has
negatively impacted small mammal diversity and that sagebrush resto-
ration could mitigate this loss of diversity. We made four predictions
about the effects of conifer encroachment and sagebrush restoration
on small mammal diversity: 1) Small mammal diversity is lower in co-
nifer encroached habitat than in sagebrush habitat; 2) Native shrub
cover and herbaceous plant density will increase in response to sage-
brush restoration; 3) Sagebrush restoration will increase small mammal
density, richness, biomass, and evenness; and 4) Small mammal com-
munity responses to sagebrush restoration will be species specic. We
expected sagebrush specialists to increase in response to sagebrush res-
toration and woodland specialists to decrease.
Study Site
The study was conducted in Great Basin National Park, South Snake
Range, White Pine County, Nevada (38.98°N, 114.30°W; Fig. 1).Eleva-
tions in the South Snake Range vary from 1 621 m in the town of Baker
to 3 982 m at the summit of Wheeler Peak. The climate is cool and arid
and varies with elevation. The elevation of the study site is 2 832 m, an-
nual precipitation 33 cm, and the mean annual temperature is 9°C
Figure 1. Studysite map, showingsmall mammal grids, vegetation transects, habitat, and areasand year of sagebrush restoration treatments.Inset map shows thestudy site in the context
of the larger Great Basin desert (gray shading).
2B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
(Western Regional Climate Center, unpublished data for Lehman
Caves). The frost-free period ranges from 60 to 90 d.
Sagebrush Restoration
Treatment goals were to reduce conifer cover from pre-treatment
levels of 20 30% to b10% cover. Quantitative outcomes for shrubs
and herbaceous vegetation were not dened, but an overall project
goal was to increase native shrubs and native herbaceous vegetation
(graminoids and forbs), without increasing cheatgrass, a non-native in-
vasive annual grass. Singleleaf piñon and Utah juniper trees were cut
with chainsaws on a total area of 32 ha over the duration of the project
(see Fig. 1). Staggered at different sites over several years, conifers were
removed only once at a given location (see Fig. 1; 20042007). Conifer
slash was disposed of through a combination of pile burning, chipping,
and fuel wood disposal. To promote the restoration of shrub and
herbaceous vegetation, a native seed mix consisting of mountain big
sagebrush, Sandberg bluegrass (Poa secunda J. Presl), bluebunch wheat-
grass, basin wildrye (Leymus cinereus [Scribn. & Merr.] Á. Löve),
squirreltail (Elymus elymoides [Raf.] Swezey), Lewis ax (Linum lewisii
Pursh), and Indian rice grass was broadcast on the conifer removal
units with belly spreaders at 9 kg of pure live seed per ha.
Study Design
Our study was set up as a Before-After-ControlImpact (BACI) de-
sign, a common, quasiexperimental study design widely used to com-
pare environmental conditions before and after human disturbance
(de Lucas et al., 2005). The BACI design predicts different patterns of
change for impact sites relative to control sites following a disturbance.
Hereafter, we use the terms impactand treatmentinterchangeably
and use referencesynonymously with control.A statistically signif-
icant interaction between time relative to treatment and treatment
(Underwood, 1992) strongly infers a causal relationship between im-
pact and effect (Block et al., 2001). Although reference sites should be
closely matched to the impact sites, the absolute similarity between im-
pact and the reference sites is less important than the trajectory of the
sites relative to each other, with respect to treatments (Underwood,
Before sagebrush restoration, habitats were stratied into two habi-
tat types: sagebrush or conifer encroached. Stratication was based on
visual assessment and pre-treatment global positioning system (GPS)
mapping of habitat patches. Twenty-four vegetation transects were ran-
domly located: 12 in sagebrush habitat and 12 in conifer-encroached
habitat, conditional on a minimum separation of 50 m between tran-
sects (see Fig. 1). Five small mammal trapping grids were randomly
chosen from conifer-encroached habitat. Sagebrush habitat was limited,
so the ve sagebrush grids were located remotely using a global infor-
mation system to maximize grid t (see Fig. 1). Although vegetation
transects were not colocated exactly with trapping grids, seven grids
were intersected by vegetation transects. The mean distance between
vegetation transects and small mammal grids was 6.4 m (range =
036 m). Giventhis close proximity, we consider vegetation transects
strongly linked to and representative of small mammal grids (see Fig. 1).
Conifer encroachment occurs on a continuum of increasing tree
cover and density described with three woodland phases (Tausch et
al., 2009). Conifers occur at low cover and density in phase I woodlands,
with shrubs and herbaceous vegetation dominating the understory. In
contrast, phase III woodlands have high conifer cover and density,
with little shrub or understory vegetation. Phase II woodlands are
codominated by conifers and shrubs and provide biological and struc-
tural attributes of both woodland and sagebrush habitats (Tausch et
al., 2009). We considered post hoc framing of our results into woodland
phases but ultimately used our a priori habitat stratication, as the
woodland phase paradigm did not exist at the onset of our study. Retro-
spectively, our conifer-encroached habitat was similar to phase III
woodlands and sagebrush habitat similar to phase II woodlands. We
recommend that inferences of our results be limited to sagebrush resto-
ration in phase II woodlands, using similar restoration methods (i.e.,
cutting conifers with chainsaws and seeding with native plant species).
Four small mammal grids and nine vegetation plots served as impact
sites. All impact sites were in sagebrush habitat, where the impact
consisted of conifer removal and seeding native vegetation (see Sage-
brush Restorationearlier). Six small mammal grids and 15 vegetation
transects served as untreated, reference sites. Of these reference sites,
1 small mammal grid and 3 vegetation transects were in sagebrush hab-
itat and 5 grids and 12 transects were in conifer-encroached habitat.
Habitat was considered a xed effect, so treated sagebrush sites were
still considered sagebrush habitat following treatments.
Time was considered in three ways. Yearwas a continuous vari-
able of calendar year and was incorporated into models asa random ef-
fect to account for interannual variation. Pre-treatment and post-
treatmentwere categorical variables (e.g., before and after treatments).
Time relative to treatment was the number of years from treatment. As
the treatments were staggered over different years, time relative to
treatment occurred in different calendar years for the different sites.
Reference sites were assigned time relative to treatment on the basis
of their proximity to the treated sites, with control sites paired with
their closest treated sites.
Vegetation Sampling
Each vegetation transect was sampled before and after sagebrush
restoration treatments in June of 2004, 2010, and 2014. Pre-treatment
cover (2004) was assessed using a line intercept method (Bonham,
1989). A major weakness of the line intercept method is that annual
grass cover is poorly sampled. Cheatgrass is the only ecologically impor-
tant annual grass on our site. As a non-native invasive species, cheat-
grass was critical to monitor. To improve sampling of cheatgrass
cover, we adjusted our methodology from a line intercept to a line
point intercept in 2010 and 2014 (Herrick et al., 2005). To justify this
change in sampling, we ran seven comparative transects and found
tree and shrub cover strongly correlated (r= 0.993 and 0.944, respec-
tively) between methods. Herbaceous cover, which included cheat-
grass, all other grasses, graminoids, and forbs, was weakly correlated
between methods (r= 0.675). Given the strong correlations for tree
and shrub cover, we used cover to assess treatment effects on shrubs,
density on herbaceous vegetation, and both cover and density to ad-
dress treatment effects on trees. Herbaceous plant density was mea-
sured in four quadrats per transect (35 cm × 35 cm). Quadrats,
combined by transect for analysis, were oriented on the east side of
transects every 10 m, sampling a total area of 0.5 m
per transect. Tree
density was sampled on larger plots (2 m × 100 m), one plot per tran-
sect (Herrick et al., 2005). All treeswith stems wholly orpartially within
the plot were tallied.
Small Mammal Sampling
Small mammals were sampled with Sherman live traps arranged in
agridconguration. Grids were sampled each July from 2004 to 2014.
Each grid consisted of 49 Sherman live traps (SFAL; 5 cm × 6 cm × 23
cm or LFA; 8 cm× 9 cm × 23 cm) separated by 15 m, in a 7 × 7 pattern,
sampling an area of approximately 1 ha. Individual trap locations were
relocated with GPS (± 1 m). Traps were locked open and prebaited
for 34 d before sampling, then rebaited and set. Millet and sunower
seeds were used as bait. Traps were set each evening between 17:00
and 20:00, checked each morning between 05:00 and 10:00, and shut
during the day. Trapping sessions consisted of 4 consecutive nights.
