Leveraging extensive soil, vegetation, fire, and land treatment data to inform restoration across the sagebrush biome

Abstract

Context
Widespread ecological degradation has prompted calls for massive global investments in ecological restoration, yet limited resources necessitate efficient application of restoration efforts. In western North America, altered fire regimes are increasing the scale of restoration needed to preserve the sagebrush (Artemisia species) biome but prioritizing and implementing effective restoration is complicated by the vast and heterogeneous sagebrush landscape, which includes gradients in climate, disturbance, and species composition.

Objectives
To develop spatially explicit and context-dependent estimates of treatment efficacy and sagebrush recovery rates.

Methods
We leveraged a suite of spatio-temporally extensive datasets to evaluate the influence of restoration treatments and environmental conditions on trends in post-disturbance sagebrush cover, with an emphasis on understanding differences between sites recovering naturally and sites receiving restoration treatments. We used estimates from these models to develop spatially explicit projections for sagebrush recovery, conditional on disturbance, restoration practice, and environmental conditions.

Results
We found seeding Artemisia spp. increased sagebrush cover over time relative to natural recovery, but this relationship depended on spring soil moisture availability and treatment methods. Natural recovery was positively influenced by soil moisture and sagebrush cover and negatively influenced by cumulative burns and annual herbaceous cover, while the influence of perennial herbaceous cover varied with soil moisture.

Conclusions
Our results provide biome-wide insights and spatially explicit tools that can inform economic cost-effectiveness analyses, restoration prioritization tools, and other scientific endeavors to ensure managers have the tools and information needed to effectively steward the sagebrush biome in a rapidly changing world.

Full text

Vol.: (0123456789)
https://doi.org/10.1007/s10980-024-01968-z
RESEARCH ARTICLE
Leveraging extensive soil, vegetation, fire, andland
treatment data toinform restoration acrossthesagebrush
biome
BryanC.Tarbox · AdrianP.Monroe· MichelleI.Jeffries· JustinL.Welty· MichaelS.O’Donnell·
RobertS.Arkle· DavidS.Pilliod· PeterS.Coates· JulieA.Heinrichs· DanielJ.Manier·
CameronL.Aldridge
Received: 2 March 2024 / Accepted: 22 August 2024
©Julie Heinrichs. Parts of this work were authored by US Federal Government authors and are not under copyright protection in the
US; foreign copyright protection may apply 2024
Abstract
Context Widespread ecological degradation has
prompted calls for massive global investments in eco-
logical restoration, yet limited resources necessitate
efficient application of restoration efforts. In western
North America, altered fire regimes are increasing the
scale of restoration needed to preserve the sagebrush
(Artemisia species) biome but prioritizing and imple-
menting effective restoration is complicated by the
vast and heterogeneous sagebrush landscape, which
includes gradients in climate, disturbance, and spe-
cies composition.
Objectives To develop spatially explicit and con-
text-dependent estimates of treatment efficacy and
sagebrush recovery rates.
Methods We leveraged a suite of spatio-temporally
extensive datasets to evaluate the influence of resto-
ration treatments and environmental conditions on
trends in post-disturbance sagebrush cover, with an
emphasis on understanding differences between sites
recovering naturally and sites receiving restoration
treatments. We used estimates from these models to
develop spatially explicit projections for sagebrush
recovery, conditional on disturbance, restoration prac-
tice, and environmental conditions.
Results We found seeding Artemisia spp. increased
sagebrush cover over time relative to natural recovery,
but this relationship depended on spring soil moisture
availability and treatment methods. Natural recovery
was positively influenced by soil moisture and sage-
brush cover and negatively influenced by cumulative
burns and annual herbaceous cover, while the influ-
ence of perennial herbaceous cover varied with soil
moisture.
Conclusions Our results provide biome-wide
insights and spatially explicit tools that can inform
economic cost-effectiveness analyses, restoration
prioritization tools, and other scientific endeavors
to ensure managers have the tools and information
Supplementary Information The online version
contains supplementary material available at https:// doi.
org/ 10. 1007/ s10980- 024- 01968-z.
B.C.Tarbox(*)· A.P.Monroe· M.S.O’Donnell·
D.J.Manier· C.L.Aldridge
U.S. Geological Survey, Fort Collins Science Center, 2150
Centre Avenue, Building C, FortCollins, CO80526, USA
e-mail: btarbox@usgs.gov
M.I.Jeffries· J.L.Welty· R.S.Arkle· D.S.Pilliod
U.S. Geological Survey, Forest andRangeland Ecosystem
Science Center, 230 N Collins Rd, Boise, ID83702, USA
P.S.Coates
U.S. Geological Survey, Western Ecological Research
Center, 800 Business Park Drive, Suite D, Dixon,
CA95620, USA
J.A.Heinrichs
Natural Resource Ecology Laboratory, Colorado State
University, FortCollins, CO80523, USA
J.A.Heinrichs
In Cooperation withU.S. Geological Survey, Fort Collins
Science Center, FortCollins, CO, USA
Landsc Ecol (2024) 39:184
/ Published online: 19 October 2024
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
needed to effectively steward the sagebrush biome in
a rapidly changing world.
Keywords Artemisia spp.· Recovery· Resilience·
Seeding· Soil moisture availability
Introduction
Ecosystem function (Cardinale et al. 2012) and
biodiversity (Newbold et al. 2015) are degraded
through intensive land use (e.g., agriculture,
development), unsustainable use of natural resources
(e.g., overgrazing), species’ invasions, and climate
change (Sage 2020). Preserving intact ecosystems
alone is unlikely to mitigate such alterations to
the environment given their extent and severity,
necessitating complementary efforts to restore
degraded ecosystems (Aronson and Alexander 2013;
Gann etal. 2019). Ecological restoration can reverse
degradation by accelerating recovery of ecosystem
structure and function, leading to improvements in
biodiversity and ecosystem services (Rey Benayas
etal. 2009). Indeed, the United Nations has called for
a global decade of ecosystem restoration (2021–2030)
to prevent, halt, and reverse the loss and deterioration
of nature (United Nations General Assembly 2019).
Despite advances in methodology, outcomes
of restoration efforts are still mixed (Svejcar et al.
2017; Jones et al. 2018). Studies informing these
efforts are often confined to specific sites that may
not be transferrable across ecosystems and regions,
potentially limiting inferences for improving
restoration outcomes (Suding 2011; Cooke et al.
2019). Furthermore, resources (e.g., time, supplies,
equipment, personnel) for implementing ecological
restoration are limited, making strategic deployment
a critical priority that can dramatically improve
restoration efficiency (Strassburg et al. 2019).
Regional analyses accounting for spatio-temporal
variability are thus needed to reveal ecological
processes that influence restoration outcomes (Perring
et al. 2015; Yirdaw et al. 2017). Such studies may
better indicate whether restoration will be effective
(Holl etal. 2003; Brudvig 2011) or necessary (Holl
and Aide 2011; Jones et al. 2018) by advancing
understanding of ecological conditions that determine
treatment outcomes.
Success of ecological restoration is influenced by
ecological resilience—the amount of disturbance
an ecosystem can endure and still recover its
fundamental structure and processes (Holling 1973;
Walker etal. 2004; Briske etal. 2008; Chambers etal.
2014b). Arid and semi-arid ecosystems (hereafter
drylands) generally exhibit low resilience given high
aridity, highly variable precipitation (i.e., prone to
extended droughts, and occasional deluge), and low
soil fertility (Maestre et al. 2012; Wu et al. 2015),
making these systems particularly susceptible to
degradation. Specifically, soil climate (i.e., soil
temperature and moisture) is a principal driver of
plant physiological function (e.g., plant growth and
microbial metabolism), contributing to primary
productivity, diversity, and ecological integrity
(Noy-Meir 1973; Boer and Puigdefábregas 2005).
Unfavorable or variable soil climates can complicate
restoration efforts by reducing recovery rates and
increasing probability of restoration failure (Wu
etal. 2015; Shriver et al. 2018; Barnard etal. 2019;
O’Connor etal. 2020). Soil climate is fundamental to
understanding resilience in drylands of western North
America dominated by sagebrush (Artemisia spp.;
Chambers et al. 2014b); indeed, the distributions of
big sagebrush subspecies (Artemisia tridentata ssp.)
are influenced by adaptations to drought conditions
(Kolb and Sperry 1999) and temperature (Hansen
etal. 2008; Chaney etal. 2017).