Trapping was interrupted on two occasions for 1 night but resumed
the following day. Captured small mammals were ear tagged, identied
to species, visually assessed for sex, weighed, and released. Recaptured
individuals were weighed and assessed for ear tag number, species, and
3B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
sex and then released. After accounting for sprung traps, trap effort
consisted of 20 920 trap nights. Ten species and 2 066 individuals
were captured over the 11 yr of sampling (Table 1). Small mammal den-
sities uctuated widely across years in both impact and reference grids
(Fig. 2). Fluctuations were qualitatively synchronous across impact and
reference sites (see Fig. 2). Small mammal sampling was conducted ac-
cording to the guidelines of Brigham Young Universitys Institutional
Animal Care and Use Committee, project code 07-0301, scientic re-
search permits from Great Basin National Park (GRBA-2007-SCI-0002)
and Nevada Department of Wildlife (S35631), and the American Society
of Mammalogists (Sikes et al., 2011).
Data Analysis (Vegetation)
To quantify pre-treatment differences in mean tree, shrub, and her-
baceous cover between sagebrush and conifer-encroached transects,
we used Studentst-tests with unequal variances and f-tests to examine
differences in variability.We examined the response of vegetation toco-
nifer removal, using a BACI design (as described earlier) implemented
in a generalized linear mixed-model framework. For vegetation analy-
ses, time was binned into two categories: pre-treatment or post-treat-
ment. Tree and shrub cover and herbaceous and tree density were
treated as xed effects in separate models. Cover was modeled with a
negative binomial distribution and density with a Poisson distribution,
both using log links. Transect was incorporatedinto models as a random
effect. Chi-squared tests were used to test for differences in observed
versus expected herbaceous density and percent composition of annual
grasses between treated and untreated sites, pre-treatment and post
treatment. Expected density, the statistical null hypothesis, was dened
as equal distribution across time and treatments.
Data Analysis (Small Mammals)
Species richness was the number of species per grid by year. Our
evenness metric was the inverse of the Simpson index (SI) calculated
as: SI = 1/[(n
1)/N(N 1)]; where n
= the number of indi-
viduals of the i
species; and N= the total number of individuals
(Magurran, 2004). As SI decreases, community evenness also decreases.
Total biomass was the sum of the mean weights of all individuals cap-
tured per grid by year.
Total density and density of individual species (deer mice, piñon
mice, cliff chipmunks, voles, Great Basin pocket mice, and western har-
vest mice) were calculated using spatially explicit capture recapture
(SECR) models. Spatially explicit capture recapture models relate the
spatial relationships of traps (detectors) and the movement of animals
between traps through a combination of a state model and anobserva-
tion model (Efford et al., 2009). The state model describes the distribu-
tion of the animal home ranges on the landscape. The observation
model (spatial detection model) relates the probability of detecting an
individual at a particular detector to the distance of the detector to a
central point in each animals home range. The distribution of home
range centers is treated as a homogenous Poisson point process. Buffer
width was set at 150 m. The detection function describes the decline
in detection probability with distance from the home range center
using a half-normal detection function. Detector types were single, as
traps weregenerally capableof catching only one animal. Full likelihood
was used to t all models. Detection was modeled as a function of dis-
tance between the trap and the individualslatentactivitycenter.
Eight candidate SECR models were analyzed to calculate density.
Each year was treated as a session, and each grid was analyzed sepa-
rately by species. Density was always t as a function of session
(year). Detection (g0) and movement (σ) were modeled as 1) constant
detection probability across occasions and detectors; 2) learned re-
sponse affecting detection; 3) trap response to time; 4) trap response
with a time trend; 5) trap response model-transient; 6) site learned re-
sponse; 7) site transient response; and 8) heterogeneity model, nite
mixture model, with two latent classes.All SECR models were computed
in the R package secr (Efford, 2018). Density estimates were highly cor-
related with raw captures (r= 0.90).
The heterogeneity model (8) was highly favored by Akaike Informa-
tion Criterion, corrected for small sample size (AIC
) when enough ani-
mals were captured to support the model structure (ΔAIC
(Burnham and Anderson, 2002). Otherwise, the null model was pre-
ferred. When there was competing weight of evidence, we used
model averaging to calculate density (individuals ha
). When there
was a clear top model (ΔAIC
b2), that model was used to estimate
density. Year-specic density estimates by sampling grid were incorpo-
rated as response variables into the mixed models.
We assessed the effects of habitat and sagebrush restoration treat-
ments on small mammal communities using the BACI design as de-
scribed earlier. Generalized linear mixed models were used to assess
the effects of sagebrush restoration and habitat on small mammal diver-
sity (species richness, total biomass, evenness, total density, and indi-
vidual species density). Habitat (sagebrush or conifer-encroached)
and sagebrush restoration effects (interaction between treatment and
time) were the primary independent variables. Time was dened in
years from treatment or binned into pre-treatment or post treatment.
For most dependent variables we compared ve models: 1) treatment
× time (pre, post) + habitat; 2) treatment × time (year relative to treat-
ment) + habitat; 3) habitat only; 4) treatment × time (pre, post); and
5) null model. Site and year were included as additive random effects
in models. Richness and evenness were modeled using a Gaussian dis-
tribution and identity link function. A negative binomial distribution
was used to model biomass and a Poisson distribution to model density,
both using log link functions (Table 2).
Zero-inated models were used to analyze the effect of sagebrush
restoration and habitat on piñon mouse, cliff chipmunk, vole, pocket
Table 1
Small mammal capturesby species forsagebrush and conifer-encroached habitats in Great
Basin National Park,White Pine County Nevada. Smallmammals were sampledfrom 2004
to 2014 for a total of 20 920 trap nights in 10, one-ha grids.
Common name Species Sagebrush Conifer
Deer mouse Peromyscus maniculatus 1 228 369
Western harvest mouse Reithrodontomys megalotis 144 0
Cliff chipmunk Tamias dorsalis 75 80
Piñon mouse Peromyscus truei 22 103
Montane vole Microtus montanus 80
Long-tailed vole Microtus longicaudus 14 1
Great Basin pocket mouse Perognathus mollipilosus 13 1
Sagebrush vole Lemmiscus curtatus 50
Uinta chipmunk Tamias umbrinus 11
Least chipmunk Tamias minimus 10
Figure2. Annual small mammaldensities (x± SE) in Julyfor reference andimpact grids in
Great Basin National Park, White Pine County, Nevada. Small mammal densities were
estimatedfrom spatially explicit capture recapture models.
4B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
mouse, and western harvest mouse densities. The proportion of grids
with no captures for these species ranged from 35% to 88%. Long-tailed
and montane vole densities were combined for analysis. Models for
piñon mice, cliff chipmunks, voles, pocket mice, and harvest mice
could support time only as pre-treatment, post-treatment in the treat-
ment time interaction. To allow model convergence, random effects
for piñon mice only included year and random effects were excluded
from density models for cliff chipmunks, voles, pocket mice, and harvest
mice. Total and deer mouse densities did not require zero-inated
To validate model t, we plotted residuals versus ttedvalues, resid-
uals versus covariates, and examined histograms of residuals for nor-
mality. We also compared models with treatment and habitat effects
to null models using Akaike Information Criterion (AIC
model selection (Burnham and Anderson, 2002). Models differing by b
units were considered equivalent. Alpha (α) was set at 0.05. Ef-
fect sizes are given following statistical results, calculated from
highest-ranked model coefcients, model averaging, or mean differ-
ences between groups. Analyses were done with Program R (RCore
Team, 2016), generalized linear mixed models in the R package
glmmADMB (Skaug et al., 2014), and model averaging with the R pack-
age MuMIn(Barton, 2018).