The sagebrush biome once extended across > 500,000
km2 of western North America but extensive and
accumulating degradation eliminated sagebrush from
nearly half its historical extent (Miller et al. 2011),
rendering this one of the most imperiled ecosystems on
the continent (Noss etal. 1995; Doherty et al. 2022). A
major challenge confronting sagebrush conservation is
changing fire regimes driven by the synergy between
exotic annual grasses and wildfire that has converted
some landscapes to annual grasslands (Balch et al.
2013; Chambers et al. 2014a). Sagebrush mortality
rates from wildfire are often high (Blaisdell 1958), and
sagebrush is a slow-growing plant that regenerates from
seeds with limited longevity (1–4 years; Wijayratne
and Pyke 2009) and dispersal distances (< 30 m;
Applestein et al. 2022), often requiring decades for
system recovery (e.g., up to 120years; Baker 2006). If
disturbance eliminates enough individuals and the seed
bank is depleted, active restoration may be necessary
to reestablish sagebrush (Pyke 2011), but research is
Landsc Ecol (2024) 39:184
Page 2 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
needed to evaluate effectiveness of specific restoration
practices across edaphic, climatic, and socio-ecological
gradients spanning this sub-continental region (Davies
et al. 2011; Pyke et al. 2013; Germino et al. 2018).
Regional syntheses and field studies indicate resilience
of sagebrush ecosystems is driven by extensive
climatic and topographic conditions (e.g., precipitation,
elevation) that influence sagebrush germination,
growth, and survival, presumably through direct and
indirect relationships with soil moisture availability
(Chambers et al. 2014a; Svejcar et al. 2017; Barnard
etal. 2019; O’Connor etal. 2020). Moreover, restoration
effectiveness also varies with local site conditions
beyond soil climate, as factors such as land-use legacy,
competition, seeding and other treatment methods, and
weather can influence restoration outcomes (Eiswerth
etal. 2009; Morris et al. 2014; Ott etal. 2016; Simler‐
Williamson et al. 2022). Studies addressing these
relationships have increasingly broad extents, which has
improved inference and understanding of ecoregional
variation (Shriver et al. 2018; Chambers et al. 2021;
Arkle et al. 2022). However, while past studies
examined similar questions at broad spatial scales, many
used a space-for-time substitution approach, rather than
a longitudinal tracking of restoration outcomes (e.g.,
Arkle etal. 2014; Knutson etal. 2014), or were limited
in duration (e.g., < 10years; Pyke etal. 2013; Copeland
etal. 2018). With wildfires rapidly increasing the need
for effective restoration (Copeland et al. 2018), there
is currently a lack of studies investigating sagebrush
recovery with the spatio-temporal breadth needed to
adequately generalize results for real-world management
questions (Pyke etal. 2013; Arkle etal. 2014; Hardegree
et al. 2016; Chambers et al. 2017a). A broader
perspective also may reveal how restoration dynamics
change over space and time, supporting strategies that
integrate regional-scale heterogeneity into restoration
planning (Hobbs and Norton 1996; Bell et al. 1997;
Brudvig 2011).
Studying restoration at broad spatio-temporal
extents is increasingly facilitated by datasets
cataloging restoration practices over space and time
(Heller et al. 2017; Pilliod et al. 2017) as well as
remote sensing products characterizing disturbance
(Hawbaker et al. 2020), vegetation cover (Rigge
et al. 2021), and soil moisture (O’Donnell and
Manier 2022b). Here, we leveraged these resources
to evaluate influence of restoration treatments and
environmental conditions on sagebrush recovery
across the biome. Specifically, we modeled trends in
post-disturbance cover of sagebrush at sites with and
without restoration treatments using data from > 1000
projects (1986–2021). We used these models to
develop spatially explicit projections for sagebrush
recovery, conditional on disturbance, restoration
practice, and environmental conditions. These
analyses may provide insights that guide effective
restoration of sagebrush-dominated ecosystems, as
well as maps to prioritize restoration efforts while
considering factors that influence spatio-temporal
variation in recovery rates from different restoration
treatments.
Methods
Study area
Our study occurred in sagebrush-dominated ecosys-
tems within the sagebrush biome (Fig.1; Jeffries and
Finn 2019). The extent incorporates multiple ecore-
gions with different climatic conditions. Major pre-
cipitation seasons vary from winter snow and rain in
the west to mostly summer rains in the east, thereby
influencing water availability, composition of plant
functional types, and interactions within the plant
community (Chambers etal. 2017a). The biome also
hosts 18 species and subspecies of Artemisia (Rem-
ington et al. 2021), with major sub-species of big
sagebrush (A. tridentata ssp.) occurring across gradi-
ents of soil conditions that range from warm and dry
(A. t. ssp. wyomingensis) to cool and moist (A. t. ssp.
vaseyana; Miller etal. 2011).
Treatment databases
We compiled land treatment data from three datasets
overlapping our study area: the Land Treatment
Digital Library (LTDL, 1917–2019; Pilliod et al.
2019), the Conservation Efforts Database (CED,
2009–2019; Heller etal. 2017), and Utah’s Watershed
Restoration Initiative (UWRI, 2005–2023; https:// wri.
utah. gov/ wri/). In organizing information contained
in these databases, we distinguished between
individual projects, each with an overall objective
(purpose, including addressing wildfire, grazing,
habitat/restoration, and hazards), and one or more
treatments associated with each project (Table S1).
Landsc Ecol (2024) 39:184 184
Page 3 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
We summarize our approach to integrating these
databases in Online Resource 1, including processing
each dataset to increase reliability and relevance of
treatment data for our objectives.
From each database, we acquired treatment infor-
mation including seeding method (aerial, drill, ground
[non-drill or unknown ground seeding], seedling
planting, greenstrip, and unknown seeding method;
defined in Table1), weed control (herbicide applica-
tion), prescribed fire, soil disturbance (e.g., chaining,
discing), and vegetation disturbance (e.g., mowing,
hand cutting juniper). For each seeding treatment, we
used species information, when documented, to deter-
mine whether a seed mix contained Artemisia spp.,
native grass, or introduced grass seeds (seed spe-
cies type). Because our objective was to evaluate the
influence of management on sagebrush recovery over
time, we examined project and treatment informa-
tion and excluded weed control, prescribed burn, soil
disturbance, and vegetation disturbance treatments
Fig. 1 Map of study area, sagebrush mask (based on LAND-
FIRE Biophysical Settings and Rangeland Condition Moni-
toring Assessment and Projection (RCMAP) ≥ 8% sagebrush
cover; Online Resource 1), sagebrush biome extent (Jeffries
and Finn 2019), and sagebrush management projects organized
by project purpose (grazing, habitat/restoration, hazards, and
wildfire)
Landsc Ecol (2024) 39:184
Page 4 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
aimed at removing sagebrush cover. In our study,
weed control treatments thus focused on herbaceous
species, whereas vegetation disturbance treatments
were largely intended to reduce pinyon pine (Pinus
spp.) or juniper (Juniperus spp.) cover. However,
we acknowledge that we may have included treat-
ments and projects that, while not specifically stating
sagebrush removal as an objective, could discourage
sagebrush growth (e.g., greenstrip, prescribed fire).
We nevertheless retained these treatments and pro-
jects in our analyses to evaluate a range of potential
outcomes from the full suite of treatment types com-
monly employed to manage environmentally diverse
sagebrush-dominated ecosystems. We did not assess
effects of Artemisia spp. seed at species- or subspe-
cies-levels in our analyses, but Artemisia triden-
tata wyomingensis was the most seeded subspecies
(included in 84% of Artemisia seeding or planting
treatments), followed by A. t. vaseyana (20%), and A.
t. tridentata (18%). Any other Artemisia (sub)species
(e.g., A. nova, A. t. parishii) were included in < 5% of
Artemisia seeding or planting treatments. Similarly,
seed mixes of herbaceous species varied dramati-
cally, so we did not assess effects of species mixes
beyond inclusion of native or introduced grasses.
However, common introduced species included Agro-
pyron species (e.g., A. desertorum, A. fragile, A. cris-
tatum), Thinopyrum species (e.g., T. intermedium, T.
ponticum), and Psathyrostachys juncea, while com-
mon native species included Elymus species (e.g.,
E. wawawaiensis, E. lanceolatus, E. elymoides), Poa
secunda, Pseudoroegneria spicata, Achnatherum
hymenoides, and Pascopyrum smithii. Seeding rate
(i.e., kg/ha) may be another important determinant
of seeding success (Ott etal. 2017), but this informa-
tion was only available for a third of sagebrush seed-
ing treatments, prohibiting its inclusion in our analy-
ses. Moreover, where documented, seeding rates did
not appear to differ considerably between seeding
methods.