Vegetation (pre-treatment)
Before sagebrush restoration, transects classied as sagebrush habi-
tat had higher herbaceous (t = 4.20, d.f. = 13.1, P = 0.001; 5.3%) and
shrub cover (t = 5.75, d.f. = 11.25, P = 0.0001; 13.4%) than conifer-
encroached transects (Fig. 3). Total plant cover did not differ between
habitats (t = 1.03, d.f. = 15.3, P = 0.32). Conifer-encroached habitat
was higher in tree cover (t = 2.69, d.f. = 16.89, P = 0.015; 13%) and
had lower variance in herbaceous (ratio of variances = 0.09), shrub
(ratio of variances = 0.01), and tree cover (ratio of variances = 0.29)
than sagebrush habitat (P b0.05 for all tests). Tree density was higher
in conifer-encroached relative to sagebrush habitat by a factor of 2 (t =
4.80, d.f. = 18, P b0.0001; 7 975 vs. 4 071 trees ha
). All conifer-
encroached transects had b2.6% shrub cover, and all sagebrush transects
had N2.6% shrub cover. Singleleaf pinyon was the dominant tree, ac-
counting for 84% of tree cover and 93% of tree density. Utah juniper
was less abundant and comprised 15% of tree cover and 5% of tree den-
sity. Curleaf mountain mahogany (Cercocarpus ledifolius Nutt.), aspen
(Populus tremuloides Michx.), and chokecherry (Prunus virginiana L.) oc-
curred on sagebrush transects but were rarely sampled.
Vegetation (post treatment)
Sagebrush restoration reduced tree cover from 28.7% to 2.2%, while
tree cover was unchanged on untreated plots (z = 11.81, Pb0.0001).
Table 2
Model comparisons of sagebrush restoration and habitat effects on smallmammal diver-
sity in Great Basin National Park.
Model Parameters AICc Delta
Richness Habitat 6 284.77 0.00 0.91
Treatment × time + habitat 9 290.76 5.99 0.05
Treatment × time 8 291.83 7.07 0.03
Null 5 292.63 7.87 0.02
Treatment × yr + habitat 25 312.52 27.76 0.00
Habitat 6 1419.76 0.00 0.93
Treatment × time + habitat 9 1424.90 5.13 0.07
Treatment × time 8 1433.56 13.80 0.00
Null 5 1434.50 14.74 0.00
Treatment × yr + habitat 25 1458.13 38.36 0.00
Evenness Null 5 221.73 0.00 0.55
Habitat 6 222.49 0.77 0.37
Treatment × time 8 226.13 4.40 0.06
Treatment × time + habitat 9 228.51 6.78 0.02
Treatment × yr + habitat 25 264.19 42.46 0.00
Treatment × time + habitat 8 689.21 0.00 0.39
Habitat 5 689.51 0.30 0.33
Treatment × time 7 690.07 0.86 0.25
Null 4 694.60 5.39 0.03
Treatment × yr + habitat 24 700.01 10.80 0.00
Richness indicates number of species/grid/yr; Evenness, inverse of Simpson index (SI),
where lower values indicate lower evenness; Total biomass, sum of mean weights for all
individuals captured/ grid/yr; Total density, sum of estimated densities for all species/yr,
estimated through spatially explicit capture recapture models. Time (pre_post), time
binned into pre-treatment or post-treatment periods; time (yrs elapsed), time binned
into yrs elapsed post treatment; habitat, habitat type, either sagebrush or conifer
Figure 3. Vegetation structure in sagebrush and conifer-encroached habitats in Great Basin National Park, White Pine County, Nevada. Herbaceous vegetation included cheatgrass,
graminoids, and forbs.Boxes show median and interquartile ranges, whiskers extend 1.5 times theinterquartile ranges, and dots show data fallingoutside the 1.5 interquartile ranges.
5B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
Tree density was reduced from 4 194 to 1 805 trees ha
on treated
plots while tree density increased from 7 120 to 7 374 trees ha
on un-
treated plots (z =, Pb0.0001). Sagebrush restoration increased shrub
cover from 13.1% to 18.6% on treated plots while shrub cover decreased
slightly on untreated plots from 4.2% to 3.7% (z = 2.50, P= 0.0123).
Herbaceous plant density increased by 146% on treated plots and was
unchanged on untreated plots (z = 4.94, Pb0.0001). The increase in
herbaceous density was driven by cheatgrass, which increased fourfold
on treated plots (Fig. 4). As a proportion of herbaceous density, cheat-
grass density nearly doubled, increasing from 42% to 81% on treated
plots while decreasing from 66% to 42% on untreated plots. When cheat-
grass was removed from the model, the increase in herbaceous density
due to conifer removal was not statistically signicant (z = 1.32, P=
0.20). Cheatgrass also increased in percent composition on treated
plots. Following sagebrush restoration treatments, cheatgrass density
was higher than expected on treated plots, while noncheatgrass herba-
ceous density was less than expected (see Fig. 4;χ
Total cheatgrass cover on post-treated transects ranged from 7% to 59%
(x= 33% ± 15%).
Small Mammal Diversity
Sagebrush restoration did not affect total biomass (yr, z b1.06, PN
0.36; pre-, post-, z = 0.81, P= 0.42); richness (yr, z b1.06, PN0.287;
pre-, post-, z = 0.49, P= 0.62); or evenness (yr, z b1.13, PN0.260;
pre-, post, z = 0.40, P= 0.69). When time was binned into years relative
to treatment, the treatment effect was signicant for total density at 8 yr
post treatment (z = 2.25, PN0.024). When time was binned into pre-
treatment or post-treatment, sagebrush restoration maintained density
Figure 4.Pe rcent cover of trees (A), percent coverof shrubs (B), densityof native herbaceousplants (C), and densityof cheatgrass (D)before and after sagebrush restoration at controland
treatment sites in the Great Basin National Park. Values are means and standard errors.
Figure 5. Total density of small mammals for impact and reference grids before and after
sagebrush restoration.Restoration maintained density on treated grids while density fell
on untreated grids (P=0.0097).
6B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
on treated grids, while density dropped on untreated grids (z = 2.59,
P= 0.0097; Fig. 5). Model comparisons supported habitat-only
models for richness and biomass, indicating that treatment effects
can be discarded in favor of the simpler habitat-only models (see
Table 2). Habitat had large effects on richness, biomass, and density
(z N3.27, Pb0.001; Table 3). Sagebrush habitat supported an addi-
tional species, 2.3 times more biomass, and 2.7 times the number
of individuals than conifer-encroached grids (see Table 3). Habitats
did not differ in evenness (z = 1.26, P= 0.21).
Deer mice were the most abundant species in both habitats, making
up 77% of captures (see Table 1). The effects of sagebrush restoration on
deer mice mirrored total density, where restoration treatments main-
tained density on treated grids and density fell on untreated grids (yr
1 post treatment, z = 2.26, Pb0.0238; pre-treatment and post-treat-
ment, z = 1.94, P= 0.0519). Deer mice were 3.8 times more abundant
in sagebrush than conifer-encroached habitat (Pb0.0001). Vole densi-
ties (Microtus sp.) were unaffected by conifer removal (z = 0.50, P=
0.612) and were higher in sagebrush habitat by a factor of 22 (P=
0.0042). Great Basin pocket mouse density was not affected by conifer
removal (z = 1.78, P= 0.076). Pocket mouse density was higher on
sagebrush than conifer-encroached grids by a factor of 16 (z =
2.42, P= 0.0155). Western harvest mice occurred only in sagebrush
increased in density in sagebrush habitat from 2004 to 2014 (z =
4.63, Pb0.001), and there was no effect of sagebrush restoration
(z = 0.00, P=1.00).Wecouldnottestforhabitatortreatmentef-
fects for sagebrush vole density, but we note that the only observa-
conifer removal. Piñon mouse density was signicantly reduced by
sagebrush restoration treatments while density increased on
untreated sites (pre-treatment and post-treatment; z = 2.91, P=
0.0036; Fig 6). Piñon mouse density was 5.9 times higher on coni-
fer-encroached than sagebrush habitat (z = 4.73, P= 0.00531).
There were no treatment effects (z = 1.32, P= 0.350) on cliff chip-
munk density. Cliff chipmunks were more abundant (1.6 times) in
sagebrush than conifer-encroached habitat (z = 2.38, P= 0.02621).