Remotely sensed vegetation products
We used the Rangeland Condition Monitoring
Assessment and Projection (RCMAP) time series
product (Rigge et al. 2021, 2022) to characterize
percent sagebrush cover as our response, as well as
annual herbaceous and perennial herbaceous cover
as covariates (refer to Sample design below). This
Table 1 Seed treatments overlapping points from a system-
atic sampling grid summarized by treatment method, number
of unique treatments (n), sum total of treatment sizes, range in
treatment completion year, and the number of treatments with
sagebrush (Artemisia spp.), native grass, and (or) introduced
grass in the seeding mix
Land treatment data were compiled from three datasets overlapping our study area in the western United States: the Land Treatment
Digital Library (LTDL, 1917–2019; Pilliod etal. 2019), the Conservation Efforts Database (CED, 2009–2019; Heller et al. 2017),
and Utah’s Watershed Restoration Initiative (UWRI, 2005–2023; https:// wri. utah. gov/ wri/)
Seeding method Definition nTotal Area (km2) Completion Year Artemisia Native grass Intro-
duced
grass
Aerial Seeding conducted by aircraft,
including fixed or rotor wing
839 16,534.5 1985–2020 505 456 365
Drill Ground-based seeding with a seed
drill
435 4482.0 1985–2019 66 269 343
Ground All other ground-based seeding,
including broadcast by ATV
or hand, or unknown ground
seeding
111 343.4 1986–2020 52 68 49
Seedling Seedlings planted by hand or
mechanical (e.g., auger)
77 499.9 1988–2019 50 0 0
Greenstrip Seeding for greenstrips, including
by aerial, drill, or other ground-
based methods
21 71.0 1990–2014 2 2 13
Unknown Seeding by unknown method 48 378.8 1986–2018 20 25 19
Landsc Ecol (2024) 39:184 184
Page 5 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
product estimates fractional components annually
(1985–2021) at 30-m resolution across the western
United States, excluding areas of cultivated crops,
open water, urban areas, major roads, snow, and ice
(i.e., non-rangelands; Rigge etal. 2020, 2022).
Soil climate product
We investigated effects of soil moisture (monthly
average climate conditions 1981–2010; O’Donnell
and Manier 2022b) as candidate covariates (30-m
raster surfaces) for modeling trends in sagebrush
cover. These data reflect estimated soil moisture based
on the Newhall simulation model (Newhall 1972;
Van Wambeke 2000), tracking vertical movement of
water within a soil profile while accounting for soil
properties, potential evapotranspiration, monthly
precipitation, and monthly air temperature. Soil
moisture (mm) captures the total monthly moisture
between 0 and 200 cm in depth (Soil Survey Staff
1999, 2014). Because this soil climate product did
not extend to the southernmost areas of the sagebrush
biome (O’Donnell and Manier 2022a), we excluded 6
projects that lacked soil climate estimates.
Sample design
Burned areas could be large (up to 77,590 ha,
encompassing > 25 million 30-m pixels), and
modeling trends across all individual pixels would not
be practical. We therefore systematically established
grids of sample points across each project based
on the number of points needed to sufficiently
characterize variability in sagebrush cover within each
project, restricted to areas of sagebrush-dominated
ecosystems (detailed description of sampling grid
establishment process in Online Resource 1 and
Fig.S1 therein). Due to overlap of multiple wildfire
and non-wildfire projects, we determined which grid
points were within 85 m of each other and retained
the grid point with the earliest fire, disturbance, or
project start year (hereafter, initial project), allowing
analysis of subsequent treatments over time. From
each sampling grid point (consisting of 3 × 3, 30-m
pixels), we extracted median sagebrush cover
each year (1986–2021). To characterize treatment
effects on sagebrush cover over time, we identified
treatment polygons that overlapped most of the 3 × 3
pixels centered at each grid point. Each grid point
time series was assigned “0” until the year before
completion of the overlapped treatment and “1”
thereafter. We buffered linear treatments by 15m to
correspond with the 30-m resolution of other spatial
datasets used in this study and to allow for slight
spatial errors when intersecting grid points. Finally,
recurrence of fire could reset seeding treatments
(Pilliod et al. 2021), so Artemisia spp. seeding (but
not seeding method) at grid points was reset to “0”
following fire in a given year, but could be assigned
“1” with subsequent treatments containing Artemisia
spp.
We extracted additional covariates that may affect
trends in sagebrush cover over space and time. At
each grid point, we determined median sagebrush
(Eiswerth etal. 2009; Germino etal. 2018), perennial
herbaceous (Germino et al. 2018; Davidson et al.
2019), and annual herbaceous cover (Ott etal. 2016;
Davidson et al. 2019) in the year before projects
began (initial disturbance or project initiation). We
also calculated mean soil moisture (annual, spring
[March–May], and fall [September–November];
O’Donnell and Manier 2022a). We estimated mean
topographic position index (TPI), a measure of
topographic position relative to the surrounding
landscape (O’Donnell et al. 2022), and slope (%;
U.S. Geological Survey 2005). TPI was calculated
using a 3 × 3 window of 10-m digital elevation
model data and resampled to 30-m resolution. We
used Landsat Burned Area (Hawbaker et al. 2020)
to determine the cumulative number of burns (sum_
burn, 1984–2021; Davies et al. 2012; Mahood and
Balch 2019; Wood et al. 2019) for each grid point
and year. Given challenges in estimating sagebrush
cover immediately following wildfire (Applestein and
Germino 2021; M. Rigge, U.S. Geological Survey,
written communication, October 21, 2022), we
excluded grid point-year samples 0–2years after fire.
Similarly, we excluded the grid points and years with
a treatment year (year 0) for vegetation disturbance,
soil disturbance, and (or) prescribed fire (but we
retained these treatments for grid points in treatment
year > 0).
Statistical analyses
We used generalized additive models (GAM; Wood
2017) to evaluate trends in sagebrush cover from
Landsc Ecol (2024) 39:184
Page 6 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
each grid point i in year t (yi,t) as outcomes of a zero-
inflated Poisson distribution with the ‘ziplss’ model
family from the mgcv package (v. 1.8-42; Wood
2023) in R (v. 4.2.1; R Core Team 2022). In prelimi-
nary models using a Poisson distribution, we detected
a greater frequency of zeros (i.e., no sagebrush) than
would be expected from a Poisson distribution (Har-
tig 2022). In this system, disturbances such as wildfire
and exotic grass invasion can create alternative states
where sagebrush absence is common (Chambers etal.
2017c; Barker etal. 2019). We therefore modeled yi,t
conditional on mean abundance (μi,t) and probability
of presence (pi,t): yi,t ~ Poisson(μi,t × pi,t). For the prob-
ability of presence, we fit covariates with a cloglog
link function including sagebrush (
sagei,−1
), annual
herbaceous (
annugi,−1
), and perennial herbaceous
(
peregi,−1
) cover in the year before project initiation
(i.e., before first fire or initial project grid points),
cumulative number of burns (
sum_burni,t
), and soil
moisture (
soili
). Vegetation cover and soil moisture
were therefore static in time, indicating the gen-
eral site condition for sagebrush presence, whereas
cumulative number of burns represented a dynamic
condition depending on occurrence of repeated wild-
fire (Mahood and Balch 2019). We also specified an
interaction between perennial herbaceous cover and
soil moisture because effects of native grass cover
can shift from facilitative to competitive depending
on soil climatic conditions (Chambers et al. 2014b;
Roundy and Chambers 2021). Additionally, we
accounted for spatial variation among grid point loca-
tions using a Duchon spline (Miller and Wood 2014)
for easting and northing (
f
(
easting
i
, northing
i)
) with
the upper limit of basis dimensions for smooth terms
(k) set at 225:
For all GAM smoothing terms, we determined the
upper limit of basis dimensions (k) using a simula-
tion-based check, increasing k until effective degrees
of freedom were less than k and statistical conclu-
sions from the model were not substantially altered
(Wood 2017).