Sagebrush restoration is recommended on millions of hectares
across the Great Basin (Wisdom et al., 2002). As the scale and rate of
these projects increase, site-specic understanding of sagebrush resto-
ration effects on wildlife communities is a major research need (Knick
et al., 2014; Bombaci and Pejchar, 2016). Small mammals are excellent
models for assessing the effects of conifer encroachment and restora-
tion. With their small home ranges, small mammals are closely tied to
local changes in resource availability (Stephens et al., 2017). Addition-
ally, small mammal communities include keystone species; habitat spe-
cialistsand generalists; a diverse guild of feeding ecologies; and are both
consumers and prey. Thus, small mammal community response is a
window into ecosystem function, the understanding of which is a fun-
damental goal of land management and restoration. This study is the
rst to assess the relationships between conifer encroachment and
sagebrushrestoration (conifer removal and seeding) on small mammal
communities using a BACI design to assign causal relationships and ran-
dom effects to increase inferential scope. In addition, the 11 yr of mon-
itoring effort is substantially longer than most BACI studies.
Conifer encroachment into sagebrush habitat has dramatically re-
duced small mammal abundance and biomass. Sagebrush restoration
(conifer removal and seeding of native plants) increased native shrub
cover and invasive, non-native cheatgrass density but did not increase
native herbaceous plant densities. Sagebrush restoration reduced the
density of the woodland specialist piñon mouse. Overall, sagebrush res-
toration treatments had few effects on small mammal diversity but res-
toration effectively maintained small mammal densities in the face of
conifer encroachment.
Large-scale ecological experiments, such as this study, are difcult to
implement but are critical to address management and conservation
questions (Soanes et al., 2018). We used the strongest possible methods
and study design, given the constraints of management and habitat on
the study site (Soanes et al., 2018). However, it is important to acknowl-
edge the limitations of our study. Due to limited sagebrush habitat, our
study did not capture the entire successional range of sagebrush ecosys-
tems (e.g., phase I, II, and III woodlands). We recommend that infer-
ences of our results be limited to sagebrush restoration in phase II
woodlands, using similar restoration methods (i.e., cutting conifers
with chainsaws and seeding with native plant species).
Conifer Encroachment and Small Mammal Diversity
We observed large differences in small mammal richness, biomass,
and density between conifer-encroached and sagebrush habitats (see
Table 3) that support our prediction that conifer encroachment has re-
duced small mammal diversity. Comparison of sagebrush and conifer-
encroached habitats can be viewed as a space for time substitution. Sev-
eral lines of evidence support the hypothesis that our study site was his-
torically more open and sagebrushdominated, and that conifer density
and cover have increased over the last century (sensu Tausch et al.,
2009). Soils on our study site are in the Badena series, a mollisol with
glacial outwash parent material (USDA Natural Resources Conservation
Service, 2009). Characterized by a mollic epipedon, mollisols develop
in the absence of conifers, primarily from organic matter derived from
grasses and shrubs (USDA Natural Resources Conservation Service,
2009). In addition, historic photos of the study site document lower
tree densities and higher shrub cover than currently occur (Appendix
1). We also regularly observed shrub skeletons under conifers, further
Table 3
Metricsof small mammal diversity for sagebrush andconifer-encroached habitats in Great
Basin National Park. Sites were sampled annually from 2004 to 2014.
Sagebrush Encroached
Richness* 2.5 ± 9.0 1.7 ± 12.4
Evenness 1.52 ± 0.65 1.41 ± 0.70
Biomass (g)* 447.8 ± 347.0 187.6 ± 213.1
Density (ind./ha)* 27.5 ± 23.3 10.1 ± 12.9
Asterisk (*) indicates signicant differences (Pb0.01).
Figure 6. Total density of piñon mice for impact and reference grids before and after
sagebrush restoration. Restoration decreased density on treated grids while density
increased on untreated grids (P= 0.0036).
7B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
indication of recent conifer encroachment (Appendix 2; Austin, 1999;
Miller et al., 2008; Tausch et al., 2009). Assuming an increase in tree
density and cover over the past century, small mammal diversity has
been lost from formerly open sagebrush habitat as a result of conifer
Similar negative relationships between conifer encroachment and
small mammal diversity were also found in northern Nevadasagebrush
ecosystems. Coincident with regional woodland expansion, small mam-
mal communities declined by 50% in abundance, biomass, and energy
use between 1920 and 2008 (Rowe et al., 2011). We found threefold
lower density and twofold lower small mammal biomass in conifer-
encroached habitat relative to sagebrush habitat, results consistent
with the Rowe et al. (2011) historic comparison.
Although these arguments are correlative, direct observation and es-
tablishment of a true causal relationship between conifer encroachment
and small mammal diversity is unlikely. More than 100 yr are required
for the development of woodlands similar to the conifer-encroached
habitat we studied (Tausch et al., 2009). Given limitations of funding
for long-term monitoring, correlative relationships, space for time sub-
stitutions, and historic comparisons provide the strongest available ev-
idence of negative effects of conifer encroachment on small mammal
If conifer encroachment causes a loss of small mammal diversity,
what mechanisms drive the process? Small mammal abundance and
biomass reect resource availability (Rowe et al., 2011). Conifer en-
croachment has resulted in the severe reduction of native understory
shrubs and herbaceous vegetation, structural complexity and variabil-
ity, plant productivity, habitat heterogeneity and an overall reduction
of resource and niche space availability to the small mammal commu-
nity (see Fig. 3).
Sagebrush Restoration and Vegetation
Sagebrush restoration increased shrub cover and decreased tree
cover and density but failed to increase native herbaceous plant density.
Restoration treatments also caused an increase in the non-native, inva-
sive, annual grass cheatgrass (see Fig. 4). Despite the increase in cheat-
grass associated with sagebrush restoration, there were no negative
effects on small mammaldiversity in our study. Some studies have cor-
related reduced small mammal diversity with high cheatgrass cover
(Ostoja and Schupp, 2009; Freeman et al., 2014). Other work has
shown that species richness was not affected by cheatgrass cover, but
species responses were related to their functional traits (Ceradini and
Chalfoun, 2017). Cover values of cheatgrass on our treated sites were
modest (33%) compared with studies linking reduced small mammal
diversity and cheatgrass (47100% cover; Freeman et al., 2014; mono-
culture and 90% standing biomass; Ostoja and Schupp, 2009). In addi-
tion, these studies on small mammal diversity and cheatgrass were
conducted at lower elevations than our study, below the zone of conifer
encroachment, where Heteromyids formed a larger component of the
small mammal community (Ostoja and Schupp, 2009; Freeman et al.,
2014; Ceradini and Chalfoun, 2017). Native shrubs and native herba-
ceous plants were also maintained on our treated sites. Thus, an in-
crease in cheatgrass resulting from sagebrush restoration may not
negatively impact small mammal diversity, provided cheatgrass density
and cover do not progress to an annual grass monoculture.
Increases in cheatgrass often occur following sagebrush restora-
tion projects (Bates et al., 2005;Baughman et al., 2010), and seeding
alone may not restore native perennial grasses and forbs (Baughman
et al., 2010). To avoid cheatgrass monocultures, sagebrush restora-
tion projects should target resilient and resistant sites (Pellant et
al., 2004; Weltz et al., 2014), anticipate post-treatment increases
in annual grasses, and incorporate several years of seeding and
herbicide treatments following conifer removal into restoration
Sagebrush Restoration and Small Mammal Communities
Contrary to prediction, small mammal diversity did not increase in re-
sponse to sagebrush restoration. With the exception of total density,
treatments did not affect diversity. However, sagebrush restoration main-
tained small mammal densities in the face of conifer encroachment. In
contrast to the short-term, high-intensity, pulseimpact of restoration,
the slow conversion of sagebrush habitat to conifer-encroached habitat
is a long-term, slow-acting pressimpact with a negative and, presum-
ably, slow effect on small mammal communities (see Underwood, 1994
for denitions of press and pulse impacts). Restoration resets succession
and reduces tree cover, increasing shrub cover and herbaceous plants. In
untreated sites, shrub and herbaceous cover continued to be lost to coni-
fer encroachment, as tree cover increased. Restoration delayed conver-
sion of sagebrush habitat to conifer-encroached habitat, effectively
maintaining small mammal densities.
Piñon miceshowed a strong and negative response to sagebrush res-
toration.We anticipated this response, as piñon mice are truewoodland
obligates (Hoffmeister, 1981; Rodhouse et al., 2010). Piñon mice were
also the only species more abundant in conifer-encroached habitat
than sagebrush (by a factor of 7). These habitat preferences are consis-
tent with historically expanding populations of piñon mice in response
to woodland expansion (Rickart et al., 2008).
Counter to expectation, cliff chipmunks, also associatedwith conifer
woodlands (Rodhouse et al., 2010), were not affected by conifer re-
moval and were slightly more abundant in sagebrush habitat. We ini-
tially speculated that the larger cliff chipmunks had correspondingly
larger home ranges than piñon mice and dispersed into treated sites
from the conifer-encroached matrix. However, the two species have
similar home range sizes of approximately 1 ha (Hoffmeister, 1981;
Hart, 1992; this study). Our methods restricted captures of diurnal spe-
cies, suchas cliff chipmunks, which may have inuenced our results. As-
sessment of treatment effects on diurnal species will require daytime
Due to low capture rates, large annual uctuations in density, and
unbalanced occurrence across habitats and treatments, species-specic
effects of sagebrush restoration and habitat were difcult to model for
most species. This is a common theme in conservation biology; species
of management concern are generally uncommon, habitat specialists
and seldom captured in sufcient numbers to make strong, statistically
valid inferences with adequate power to detect treatment effects. While
we found no treatment effects for voles, pocket mice, or harvest mice,
we noted that these species were much more abundant in sagebrush
than conifer-encroached habitat. Harvest mice and sagebrush voles in-
clude annual grasses in their range of habitat preferences, and our
study found those species only on treated sites post sagebrush restora-
tion. Future studies should test the hypothesis that some small mammal
species may increase in density or occupancy as a result of sagebrush
restoration and increased annual grasses. Additional work should
focus on functional and aggregate diversity and attempt to model resto-
ration effects on rarer species, particularly sagebrush specialists. Future
questions should be addressed across the successional range of conifer
encroachment in sagebrush ecosystems and, ideally, replicated regionally.
Management Implications
Negative consequences of conifer encroachment on small mammals
far outweigh the short-term, minor impacts of sagebrush restoration.
Given the ecological signicance of small mammals as keystone species,
consumers, and prey species, maintenance of small mammal density is a
desirable ecological outcome for sagebrush restoration. Unless wood-
land specialists such as piñon mice and cliff chipmunks are species of
management concern, sagebrush restoration is an appropriate tool for
8B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
maintenance of small mammal diversity in the face of conifer encroach-
ment. Moderate increases in cheatgrass as a result of sagebrush restora-
tion may not negatively affect small mammal communities, provided
native shrubs and herbaceous plants are maintained and cheatgrass
cover values are low. However, sagebrush restoration projects should
anticipate increased annual grasses and plan for several years of herbi-
cide application and seeding following restoration treatments.
We would like to thank the eld assistants who helped to collect
these data: T. Athens, A. V. Conrad, R. Conrad, R. Cook, N. Darby, J. Dulgar,
P. Dupre, B. Eastman, K. Geyer, M. C. Grover, J. Keehn, E. J. Kolada, T.
Kollenbroich, M. Pepper, D. McKinney, J. Minott, S. Olind, J. Reynolds,
C. Rhine, J. Russell, R. Thomas, D. Walsh, D. Watrous, B. Williams, and
A. Yoder. T. Williams, K. Heister, and B. Roberts provided logistical sup-
port. S. Knick and E. Rickart provided valuable advice and reviewed
manuscript drafts.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
Austin, D.D., 1999. Changes in plant composition within a pinyon-juniper woodland. In:
Monsen, S.B., Stevens, R. (Eds.), Proceedings: ecology and management of pinyon-ju-
niper communities within the interior West: 1518 Sept, 1997; Provo, UT. Proceed-
ings of the RMRS-P-9. US Department of Agriculture, Forest Service, Rocky Mountain
Research Station, Ogden, UT, pp. 138140.
Barton, K.,2018. MuMIn: Multi-Model Inference. R package version 1.42.1. https://CRAN., Accessed date: 11 September 2018.
Baruch-Mordo, S., Evans, J.S., Severson, J.P., Naugle, D.E., Maestas, J.D., Kiesecker, J.M.,
Falkowski, M.J., Hagen, C.A., Reese, K.P., 2013. Saving sage-grouse from the trees: a
proactive solution to reducing a key threat to a candidate species. Biological Conser-
vation 167, 233241.
Baughman, C., Forbis, T.A., Provencher, L., 2010. Response of two sagebrush sites to low-
disturbance, mechanical removal of piñyon and juniper. Invasive Plant Science and
Management 3, 122129.
Bates, J.D., Miller, R.F., Svejcar, T., 2005.Long-term successional trends following western
juniper cutting. Rangeland Ecology & Management 58, 533541.
Bekoff, M., 1977. Canis latrans. Mammalian Species 19.
Block, W.M., Franklin, A.B., Ward, J.P., Ganey, J.L., White, G.C.,2001. Design and implemen-
tation of monitoring studies to evaluate the success of ecological restoration on wild-
life. Restoration Ecology 9, 293303.
Bombaci, S., Pejchar, L., 2016. Consequences of pinyon and juniper woodland reduction
for wildlife in North America. Forest Ecology and Management 365, 3450.
Bonham, C.D., 1989. Measurements for terrestrial vegetation. John Wiley and Sons, New
York, NY, USA.
Burnham, K.P., Anderson, D.R., 2002. Modelselection and multimodel inference: a practi-
cal informationtheoretic approach. Springer, Berlin, Germany 488 p.
Ceradini, J.P., Chalfoun, A.D.,2017. Speciestraits help predict small mammal responses to
habitat homogenization by an invasive grass. Ecological Applications 27, 14511465.
Chambers, J.C., VanderWall, S.B., Schupp, E.W., 1999. Seed and seedling ecology of pinon
and juniper species in the pygmy woodlands of Western North America. Botanical
Review 65, 138.
Clements, C.D., Young, J.A., 1996. Inuence of rodent predation on antelope bitterbrush
seedlings. Journal of Range Management 49, 3134.
de Lucas, M., Janss, G.F.E., Ferrer, M., 2005. A bird and small mammal BACI and IG design
studies in a wind farm in Malpica (Spain). Biodiversity & Conservation 14,
Efford, M.G., Dawson, D.K., Borchers, D.L., 2009. Population density estimated from loca-
tions of individuals on a passive detector array. Ecology 90, 26762682.
Efford, M.G., 2018. secr: Spatially explicit capture-recapture models. R package version
3.1.5., Accessed date: 11 September 2018.
Everett, R.L., Meeuwig, R.O., Stevens, R., 1978. Deer mouse preference for seed of com-
monly planted species, indigenous weed seed, and sacrice Foods. Journal of Range
Management 31, 7073.
Felicetti, L.A., Schwartz, C.C., Rye, R.O., Haroldson, M.A., Gunther, K.A., Phillips, D.L.,
Robbins, C.T., 2003. Use of sulfur and nitrogen stable isotopes to determine the im-
portanceof whitebark pine nutsto Yellowstone grizzly bears. Canadian Journal of Zo-
ology 81, 763770.
Freeman, E.D., Sharp, T.R., Larsen, R.T., Knight, R.N., Slater, S.J., McMillan, B.R., 2014. Neg-
ative effects of an exotic grass invasion on small-mammal communities. PLoS ONE 9,
Glaudas, X., Jezkova, T., Rodriguez-Robles, J.A., 2008. Feeding ecology of the Great Basin
Rattlesnake (Crotalus lutosus, Viperidae). Canadian Journal of Zoology 86, 723-234.
Gruell, G.E., Eddleman, L.E., Jaindl, R., 1994. Fire history of the pinyon-juniper wood-
lands of Great Basin National Park. Technical Report NPS/PNROSU/NRTR-94/01,
pp. 127.