We fit parameters to the abundance model
including J covariates (Xi,t) for slope, TPI, soil
moisture, purpose (wildfire, grazing, habitat/
cloglog(p
i,t
)=α
0
+α
1
sage
i
+α
2
annug
i
+α
3peregi+α
4soili+α
5peregi×soili
+α
6sum_burni,t+
f
(eastingi, northingi).
restoration, hazards), seeding method (aerial, drill,
ground, seedling planting, greenstrip, unknown), seed
species type (Artemisia spp., native grass, introduced
grass), soil disturbance, vegetation disturbance, weed
control, and prescribed fire. Because soil moisture
availability and perennial grasses can jointly influence
success of Artemisia seeding and transplanting
(Minnick and Alward 2012; Roundy and Chambers
2021), we included two-way interactions between
soil moisture and treatment type and between soil
moisture and seed species type (e.g., aerial seeding
of native grasses), as well as between soil moisture
and perennial herbaceous cover in the preceding
year. Additionally, we specified two-way interactions
between the three seed species types to test whether
their co-occurrence in a seed mix affected subsequent
trends in sagebrush cover (e.g., Artemisia spp. and
native grass; Reichenberger and Pyke 1990; Germino
et al. 2018). We specified density terms for the
natural log of sagebrush, perennial herbaceous, and
annual herbaceous cover in t − 1 and an offset for the
natural log of sagebrush cover in t − 1 to model the
annual rate of change from t − 1 to t (Monroe etal.
2020). We again accounted for additional spatio-
temporal variation among grid points using a Duchon
spline for easting and northing (with k = 196 knots),
and allowed this global smoother to vary by year
( fYEAR
i,t(
eastingi, northingi
)
, with k = 14, 14, and 35
for easting, northing, and year, respectively; Pedersen
etal. 2019):
Except when transformed with a natural log, we
standardized all continuous covariates (subtract the
sample mean, divide by the sample standard devia-
tion). We fit separate models with either spring, fall,
or annual soil moisture covariates and compared sup-
port for soil moisture season using Akaike’s Informa-
tion Criterion (AIC; Akaike 1973). We report addi-
tional methods and results of model fit and accuracy
evaluation in Online Resource 1.
log(μ
i,t
)=β
0
+β
1:J
X
i,t,1:J
+f(easting
i
, northing
i
)
+fYEARi,t(eastingi, northingi)
+
offset(ln[sagei,t−1
+
1]).
Landsc Ecol (2024) 39:184 184
Page 7 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
Recovery projections
To evaluate the implications of our final model for
sagebrush recovery across the sagebrush biome,
we created three scenarios to compare treatment
effects on sagebrush trends following wildfire, an
important disturbance in the sagebrush biome: no
treatment (i.e., natural recovery; scenario 1), aerial
seeding with Artemisia spp. in the seed mix (scenario
2), and drill seeding with Artemisia spp. in the
seed mix (scenario 3). We began with 2021 spatial
layers for sagebrush, perennial herbaceous, and
annual herbaceous cover from RCMAP, cumulative
number of burns (1984–2021), and static layers for
soil moisture, slope, and TPI. We aggregated all
30-m layers to 300m for predictions, either by the
medians for vegetation cover and cumulative number
of burns or by the means for soil moisture, slope,
and TPI, so that simulations could be completed
in a reasonable length of time. In all scenarios,
we assumed an initial cover of sagebrush of 1% at
t = 0, we added 1 to the cumulative burn layer from
2021, and we used 2021 layers for cover before
project initiation. As a threshold for “recovery”, we
used the 95th percentile of percent sagebrush cover
for each 300-m pixel across the 1985–2021 time
series (Monroe et al. 2022). We then used our best-
ranked model to sequentially simulate percent cover
of sagebrush in year t at each pixel, conditional on
t − 1. We repeated this annually for t = 1–30. While
sagebrush-dominated ecosystems can sometimes
recover within 30years (Lesica etal. 2007; Nelson
et al. 2014), sagebrush often requires > 30 years to
attain pre-fire canopy cover (Wambolt et al. 2001;
Baker 2006; Shinneman and McIlroy 2016); however,
we chose this timeframe to align with our dataset’s
length and to reduce uncertainty from a changing
climate relative to longer projections. When a pixel
in a given scenario exceeded its recovery threshold
in a given year (t = 1–30), we halted simulation for
that pixel and fixed sagebrush cover at the recovery
threshold. For pixels that did not recover by t = 30,
we recorded predicted percent cover in year 30.
Since we were simulating sagebrush cover into the
future, we excluded the global smoother varying by
year (
fYEAR
i,t(
eastingi, northingi
)
). We repeated this
sequence of simulations n = 100 times, and in each
iteration, we used a random sample from the model
parameters’ posterior distributions as estimated with a
Metropolis Hastings sampler (Wood 2023). We used
default random walk settings (40 degrees of freedom
for static multivariate t proposal, 0.25 random-walk
scale), discarding the first 2000 iterations and saving
100 posterior samples after thinning by 10. Applying
posterior draws from parameter estimates to predict
sagebrush cover for each pixel incorporated parameter
uncertainty in our simulations. Across simulation
iterations for each scenario, we determined the
proportion of simulations where each pixel reached
the recovery threshold (recovery probability; P). We
also summarized percent cover for each pixel across
simulation iterations by its median and SD and
calculated the difference in recovery probability and
the percent change in sagebrush cover between the
two seeding treatments and the no treatment scenario
by year 30 (median and SD).
Although the 95th percentile of sagebrush cover
provides a reasonable threshold for recovery of
sagebrush based on the ecological potential of each
pixel, we acknowledge that other thresholds may be
considered depending on management objectives (but
refer to Smith et al. 2020). We therefore compared
the proportion of the study area that reached different
recovery probabilities (P ≥ 0.7, P ≥ 0.8, P ≥ 0.9, or
P = 1) relative to the 95th percentile of sagebrush
cover across the time series. We also evaluated
the proportion of the study area where median
sagebrush cover (across simulation iterations for a
pixel) met or exceeded the 95th percentile as well
as uniform thresholds of sagebrush cover, including
5%, 10%, and 15%. The uniform thresholds may
indicate relative rates of recovery after 30 years but
do not consider local potential for sagebrush cover.
To facilitate interpretation of recovery to uniform
thresholds, we also calculated the proportion of the
study area where the 95th percentile of sagebrush
cover met or exceeded each of these thresholds.
To further interpret simulation results, we took
a random 1% sample of pixels from each scenario
(n = 96,914). We then fit a GAM with percent change
in sagebrush cover from either seeding treatments as
the response to environmental covariates at each pixel
(2021 sagebrush cover, 2021 perennial herbaceous
cover, 2021 annual herbaceous cover, cumulative
number of burns, slope, and TPI), including a Duchon
spline for location (k = 400).
Landsc Ecol (2024) 39:184
Page 8 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
Results
Our final sample included 1,140 initial projects
(projects with the earliest fire, disturbance, or project
start year), including 12,215 grid points (Fig. 1) and
173,933 point-by-year samples across 1987–2021
(1986 was used for t − 1 covariates). We estimated
an average 2.2 treatments (seedings, weed control,
soil disturbance, vegetation disturbance, prescribed
burn) overlapping one or more grid points per
initial project. The most common purpose of initial
projects was related to wildfire (n = 823), followed by
habitat/restoration (n = 153), hazards (n = 146), and
grazing (n = 18). Our sample included 1,531 seeding
treatments overlapping one or more grid points,
with most treatments from aerial and drill seeding,
followed by ground seeding (non-drill or unknown
ground seeding), seedling planting, unknown
seeding method, and greenstrip (Table 1). Aerial
seedings included Artemisia spp. in seed mixes more
frequently than native or introduced grass, whereas
Artemisia spp. was less common in drill seedings
than native or introduced grass seed types. Native
and introduced grass were absent from seedling
treatments and Artemisia spp. and native grass seed
were seldom used in greenstrips. Other treatments
included weed control (n = 375), soil disturbance
(n = 353), vegetation disturbance (n = 191), and
prescribed fire (n = 79). The range in years since
initial projects began was 1–36 (median = 11 years)
and the median number of wildfires at a given grid
point was 1 (range = 0, 9 burns). The best supported
model was fit with spring soil moisture (Online
Resource 1: Table S2), and hereafter we used
this model for interpretation and projections. For
additional results of model evaluation, see Online
Resource 1.
The best supported model’s zero-inflation com-
ponent indicated probability of sagebrush presence
increased with greater soil moisture and declined
with cumulative number of burns (Table2). Based on
vegetation estimates before an initial project began,
probability of sagebrush presence increased with
greater pre-project sagebrush cover, declined with
greater annual herbaceous cover, and declined with
greater perennial herbaceous cover, particularly under
low soil moisture conditions.