Hanser, S.E., Knick, S.T., 2011. Greater sage-grouse as an umbrella species for shrubland
passerine birds: a multiscale assessment. In: Knick, S.T., Connelly, J.W. (Eds.), Greater
sage grouse:ecology and conservation of a landscapespecies and its habitats.Studies
in avian biology No. 38. University ofCalifornia Press, Berkeley, CA,USA, pp. 475488.
Hart, E.B., 1992. Tamias dorsalis. Mammalian Species 16.
Herrick, J.E., Zee, J.W.V., Havstad, K.M., Burkett, L.M., Whitford, W.G., 2005. Monitoring
manual for grassland, shrubland, and savannaecosystems. Volume II: design, supple-
mentary methods, and interpretation. USDA-ARS Jornado Experimental Range, Las
Cruces, NM, USA 200 p.
Hoffmeister, D.F., 1981. Peromyscus truei. Mammalian Species 15.
Hollander, J.L., Vander Wall, S.B., Baguley, J.G., 2010. Evolution of seed dispersal in North
American ephedra. Evolutionary Ecology 24, 333345.
Hormay, A.L., 1943. Bitterbrush in California. US Department of Agriculture Forest Service,
Forest Research Note 34, 113.
Huntly, N., Inouye, R., 1988. Pocket gophers in ecosystems: patterns and mechanisms.
BioScience 38, 786793.
Keane, R.E., Ryan, K.C., Veblen, T.T., Allen, C.D., Logan, J., Hawkes, B., 2002. Cascading ef-
fects of re exclusion in the Rocky Mountain ecosystems: a literature review. USDA
Forest Service General Technical Report RMRSGTR-91, pp. 124.
Kitchen, S.G., Jorgensen, G.L., 1999. Annualization of rodent burrow clusters andwinterfat
decline in a salt-desert community. In: McArthur, E.D., Ostler, W.K., Wambolt, C.L.
(Eds.), Proceedings: shrubland ecotones; 1214 Aug 1998; Ephraim, UT. Proc.
RMRS-P-11. US Department of Agriculture, Forest Service, Rocky Mountain Research
Station, Ogden, UT, USA, pp. 175180.
Knick, S.T., Hanser, S.E., Leu, M., 2014. Ecological scale of bird community response to
piñon-juniper removal. Rangeland Ecology & Management 67, 553562.
Magurran, A.E., 2004. Measuring biological diversity. Blackwell Publishing, Malden, MA,
USA, p. 256.
McAdoo, J.K., Evans, C.C., Roundy, B.A., Young, J.A., Evans, R.A., 1983. Inuence of
heteromyid rodents on Oryzopsis hymenoides germination. Journal of Range Man-
agement 36, 6164.
McIver, J., Macke, E., 2014. Short-term buttery response to sagebrush steppe restoration
treatments. Rangeland Ecology & Management 67, 539552.
Miller, M.E.,2008. Broad-scale assessment of rangeland health,Grand StaircaseEscalante
National Monument, USA. Rangeland Ecology & Management 61, 249262.
Miller, R.F., Tausch, R.J., 2001. The role of re in pinyon and juniper woodlands: a descrip-
tive analysis. Proceedings of the Invasive Species Workshop: the role of re in the
control and spread of invasive species. Fire Conference 2000: the First National Con-
gress on Fire Ecology, Prevention, and Management. Miscellaneous Publication No.
11. Tall Timbers Research Station, Tallahassee, FL, USA, pp. 1530.
Miller, R.F., Svejcar,T.J., West, N.E., 1994. Implications of livestock grazing in thethe inter-
mountain sagebrush region: plant composition. In: Vavra, M., Laycock, W.A., Pieper,
R.D. (Eds.), Ecological implications of livestock herbivory in the West. Society for
Range Management, Denver, CO, USA, pp. 101146.
Miller, R.F., Bates, J.D., Svejcar,T.J., Pierson, F.B., Eddleman, L.E., 2005.Biology, ecology, and
management of the western juniper. Oregon State University Agricultural Experi-
ment Station, Corvallis, OR, USA Technical Bulletin 52, June 2005.
Miller, R.F., Tausch, R.J., McArthur, E.D., Johnson,D.D., Sanderson, S.C., 2008. Age structure
and expansion of piñon-juniper woodlands: a regional perspective in the intermoun-
tain West. Research paper. RMRS-RP-69. US Department of Agriculture, Forest Ser-
vice, Rocky Mountain Research Station, Fort Collins, CO, USA 15 p.
Ostoja, S.M., Schupp, E.W., 2009. Conversion of sagebrush shrublands to exotic annual
grasslands negatively impacts small mammal communities. Diversity and Distribu-
tions 15, 863870.
Pellant, M., Abbey, B., Karl, S., 2004. Restoring the Great Basin Desert, U.S.A.: integrating
science, management, and people. Environmental Monitoring and Assessment 99,
R Core Team, 2016. R: A language and environment for statistical computing. R Founda-
tion for Statistical Computing, Vienna, Austria URL,
Accessed date: 11 September 2018.
Rapp, V., 2004. Western forests, re risk, and climate change. Pacic Northwest Research
Station Science, pp. 112 Update January.
Rickart, E.A., Robson, S.L., Heaney, L.R., 2008. Mammals of Great Basin National Park, Ne-
vada: Comparative Field Surveys and Assessment of Faunal Change. Western North
American Naturalist 4, 77114.
Rodhouse, T.J., Hirnyck, R.P., Wright, R.G., 2010. Habitat selection of rodents along a
piñonjuniper woodlandsavannah gradient. Journal of Mammalogy 91, 447457.
Rowe, R.J., Terry, R.C., Rickart, E.A., 2011. Environmental change and declining resource
availability for small-mammal communities in the Great Basin. Ecology 92,
Sikes, R.S., Gannon, W.L., The Animal Care and Use Committee ofthe American Society of
Mammalogists, 2011. Guidelines of the American Society of Mammalogists for the
use of wild mammals in research. Journal of Mammalogy 92, 235253.
Sirotnak, J.M., Huntly, N.J., 2000.Direct and Indirect Effects of Herbivores on Nitrogen Dy-
namics: Voles in Riparian Areas. Ecology 81, 7887.
Skaug, H., Fournier, D., Bolker, B., Magnusson, A., Nielsen, A., 2014. Generalized Linear
Mixed Models using AD Model Builder R package version 0.8.0.
Soanes, K., Taylor, A.C., Sunnucks, P., Vesk, P.A., Cesarini, S., van der Ree, R., 2018. Evaluat-
ing the success of wildlife crossing structures using genetic approaches and an exper-
imental design: Lessons from a gliding mammal. Journal of Applied Ecology 55,
Stephens, R.B., Hocking, D.J., Yamasaki, M., Rowe, R.J., 2017. Synchrony in small mammal
community dynamics across a forested landscape. Ecography 40, 11981209.
9B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
Tausch, R.J., Roundy, B.A., Chambers, J.C., 2009.Piñon and Juniper Field Guide: Asking the
right Questions to select appropriate managment actions. U.S. Geological Survey Cir-
cular 1335, 196.
Underwood, A.J., 1992. Beyond BACI: the detection of environmental impacts on popula-
tions in the real, but variable, world. Journal of Experimental Marine Biology and
Ecology 161, 145178.
Underwood, A.J., 1994.On Beyond BACI: Sampling Designs that Might Reliably Detect En-
vironmental Disturbances. Ecological Applications 4, 415.
USDA NaturalResources Conservation Service, 2009.Soil of Great Basin National Park, Ne-
vada. Accessible online at.
Weltz, M.A., Spaeth, K., Taylor, M.H., Rollins, K., Pierson,F., Jolley, L., Nearing, M., Goodrich,
D., Hernandez, M., Nouwakpo, S.K., Rossi, C., 2014. Cheatgrass invasion and woody
species encroachment in the Great Basin: Benets of conservation. Journal of Soil
and Water Conservation 69, 39A44A.
White, C.G., Flinders, J.T., Cates, R.G., Blackwell, B.H., Smith, H.D., 1999. Dietary Use of Utah
Juniper Berries by Gray Fox in Eastern Utah. In: Monsen, S.B., Stevens, R. (Eds.), Pro-
ceedings: ecology and management of pinyon-juniper communities within the
Interior West. Provo, UT, USA: 15 18 September; Proceedings of the RMRS-P-9.
US Department of Agriculture, Forest Service, Rocky Mountain Research Station,
Ogden, UT, USA, pp. 219232.
Wisdom, M.J., Wales, B.C., Rowland, M.M., Raphael, M.G., Holthausen, R.S., Rich, T.D., Saab,
V.A., 2002. Performance of greater sage-grouse models for conservation assessment
in the Interior Columbia Basin, U.S.A.The Journal of the Society for Conservation Biol-
ogy 16, 12321242.
Woods, B.A., Rachlow, J.L., Bunting, S.C., Johnson, T.R., Bocking, K., 2013. Managing high-
elevation sagebrush steppe: do conifer encroachment and prescribed re affect hab-
itat for pygmy rabbits? Rangeland Ecology & Management 66, 462471.
Young, J.A., Clements, C.D., 2002. Purshia: the wild and bitter roses. University ofNevada
Press, Reno and Las Vegas, NY, USA.
Young, J.A., Clements, C.D., 2009. Cheatgrass: re and forage on the range. University of
Nevada Press, Reno, NV, USA.
10 B.T. Hamilton et al. / Rangeland Ecology & Management xxx (xxxx) xxxxxx
Please cite this article as: Hamilton, B.T., et al., Effects of Sagebrush Restoration and Conifer Encroachment on Small Mammal Diversity in
Sagebrush Ecosystem, (2018),
... For example, Bachen et al. (2018) experimentally manipulated litter to represent heavily cheatgrass-invaded areas, then found less consumption of food by rodents compared to control areas with less litter. Hamilton et al. (2019) found that conifer encroachment into sage steppe-a strong change in habitatled to much lower densities of pocket mice. Further, we did not examine other possible predictors of movements, such as the spatial distribution of seeds (Guo et al. 1998;Barga and Leger 2018) or the risk of predation (Stoddard et al. 2019). ...
Foraging animals choose habitats based on characteristics that often cannot be satisfied simultaneously, such as easy mobility, abundant or high-quality foods, and safety from predators. Invasive plants may alter habitat structure and provide novel foods; thus, measuring how animals forage in invaded landscapes offers insights into these new ecological relationships. We examined the movements of Great Basin pocket mice (Perognathus parvus) in sage-steppe habitats invaded by cheatgreass (Bromus tectorum) in southcentral British Columbia, Canada. The pathway tortuosity (fractal D) of pocket mice increased with vegetative cover and population density and decreased with open habitat, but these variables explained little of the variation in tortuosity. The fractal dimension of movement pathways of pocket mice was consistent over spatial scales ranging from 0.5 m to two-thirds of the home range size, unlike in other species where fractal dimensions are not consistent over multiple spatial scales. Collectively, our results indicate that foraging movements of pocket mice were not affected by the low densities of cheatgrass in this system.
... 63,64 Changes in plant communities, in turn, indirectly affect rodent populations through direct effects on availability of food and cover and other aspects of habitat suitability. Examples include increased rodent densities in response to juniper removal and rangeland reseeding, 65 reduced populations of a woodland rodent species in response to sagebrush restoration in areas impacted by conifer expansion, 66 and changes in rodent populations resulting from biological control of introduced woody plants. 67 Thus, the essential roles of scatter-hoarding rodents that we have highlighted here impact not only the plant populations involved, but may feedback in a cyclical fashion to yield habitat effects that impact rodent populations and communities. ...
Full-text available
Precipitation events have been predicted and observed to become fewer, but larger, as the atmosphere warms. Water-limited ecosystems are especially sensitive to changes in water cycling, yet evidence suggests that productivity may either increase or decrease in response to precipitation intensification. Interactions among climate, soil properties, and vegetation type may explain different responses, but this is difficult to experimentally test over large spatial scales. Simulation modeling may reveal the mechanisms through which climate, soils, and vegetation interact to affect plant growth. We use an individual-based plant ecohydrological model to simulate the effects of 25%, 50%, and 100% increases in precipitation event sizes on water cycling and shrub, grass, and forb biomass in 200 shrub-steppe sites spanning 651,000 km² of the Intermountain West, USA. Simulations did not change annual precipitation amounts and were performed for 0, 3, and 5 °C warming. Larger precipitation events decreased evaporation and ‘pushed’ water into shrub root zones in arid and semi-arid sites, but ‘pushed’ water below shrub root zones in mesic sites resulting in increased shrub biomass in arid and semi-arid, but not mesic, sites. Positive effects of precipitation intensification on shrub growth partially counteracted negative effects of warming. Grasses and forbs showed no consistent response to precipitation intensification. Results indicate that increased precipitation intensity creates a competitive advantage for shrubs in arid and semi-arid sites. This advantage results in greater shrub relative abundance and suggests that precipitation intensification contributes to woody plant encroachment observed globally in arid and semi-arid ecosystems.
Full-text available
Adaptive management and monitoring efforts focused on vegetation, habitat, and wildlife in the sagebrush (Artemisia spp.) biome help inform management of species and habitats, predict ecological responses to conservation practices, and adapt management to improve conservation outcomes. This chapter emphasizes the adaptive resource management framework with its four stages: (1) problem definition, (2) outcomes, (3) decision analysis, and (4) implementation and monitoring. Adaptive resource management is an evolving process involving a sequential cycle of learning (the accumulation of understanding over time) and adaptation (the adjustment of management over time). This framework operationalizes monitoring a necessary component of decision making in the sagebrush biome. Several national and regional monitoring efforts are underway across the sagebrush biome for both vegetation and wildlife. Sustaining these efforts and using the information effectively is an important step towards realizing the full potential of the adaptive management framework in sagebrush ecosystems. Furthermore, coordinating monitoring efforts and information across stakeholders (for example, Federal, State, nongovernmental organizations) will be necessary given the limited resources, diverse ownership/management, and sagebrush biome size.
Full-text available
Rangelands of the United States provide ecosystem services that sustain biodiversity and rural economies. Native tree encroachment is a recognized and long-standing conservation challenge to these landscapes, but its impact is often overlooked due to the slow pace of tree invasions and the positive public perception of trees. Here we show that tree encroachment is a dominant change agent in U.S. rangelands; tree cover has increased by more than 77,000 km ² over 30 years, and more than 25% of U.S. rangelands are now experiencing sustained tree cover expansion. Further, we use machine learning methods to estimate the potential herbaceous production (forage) lost to tree encroachment. Since 1990 roughly 300 Tg of herbaceous biomass has been lost, totaling some $5 billion in foregone revenue to agricultural producers. These results suggest that tree encroachment is similar in scale and magnitude to row-crop conversion, another primary cause of rangeland loss in the U.S. Prioritizing conservation efforts to prevent tree encroachment in rangelands can bolster ecosystem and economic sustainability of these landscapes, particularly among privately-owned lands threatened by land-use conversion.
Globally, invasive species in grasslands and shrublands have substantially altered trophic interactions and ecosystem processes. In North America, cheatgrass (Bromus tectorum) has invaded much of the western sage-steppe, leading to substantial shifts in plant communities and fire regimes. The effects of this cheatgrass invasion on small mammals are inconsistent among species and regions, and the mechanisms behind these patterns may be trophic or non-trophic. Cheatgrass alters cover and the amount of bare ground, thus affecting movements and predation risk, but cheatgreass also provides food for some rodents. We examined the extent to which low-density cheatgrass invasion (<20% cover) affected population densities and habitat selection of mice in southern British Columbia, Canada. In this region, Great Basin pocket mice (Perognathus parvus) were far more common than western harvest mice (Reithrodontomys megalotis) and deermice (Peromyscus maniculatus). Density of Great Basin pocket mice increased with bare ground and annual grasses, but decreased with shrub cover; when pocket mice foraged, they selected habitats with cheatgrass and bare, open ground. Cheatgrass did not predict movements of deermice or harvest mice. These results indicate that low cheatgrass densities may not be negative for these three species of mice, at least in the northern distribution of sage-steppe.