Conditional on presence, increases in sagebrush
cover were greater when project purpose was related
Table 2 Mean and standard error (SE) of model coefficient
estimates for the top-ranked model (spring soil moisture) fit
to remote sensing products from the Rangeland Condition
Monitoring Assessment and Projection (RCMAP), with coef-
ficients reported separately for the probability of presence and
for abundance (trends in sagebrush cover in the western United
States)
Mean SE
Coefficient (presence)
Intercept 0.093 0.006
 Sage−1
a0.269 0.005
 Annug−1
b− 0.285 0.007
 Pereg−1
c− 0.209 0.006
Soil moisture (spring) 0.073 0.005
 Pereg−1 × soil moisture 0.040 0.004
Cumulative burns − 0.103 0.005
Coefficient (abundance)
Intercept − 0.503 0.014
Soil moisture (spring) 0.007 0.007
ln(saget−1 + 1) 0.126 0.003
ln(annugt−1 + 1) − 0.034 0.002
ln(peregt−1 + 1) 0.017 0.003
ln(peregt−1 + 1) × soil moisture 0.004 0.002
Purpose (grazing) − 0.041 0.010
Purpose (habitat/restoration) − 0.020 0.006
Purpose (hazards) − 0.004 0.007
Slope 0.011 0.002
Topographic position index − 0.001 0.001
Aerial seeding − 0.010 0.010
Drill seeding − 0.020 0.016
Ground seeding 0.016 0.011
Seedling planting 0.028 0.023
Unknown seeding method − 0.014 0.016
Artemisia spp. 0.089 0.017
Native grass 0.013 0.022
Introduced grass 0.001 0.024
Aerial seeding × Soil moisture − 0.011 0.005
Drill seeding × Soil moisture − 0.010 0.008
Ground seeding × Soil moisture 0.000 0.006
Seedling planting × Soil moisture − 0.031 0.014
Unknown seeding method × Soil moisture − 0.008 0.007
Artemisia spp. × Soil moisture 0.004 0.005
Native g rass × Soil moisture − 0.010 0.006
Introduced grass × Soil moisture 0.026 0.006
Aerial seeding × Artemisia spp. − 0.060 0.015
Drill seeding × Artemisia spp. 0.005 0.015
Ground seeding × Artemisia spp. − 0.043 0.016
Seedling planting × Artemisia spp. − 0.061 0.035
Unknown seeding method × Artemisia spp. − 0.037 0.019
Landsc Ecol (2024) 39:184 184
Page 9 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
to wildfire rather than grazing or habitat/restoration
(Table 2). We also estimated increases in sagebrush
cover (in year t) as a function of greater slope and
sagebrush cover (in year t − 1) and decreases as
a function of greenstrips and annual herbaceous
cover (t − 1; Table 2). Sagebrush cover increased
with greater perennial herbaceous cover (t − 1),
particularly among high soil moisture sites.
We estimated an overall positive effect of Artemi-
sia spp. in seeding treatments (Table 2), but impli-
cations for trends in sagebrush cover, relative to no
treatment, depended on seeding method and soil
moisture (Fig. 2). Aerial seeding, drill seeding, and
seedling planting Artemisia spp. (without co-seeding
native or introduced grass) increased sagebrush cover
under low, but not high, soil moisture conditions,
whereas ground seeding (non-drill or unknown seed-
ing) Artemisia spp. had consistently positive effects
across the soil moisture gradient. Co-seeding Arte-
misia spp. with introduced grass increased sagebrush
cover with aerial, drill, and ground seeding under
higher soil moisture conditions (Fig. 2). Ground
seeding introduced grass without Artemisia spp. in
seed mixes increased sagebrush cover in higher soil
moisture sites, whereas drill seeding introduced grass
reduced sagebrush cover in lower soil moisture sites.
Aerial and drill seeding native grasses with or with-
out Artemisia spp. in seed mixes increased sagebrush
cover in low but not high soil moisture sites. Among
other treatments, interactions with soil moisture were
supported for soil disturbance and prescribed fire
(Table2), with declines in sagebrush cover following
soil disturbance or prescribed fire treatments under
high soil moisture conditions relative to no treatment
(Online Resource 1: Fig.S4).
Sagebrush recovery projections
Based on environmental conditions in 2021
(Fig.3a–e), model estimates indicated that an addi-
tional wildfire would result in probabilities of sub-
sequent sagebrush presence ranging from near 1.0
across much of Colorado and Wyoming to < 0.2 in
the Great Basin, Montana, and Washington (Fig.3f).
Without treatment, we found a small proportion
of the study area could be expected to reach the
95th percentile of sagebrush cover within 30years
(< 15% of pixels reaching the recovery threshold
with P ≥ 0.7; Fig. 4a; Table S3), with more areas
likely to recover in Wyoming and Colorado than
in Nevada and Montana (Fig. 3f). Relative to no
treatment, more of the landscape was predicted to
recover with aerial (up to 16% with P = 0.7; Fig.4b)
and drill (up to 25%; Fig.4c; TableS3) seeding Arte-
misia spp., with substantially greater recovery prob-
ability from drill seeding Artemisia spp. occurring
locally (Fig. 4d–f). Median sagebrush cover after
30years met the 95th percentile recovery threshold
across up to 28% of the study area (with drill seed-
ing; TableS4). Less of the landscape exceeded uni-
form recovery thresholds for sagebrush cover (i.e.,
5%, 10%, or 15%), with as little as 2% of the study
area recovering to 15% sagebrush cover without
treatment and up to 23% of the study area recovering
to 5% cover with drill seeding (TableS4). For refer-
ence, we also report the proportion of the landscape
where the 95th percentile of sagebrush cover met or
Table 2 (continued)
Mean SE
Aerial seeding × Native g rass 0.031 0.022
Drill seeding × Native grass 0.023 0.022
Ground seeding × Native grass − 0.049 0.025
Unknown seeding method × Native grass 0.003 0.029
Aerial seeding × Introduced grass 0.004 0.020
Drill seeding × Introduced grass − 0.010 0.021
Ground seeding × Introduced grass 0.055 0.024
Unknown seeding method × Introduced
grass
0.011 0.027
Artemisia spp. × Native g rass − 0.041 0.014
Artemisia spp. × Introduced grass 0.020 0.012
Native g rass × Introduced g rass − 0.033 0.015
Greenstrip − 0.234 0.078
Soil disturbance − 0.008 0.006
Vegetation disturbance 0.004 0.007
Weed control 0.005 0.006
Prescribed fire − 0.006 0.011
Soil disturbance × soil moisture − 0.021 0.005
Vegetation disturbance × soil moisture − 0.007 0.005
Weed control × soil moisture 0.007 0.005
Prescribed fire × soil moisture − 0.016 0.007
a Percent sagebrush cover in the year before project initiation
(− 1) or in the previous year (t − 1)
b Percent annual herbaceous cover in the year before project
initiation (− 1) or in the previous year (t − 1)
c Percent perennial herbaceous cover in the year before project
initiation (− 1) or in the previous year (t − 1)
Landsc Ecol (2024) 39:184
Page 10 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
exceeded uniform thresholds in Online Resource 1:
TableS4.
At 30 years following wildfire without treat-
ment, sagebrush cover estimates followed similar
spatial patterns as observed for recovery probabil-
ity (Fig. 5a). We identified areas where sagebrush
cover declined or increased with aerial and drill
seeding Artemisia spp., relative to no treatment
(Figs. 5b, c). The range in percent change com-
pared with the no treatment scenario was greater for
drill (0% to 437% change in sagebrush cover) than
aerial seeding (– 28% to 95%). Increases in sage-
brush cover also varied regionally, ranging from up
to 95% increase from aerial seeding Artemisia spp.
in Oregon (Fig. 5b) to a 437% increase from drill
seeding Artemisia spp. in California (Fig. 5c). We
provide uncertainty maps associated with Fig.5 in
Online Resource 1: Fig.S5. As with recovery prob-
ability, we observed important local differences in
outcomes from the two seeding methods (Fig. 5e,
f) relative to no treatment (Fig. 5d). Examining
environmental conditions associated with aerial
and drill seeding (Online Resource 1: Table S5)
revealed differences in percent change increased in
areas with more burns and greater sagebrush cover
before disturbance and declined (were more simi-
lar to no treatment) with greater annual herbaceous
cover before disturbance. Differences from no treat-
ment also increased for aerial and drill seedings
occurring in areas with less spring soil moisture
Fig. 2 Contrast plots (
x
and 95% confidence interval
(CI)) for predicted annual
change in sagebrush cover
(%) across a gradient of
relatively low, intermediate,
and high spring soil mois-
ture (mm) and by seeding
method and seed species,
compared to no treatment.
Estimates excluded spatio-
temporal smoothing terms
for abundance (trends) and
occurrence probability and
assumed initial sagebrush
cover = 1% with pur pose
related to wildfire response
Landsc Ecol (2024) 39:184 184
Page 11 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
and perennial herbaceous cover before disturbance.
Treatment benefits (differences from no treatment)
increased with greater slope for aerial seedings, but
not drill seedings.
Discussion
Efficacy of restoration treatments
Compared to untreated sites, we generally found the
greatest effects of seeding Artemisia spp. without
co-seeding native or introduced grasses at dryer
sites (Fig.2). Higher soil moisture sites have greater
resistance to invasion and resilience from disturbance
(Chambers et al. 2014b; Wood et al. 2019), which
could diminish differences between sites with and
without seeding treatment. This pattern has also
been observed in one of the largest studies using
field data, where sagebrush occupancy was 3.3
times greater in seeded areas with low soil moisture,
but the seeding effect diminished as soil moisture
increased (Arkle et al. 2022). Seedings including
native grasses responded similarly, whereas wetter
sites benefited from seed mixes with both Artemisia
spp. and introduced grasses (Fig. 2). Perennial
grasses are typically seeded to stabilize disturbed
sites (i.e., prevent erosion and establishment of
invasive annuals; Ott etal. 2003; Davies etal. 2014;
Davies and Boyd 2021), with introduced species
such as crested wheatgrass (Agropyron cristatum)
often preferred because they establish more rapidly
than native species (Davies and Boyd 2021).
Introduced grasses can compete with naturally
recruiting sagebrush seedlings when resources are
limited (Cook and Lewis 1963; Shaw et al. 2005;
Fig. 3 Baseline datasets for
simulating trends in sage-
brush cover following wild-
fire, including spring soil
moisture1 1981–2010 (a),
sagebrush cover2 in 2021
(b), perennial herbaceous
cover2 in 2021 (c), annual
herbaceous cover2 in 2021
(d), and cumulative number
of burns 1984–20213 (plus
1 for wildfire scenario; (e).
From these baseline layers,
we derived the probability
of sagebrush presence (p)
for use in simulations (f).
Extent of the sagebrush
biome (Jeffries and Finn
2019) is outlined in green.
1O’Donnell and Manier
(2022b), 2Rigge etal.
(2022), 3Hawbaker etal.
(2020)
Landsc Ecol (2024) 39:184
Page 12 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
Gunnell et al. 2010), but wetter sites could mitigate
their detrimental impacts. Indeed, we found drill
seeding introduced grasses without Artemisia reduced
sagebrush cover under dryer soil conditions.
Among seeding treatments, we found greater
recovery probability and sagebrush cover from drill
seeding Artemisia spp. compared with aerial seeding
(Fig. 4). Drill or ground-based broadcast seeding of
Artemisia are recommended as more effective than
aerial seeding, especially when aerial seeding is
conducted without covering Artemisia spp. seeds
to promote establishment (Lambert 2005; Shaw
et al. 2005; Knutson et al. 2014). Previous studies
documented increases in sagebrush occurrence
within the first several years following aerial seeding
(Eiswerth etal. 2009; Germino etal. 2018; Arkle etal.
2022), yet long-term implications are less clear and
may depend on multiple factors at planting, including
weather and timing of seedings (Knutson etal. 2014;
Shriver etal. 2018; O’Connor etal. 2020; Applestein
etal. 2021). Although we used a static measure of soil
moisture availability rather than temporally varying
measures of weather, contingent weather factors may
only marginally increase predictive performance
over long-term averages (Applestein et al. 2021;
Simler‐Williamson etal. 2022). Limited increases in
sagebrush cover from aerial seeding relative to drill or
other ground seedings also could be a function of the
data and uncertainty of where seeds were distributed.
Aerial seeding can cover a large area and, depending
on weather conditions, seed applications may be
subject to drift (Ott et al. 2019). We also did not
include seeding rates or seed source in our models,
which may affect seeding success (Thompson etal.
2006; Ott et al. 2017, 2019; Germino et al. 2018).
Additionally, when seed supplies are low, crews may
increase seeding rates but apply seed at every other
pass, creating a striped pattern of sagebrush seed
application (or see alternatives in Lambert 2005;
Grant-Hoffman and Plank 2021). If a treatment
polygon covers the entire planned treatment area
but received only half the application, these data
could increase uncertainty in estimates and reduce
differences with untreated areas.
Fig. 4 Probability of recovery (relative to the 95th percentile
of sagebrush cover, 1985–2021) 30years after a wildfire (a),
based on 100 simulations of model predictions and the refer-
ence scenario (no treatment). We also report the difference in
probability (Δ) of recovering between aerial seeding with Arte-
misia spp. and the reference (b) and between drill seeding with
Artemisia spp. and the reference (c). Areas in gray lacked pre-
dictions. We present maps at a smaller spatial extent (dashed
lines) for probability for sagebrush cover without treatment
(d), and the difference in probability of recovering between
aerial seeding with Artemisia spp. and the reference (e) and
between drill seeding with Artemisia spp. and the reference (f).
Areas in gray lacked predictions
Landsc Ecol (2024) 39:184 184
Page 13 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
In contrast to aerial and drill seeding Artemisia
spp., ground seedings (non-drill or unknown ground
seeding) increased sagebrush recovery across soil
climates (Fig. 2), but apparent differences in out-
comes between drill and ground seeding methods in
our analyses should be interpreted with caution. Drill
seeding is considered an effective method of estab-
lishing Artemisia spp. but special alterations must
be made to equipment for proper application of these
smaller seeds (Stevens and Monsen 2004; Lambert
2005). While these methods use drill seeding equip-
ment, they are sometimes described as broadcast
seedings; similarly, broadcast seeding of Artemi-
sia often occurs in the wake of drill seeding large-
seeded species (Stevens and Monsen 2004; Shaw
etal. 2005). Together, these practices generate uncer-
tainty about how Artemisia seedings were applied
and documented in treatment databases. Furthermore,
documentation of seeding methods sometimes lacked
details, and treatments labeled as ground seeding in
the LTDL include treatments applied from the ground
but with unknown methods beyond that (i.e., poten-
tially including drill seedings).
We found the most negative effects of
prescribed fire and soil disturbance on sagebrush
recovery occurred under higher soil moisture
conditions. While these results appear to contradict
expectations based on long-term field studies that
disturbance-mimicking treatments are least risky
under more resilient site conditions (Davies and
Bates 2017; Chambers etal. 2021), our findings are
relative to trends in natural recovery of sagebrush
after disturbance. Since sagebrush recovers more
rapidly at high soil moisture sites, treatments that
again disturb vegetation could impede rather than
facilitate recovery. Likewise, these treatments
may have little influence on the limited recovery
occurring at dryer sites. Moreover, treatments
that potentially reduce sagebrush cover in the
short term also could facilitate spread of exotic
annuals (Roundy et al. 2018; Shinneman et al.
2023), thereby limiting increases in sagebrush
cover relative to untreated sites. Similarly, we
Fig. 5 Mean sagebrush cover (%) 30 years after a wildfire
(a), based on 100 simulations from model predictions and the
reference scenario (no treatment). We also report the percent
change in mean sagebrush cover from simulations of aerial
seeding with Artemisia spp. relative to the reference (b) and
from simulations of drill seeding with Artemisia spp. relative
to the reference (c). For sub-figures a–c we present maps at a
smaller spatial extent (dashed lines) for sagebrush cover with
no treatment (d), and percent change in mean sagebrush cover
between simulations of aerial seeding with Artemisia spp. and
the reference (e) and between simulations of drill seeding with
Artemisia spp. and the reference (f). Areas in gray lacked pre-
dictions
Landsc Ecol (2024) 39:184
Page 14 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
estimated a negative effect of greenstrip seeding
on subsequent sagebrush cover, as expected for a
treatment intended to reduce spread of wildfire by
replacing fire-prone vegetation with introduced
herbaceous species (Pellant 1990, 1994). Studies
evaluating benefits and effectiveness of fuel breaks
while accounting for their direct ecological effects
across these landscapes are needed and ongoing
(Shinneman etal. 2019).