Full-text available
As part of the Sagebrush Steppe Treatment Evaluation Project (SageSTEP), butterflies were surveyed pretreatment and up to 4 yr posttreatment at 16 widely distributed sagebrush steppe sites in the interior West. Butterfly populations and communities were analyzed in response to treatments (prescribed fire, mechanical, herbicide) designed to restore sagebrush steppe lands encroached by piñon-juniper woodlands (Pinus, Juniperus spp.) and invaded by cheatgrass (Bromus tectorum). Butterflies exhibited distinct regional patterns of species composition, with communities showing marked variability among sites. Some variation was explained by the plant community, with Mantel's test indicating that ordinations of butterfly and plant communities were closely similar for both woodland sites and lower-elevation treeless (sage-cheat) sites. At woodland sites, responses to stand replacement prescribed fire, clear-cutting, and tree mastication treatments applied to 10–20-ha plots were subtle: 1) no changes were observed in community structure; 2) Melissa blues (Plebejus melissa) and sulfurs (Colias spp.) increased in abundance after either burning or mechanical treatments, possibly due to increase in larval and nectar food resource, respectively; and 3) the juniper hairstreak (Callophrys gryneus) declined at sites at which it was initially present, probably due to removal of its larval food source. At sage-cheat sites, after prescribed fire was applied to 25–75-ha plots, we observed 1) an increase in species richness and abundance at most sites, possibly due to increased nectar resources for adults, and 2) an increase in the abundance of skippers (Hesperiidae) and small white butterflies. Linkages between woody species removal, the release of herbaceous vegetation, and butterfly response to treatments demonstrate the importance of monitoring an array of ecosystem components in order to document the extent to which management practices cause unintended consequences.
Full-text available
Guidelines for use of wild mammal species are updated from the American Society of Mammalogists (ASM) 2007 publication. These revised guidelines cover current professional techniques and regulations involving mammals used in research and teaching. They incorporate additional resources, summaries of procedures, and reporting requirements not contained in earlier publications. Included are details on marking, housing, trapping, and collecting mammals. It is recommended that institutional animal care and use committees (IACUCs), regulatory agencies, and investigators use these guidelines as a resource for protocols involving wild mammals. These guidelines were prepared and approved by the ASM, working with experienced professional veterinarians and IACUCs, whose collective expertise provides a broad and comprehensive understanding of the biology of nondomesticated mammals in their natural environments. The most current version of these guidelines and any subsequent modifications are available at the ASM Animal Care and Use Committee page of the ASM Web site (
Millions of dollars are spent on wildlife crossing structures intended to reduce the barrier effects of roads on wildlife. However, we know little about the degree to which these structures facilitate dispersal and gene flow. Our study incorporates two elements that are rarely used in the evaluation of wildlife crossing structures: an experimental design including a before and after comparison, and the use of genetic techniques to demonstrate effects on gene flow at both population and individual levels. We evaluated the effect of wildlife crossing structures (canopy bridges and glider poles) on a gliding mammal, the squirrel glider (Petaurus norfolcensis). We genotyped 399 individuals at eight microsatellite markers to analyse population structure, first-generation migrants and parentage relationships. We found that the freeway was not a complete genetic barrier, with a strong effect evident at only one site. We hypothesise that the presence of corridors alongside the freeway and throughout the surrounding landscape facilitated circuitous detours for squirrel gliders. Installing a crossing structure at the location with a strong barrier effect restored gene flow within just five years of mitigation. Synthesis and applications. Our study highlights the importance of using genetic techniques not just to evaluate the success of road crossing structures for wildlife, but also to guide their placement within the landscape. Managers wishing to reduce the effects of linear infrastructure on squirrel gliders and other arboreal mammals should aim to preserve and enhance vegetation along roadsides and within centre medians, as well as mitigate large gaps by implementing wildlife crossing structures.
Invasive plants can negatively affect native species, however, the strength, direction, and shape of responses may vary depending on the type of habitat alteration and the natural history of native species. To prioritize conservation of vulnerable species, it is therefore critical to effectively predict species’ responses to invasive plants, which may be facilitated by a framework based on species’ traits. We studied the population and community responses of small mammals and changes in habitat heterogeneity across a gradient of cheatgrass (Bromus tectorum) cover, a widespread invasive plant in North America. We live-trapped small mammals over two summers and assessed the effect of cheatgrass on native small mammal abundance, richness, and species-specific and trait-based occupancy, while accounting for detection probability and other key habitat elements. Abundance was only estimated for the most common species, deer mice (Peromyscus maniculatus). All species were pooled for the trait-based occupancy analysis to quantify the ability of small mammal traits (habitat association, mode of locomotion, and diet) to predict responses to cheatgrass invasion. Habitat heterogeneity decreased with cheatgrass cover. Deer mouse abundance increased marginally with cheatgrass. Species richness did not vary with cheatgrass, however, pocket mouse (Perognathus spp.) and harvest mouse (Reithrodontomys spp.) occupancy tended to decrease and increase, respectively, with cheatgrass cover, suggesting a shift in community composition. Cheatgrass had little effect on occupancy for deer mice, 13-lined ground squirrels (Spermophilus tridecemlineatus), and Ord's kangaroo rat (Dipodomys ordii). Species’ responses to cheatgrass primarily corresponded with our a priori predictions based on species’ traits. The probability of occupancy varied significantly with a species’ habitat association but not with diet or mode of locomotion. When considered within the context of a rapid habitat change, such as caused by invasive plants, relevant species’ traits may provide a useful framework for predicting species’ responses to a variety of habitat disturbances. Understanding which species are likely to be most affected by exotic plant invasion will help facilitate more efficient, targeted management and conservation of native species and habitats. This article is protected by copyright. All rights reserved.
Long-term studies at local scales indicate that fluctuations in abundance among trophically similar species are often temporally synchronized. Complementary studies on synchrony across larger spatial extents are less common, as are studies that investigate the subsequent impacts on community dynamics across the landscape. We investigate the impact of species population fluctuations on concordance in community dynamics for the small mammal fauna of the White Mountain National Forest, USA. Hierarchical open population models, which account for imperfect detection, were used to model abundance of the most common species at 108 sites over a three year period. Most species displayed individualistic responses of abundance to forest type and physiographic characteristics. However, among species, we found marked synchrony in population fluctuations across years, regardless of landscape affinities or trophic level. Across the region, this population synchrony led to high within-year concordance of community composition and aggregate properties (e.g., richness and diversity) independent of forest type and low among-year similarity in communities, even for years with similar species richness. Results suggest that extrinsic factors primarily drive abundance fluctuations and subsequently community dynamics, although local community assembly may be modified by species dispersal abilities and biotic interactions. Concordant community dynamics across space and over time may impact the stability of regional food webs and ecosystem functions. This article is protected by copyright. All rights reserved.
Tools for performing model selection and model averaging. Automated model selection through subsetting the maximum model, with optional constraints for model inclusion. Model parameter and prediction averaging based on model weights derived from information criteria (AICc and alike) or custom model weighting schemes. [Please do not request the full text - it is an R package. The up-to-date manual is available from CRAN].
This is a scientific and historical account of the cheatgrass invasion. It is a study of the plant which has changed the ecology of millions of acres of western rangeland.
Antelope bitterbrush (Purshia tridentata(Pursh) DC) is the most important browse species on many mule deer (Odocoileus hemionus) ranges. California-Nevada interstate mule deer herds are critically dependent on antelope bitterbrush stands, in which many of these stands have been and are currently exhibiting little recruitment. Lassen is the only established cultivar of antelope bitterbrush. Rodent predation on Lassen antelope bitterbrush seedlings was studied in burned and unburned antelope bitterbrush communities in northeastern California during 1993. Rodent population densities were 15/ha and 14/ha in the burned and unburned habitats, respectfully. Rodent compositions consisted of the Ord's kangaroo rat (Dipodomys ordii), deer mouse (Peromyscus maniculatus), and the Great Basin pocket mouse (Perognathus parvus). Rodents significantly decreased antelope bitterbrush recruitment through grazing and disturbance of antelope bitterbrush seedlings. Ord's kangaroo rats preyed on higher numbers of antelope bitterbrush seedlings than did the other 2 common rodent species.