Environmental contributors to sagebrush cover
We evaluated influence of environmental factors on
sagebrush cover with two joint processes, includ-
ing probability of sagebrush presence, and trends
in sagebrush cover given presence. Consistent with
previous field studies, we found sagebrush presence
was positively associated with the amount of initial
(pre-project) sagebrush cover (Eiswerth et al. 2009;
Arkle etal. 2022) and negatively with initial annual
herbaceous cover and cumulative number of burns
(Eiswerth et al. 2009; Mahood and Balch 2019).
Interactions between wildfire and invasive annual
grasses are common in the study region and a leading
driver of sagebrush loss (Knutson etal. 2014; Barker
etal. 2019). In our study, their effects on sagebrush
presence were apparent across areas of the Great
Basin (Fig.3), limiting sagebrush recovery irrespec-
tive of treatment (Fig.4). Where sagebrush persisted
or re-established after disturbance, sagebrush and
annual herbaceous cover in the previous year influ-
enced trends in sagebrush cover similar to probability
of sagebrush presence, suggesting these vegetation
components continue to affect recovery after distur-
bance. We also estimated sagebrush cover increasing
with slope, in contrast with previous studies finding
ambivalent (Johnson and Payne 1968) or negative
(Davidson et al. 2019) effects. Topographic charac-
teristics affect microclimate and growing conditions
(Burke et al. 1989; Rodhouse et al. 2014; Mata-
González etal. 2018), and therefore the soil moisture
covariate in our model may alter the importance of
topography and (or) highlight factors other than soil
moisture associated with topography such as large
herbivores avoiding steep slopes (Clark et al. 2014;
Anthony and Germino 2022).
The neutral to negative effect of perennial
herbaceous cover on sagebrush presence was
somewhat surprising, although large field studies
have found the relationship between perennial
grasses and sagebrush establishment to be parabolic,
where low and high cover of perennial grass are
detrimental but moderate cover is facilitative
(Chambers et al. 2007, 2014a; Arkle et al. 2022).
Furthermore, remotely sensed estimates of perennial
herbaceous cover may include misclassifications of
annual herbaceous cover (M. Rigge, U.S. Geological
Survey, written communication, January 19, 2024),
or introduced perennial grasses and forbs that could
be more competitive with sagebrush seedlings than
native bunchgrasses, especially where resources
are limited (Booth etal. 2010; McAdoo etal. 2013;
Chambers et al. 2017b; Davies et al. 2020a). In
contrast to sagebrush presence, we found a neutral
to positive effect of perennial herbaceous cover on
sagebrush cover trends, especially under moister
conditions. Once established, greater rooting depths
of older sagebrush plants may reduce the competitive
effects from perennial grasses, including introduced
species (Gunnell et al. 2010; Davies et al. 2020b).
Moreover, positive effects on sagebrush cover in areas
with higher spring soil moisture may reflect the role
that perennial herbaceous cover plays in resisting
invasion by annual grasses such as Bromus tectorum
(Chambers et al. 2007, 2014a; Davies and Johnson
2017).
Sagebrush recovery projections
Although long-term soil moisture averages
(1981–2010) mostly overlapped our study period (fire
data: 1984–2021; sagebrush cover data: 1985–2021),
these conditions are unlikely to persist due to climate
change (Berg etal. 2017; Williams etal. 2020). Our
projections of sagebrush recovery could thus be
improved and expanded if estimates of future soil
climate conditions become available. We also note
that over 26% of post-fire treatments have re-burned
one or multiple times within the first 20 years since
initial treatment, and that under current fire regimes
17% of treated hectares are unlikely to reach this
30-year mark (Pilliod etal. 2021). Additionally, our
analyses were based on other modeled datasets and
thus inherit their limitations. For example, RCMAP
tends to overestimate sagebrush cover at low values
and underestimate cover at high values (Rigge etal.
2020; Applestein and Germino 2021), and is more
Landsc Ecol (2024) 39:184 184
Page 15 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
likely to misclassify vegetation in wetter areas due
to lower contrast in their spectral signals (M. Rigge,
U.S. Geological Survey, written communication,
January 19, 2024). For these reasons, we emphasize
relative differences among treatments over space and
time, rather than absolute measures. Nevertheless,
our predictions of sagebrush recovery show patterns
broadly similar to recent maps of sagebrush
ecosystem resistance and resilience (Chambers etal.
2023), with some important divergences such as in
southern Montana (Fig. 5), where we also had fewer
data (Fig.1).
Irrespective of the recovery threshold, we found
seeding treatments, particularly drill seeding
Artemisia spp., increased the rate of sagebrush
recovery over a larger area than natural recovery
(no treatment). Our simulations also indicated
environmental conditions where restoration
treatments could be more effective than natural
recovery. We observed greater benefits from
seeding treatments in dry areas with more repeated
burns, highlighting the ability of seedings to
increase sagebrush cover in areas with frequent
wildfire. However, we also found greater benefits
in areas with more sagebrush cover and less annual
and perennial herbaceous cover before disturbance.
Such areas may possess favorable conditions
for sagebrush recovery (e.g., better seed banks,
microsites; Arkle et al. 2022), yet our simulations
suggested a role for sagebrush seedings to
accelerate recovery relative to no treatment. Areas
with steeper slopes were associated with greater
benefits from aerial seedings than natural recovery,
but not from drill seeding. Topography may present
challenges to implementing drill seedings (Stevens
and Monsen 2004; Pyke et al. 2013), potentially
reducing their efficacy, or their implementation in
the first place, thereby limiting our sample of such
treatments.
We identified areas where sagebrush cover may be
reduced 30 years after aerial seeding Artemisia spp.
relative to natural recovery. This result could be a
function of the observational nature of our analyses,
where treatments may not be applied randomly on the
landscape, and instead resource managers may choose
not to treat sites in better pre-fire condition (Arkle
et al. 2022; Simler-Williamson and Germino 2022).
We attempted to reduce this bias by sampling from
grids established systematically across wildfires and
projects and including covariates for site condition
such as soil moisture and annual herbaceous cover,
but we acknowledge these are not substitutes for
randomized and controlled experiments. However,
there could be instances where a seeding treatment
yields worse outcomes than an untreated counterpart,
such as when using non-local seed sources (Brabec
et al. 2015, 2017; Baughman et al. 2019). Areas
projected to decline from aerial seeding Artemisia
spp. therefore warrant additional study, but at a
minimum, and without additional information, these
areas are not expected to benefit from this treatment
following wildfire.
How one determines “recovery” depends on
objectives and may not necessarily be aligned with
the 95th percentile of sagebrush cover across the
1985–2021 time series used here (Gann etal. 2019).
While recovery was limited under the strictest
probability (e.g., < 15% of the study area recovered
under all simulation iterations when drill seeded, or
P = 1.0), less conservative thresholds (e.g., P = 0.7,
or using median sagebrush cover) nearly doubled
recovery projections (up to 28% of the study area
under drill seeding). Similarly, recovery to uniform
thresholds of 5% sagebrush cover occurred in < 25%
of the study area, but these recovered areas declined
when the uniform threshold was increased to 15%
sagebrush cover. However, uniform thresholds should
be interpreted in the context of local potential for
sagebrush cover. For example, 18% of the study area
had the capacity to support 15% sagebrush cover
(based on the 95th percentile of the time series), and
therefore a 6% recovery to at least 15% sagebrush
cover with drill seeding actually represents recovery
of one third of the area of potentially high sagebrush
cover. Our projections thus identify considerable
opportunities for targeting sagebrush restoration
to improve sagebrush recovery trajectories after
disturbance, potentially providing important habitat
for sagebrush dependent and associated wildlife
species (Pyke et al. 2020). While our predictions
suggest that considerable amounts of habitat could
be restored after 30years, still more of the study area
is not expected to recover within that time frame,
highlighting the importance of protecting sagebrush
habitats in the first place (Arkle etal. 2014; Orning
etal. 2023). Sagebrush protection is especially crucial
for species depending on greater cover of sagebrush
(e.g., > 15%), yet areas restored to lower cover could
Landsc Ecol (2024) 39:184
Page 16 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
be used for specific life stages or resource needs of
some species (e.g., greater sage-grouse brood-rearing
habitat requiring less sagebrush cover than nesting
habitat; Connelly et al. 2000; Hagen etal. 2007) or
provide habitat for wildlife with less stringent cover
requirements.
Conclusions
Preserving the sagebrush biome and its dependent
species requires effective and efficient restoration
of sagebrush cover across a broad geographic
area with a dynamic range of growing conditions.
With millions of dollars spent annually to restore
thousands of hectares of sagebrush across the
biome (see Pilliod et al. 2017; Copeland et al.
2018), increasing threats to sagebrush ecosystems
mean the need to anticipate how these ecosystems
will respond to restoration efforts will only become
more important. The complex relationships we
observed among seeding method, seed species, and
soil moisture preclude anticipating effects of these
variables on sagebrush trends in isolation across this
biome. Our results generally confirmed expectations
from a resilience-based conceptual framework,
where sagebrush recovery and interactions between
plant functional groups are predicted to vary based
on resource availability and life histories (Gómez-
Aparicio 2009; Chambers et al. 2017c). Qualified
with accompanying uncertainty maps (Online
Resource 1: Fig. S5), our recovery projections
could be used to guide strategic conservation
efforts based on resistance and resilience concepts
such as ‘Resist-Accept-Direct’ (Crausbay et al.
2021) and ‘Defend the Core’ (e.g., Doherty et al.
2022). For example, resist or defend actions could
be targeted at sagebrush areas that are crucial for
meeting conservation objectives and unlikely to
recover from wildfire. Where sagebrush recovery
is more likely, our results could be applied to
prioritizing restoration efforts (see Duchardt et al.
2021), differentiating between areas likely to
recover naturally and areas most likely to benefit
from specific restoration actions. We summarize
the implications of our findings for management
decisions as follows:
1. Seeding and seedling planting Artemisia spp.
facilitated sagebrush recovery, particularly in
dryer areas, and ground-based methods for seed-
ing Artemisia spp. were generally more effective
than aerial seeding.
2. Co-seeding native grasses was more likely
to improve sagebrush recovery in dryer sites,
whereas co-seeding introduced grasses may
increase sagebrush cover in moister sites.
However, seeding introduced grasses could be
counter-productive for restoring sagebrush in
dryer sites.
3. Configuration of existing vegetation components
influenced sagebrush establishment and
subsequent cover trends and could therefore
determine success of restoration treatments.
4. Repeated wildfires inhibited sagebrush recovery
but sagebrush seeding treatments could overcome
this effect to some extent, improving trajectories
over natural recovery.
We emphasize that the maps provided here are not
meant to be prescriptive or substitute for local, on-
the-ground knowledge of experts in the field; rather,
they provide a broad perspective of potential out-
comes from different restoration practices. The spatial
and temporal breadth of our study offered a unique
opportunity to evaluate the efficacy of dryland resto-
ration treatments in an imperiled biome and predict
outcomes decades into the future. Although our find-
ings largely concur with those from more recent res-
toration efforts (Arkle et al. 2022), trends measured
here may not reflect more recent advances in restora-
tion practices and technology (Germino etal. 2021),
particularly given it may take decades before long-
term consequences of treatments become apparent in
this system. This recognition highlights the impor-
tance of recording spatio-temporal data on treatment
applications in regional, national, or global databases
(e.g., Global Restore Project https:// globa lrest orepr
oject. com/, RestoreNET https:// dryla ndeco logy. com/
proje cts/ resto renet) and taking advantage of existing
standardized monitoring programs (e.g., Assessment,
Inventory and Monitoring https:// www. blm. gov/ aim),
so that models and predictions may be updated over
time. As restoration, data management, and spa-
tial modeling practices continue to develop, stud-
ies can leverage these important resources to inform
Landsc Ecol (2024) 39:184 184
Page 17 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
restoration planning and delivery across dynamic and
imperiled ecosystems around the world.
Acknowledgements We thank G. Simpson for helpful sug-
gestions on simulated predictions from GAMs, T. Conkling
for reviewing the associated data release, D. Major for review-
ing this manuscript, and the Conservation Efforts Database,
the Land Treatment Digital Library, and Utah’s Watershed
Restoration Initiative for providing land treatment data. This
work was supported by the U.S. Geological Survey, the U.S.
Department of the Interior Bureau of Land Management, and
the Wyoming Landscape Conservation Initiative. Any use of
trade, firm, or product names is for descriptive purposes only
and does not imply endorsement by the U.S. Government.
Author contributions BCT compiled spatial layers,
processed treatment data, contributed to study design, and led
development of the manuscript. APM and CLA conceived
the study and acquired project funding. APM initiated data
collection, conducted analyses, and contributed to manuscript
writing. CLA, RSA, PSC, MIJ, JAH, DJM, MSO, DSP,
and JLW each contributed to study design and manuscript
development, review, and revision.
Funding Funding was provided by U.S. Bureau of Land
Management, Wyoming Landscape Conservation Initiative and
U.S. Geological Survey.
Data availability Data and resulting simulation maps are
available in a U.S. Geological Survey data release (Monroe
etal. 2024).
Declarations
Competing interests The authors declare no competing inter-
ests.
Open Access This article is licensed under a Creative
Commons Attribution-NonCommercial-NoDerivatives 4.0
International License, which permits any non-commercial
use, sharing, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original
author(s) and the source, provide a link to the Creative
Commons licence, and indicate if you modified the licensed
material. You do not have permission under this licence to
share adapted material derived from this article or parts of
it. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless
indicated otherwise in a credit line to the material. If material
is not included in the article’s Creative Commons licence and
your intended use is not permitted by statutory regulation or
exceeds the permitted use, you will need to obtain permission
directly from the copyright holder. To view a copy of this
licence, visit http:// creat iveco mmons. org/ licen ses/ by- nc- nd/4.
0/.
References
Akaike H (1973) Information theory and an extension of the
maximum likelihood principle. In: Petran B, Csaaki F
(eds) International symposium on information theory.
Akadeemiai Kiadi, Budapest, pp 267–281
Anthony CR, Germino MJ (2022) Predictive models of selec-
tive cattle use of large, burned landscapes in semiarid
sagebrush-steppe. Rangel Ecol Manag 85:1–8
Applestein C, Germino MJ (2021) Detecting shrub recovery in
sagebrush steppe: comparing Landsat-derived maps with
field data on historical wildfires. Fire Ecol 17(1):5
Applestein C, Caughlin TT, Germino MJ (2021) Weather
effects post-fire recovery of sagebrush-steppe commu-
nities and model transferability among sites. Ecosphere
12(4):e03446
Applestein C, Caughlin TT, Germino MJ (2022) Post-fire seed
dispersal of a wind-dispersed shrub declined with dis-
tance to seed source, yet had high levels of unexplained
variation. AoB Plants 14(6):plac045
Arkle RS, Pilliod DS, Hanser SE etal (2014) Quantifying res-
toration effectiveness using multi-scale habitat models:
implications for sage-grouse in the Great Basin. Eco-
sphere 5(3):31
Arkle RS, Pilliod DS, Germino MJ, Jeffries MI, Welty JL
(2022) Reestablishing a foundational species: limitations
on post-wildfire sagebrush seedling establishment. Eco-
sphere 13(8):e4195
Aronson J, Alexander S (2013) Ecosystem restoration is now
a global priority: time to roll up our sleeves. Restor Ecol
21(3):293–296
Baker WL (2006) Fire and restoration of sagebrush ecosys-
tems. Wildl Soc Bull 34(1):177–185
Balch JK, Bradley BA, D’Antonio CM, Gomez-Dans J (2013)
Introduced annual grass increases regional fire activity
across the arid western USA (1980–2009). Glob Chang
Biol 19(1):173–183
Barker BS, Pilliod DS, Rigge M, Homer CG (2019) Pre-fire
vegetation drives post-fire outcomes in sagebrush eco-
systems: evidence from field and remote sensing data.
Ecosphere 10(11):e02929
Barnard DM, Germino MJ, Arkle RS etal (2019) Soil char-
acteristics are associated with gradients of big sage-
brush canopy structure after disturbance. Ecosphere
10(6):e02780
Baughman OW, Agneray AC, Forister ML etal (2019) Strong
patterns of intraspecific variation and local adaptation
in Great Basin plants revealed through a review of 75
years of experiments. Ecol Evol 9(11):6259–6275
Bell SS, Fonseca MS, Motten LB (1997) Linking restoration
and landscape ecology. Restor Ecol 5(4):318–323
Berg A, Sheffield J, Milly PCD (2017) Divergent surface and
total soil moisture projections under global warming.
Geophys Res Lett 44(1):236–244
Blaisdell JP (1958) Ecological effects of planned burning
of Sagebrush-Grass range on the Upper Snake River
Plains. Technical Bulletin No. 1075. U.S. Department
of Agriculture, Washington
Landsc Ecol (2024) 39:184
Page 18 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
Boer MM, Puigdefábregas J (2005) Assessment of dryland
condition using spatial anomalies of vegetation index
values. Int J Remote Sens 26(18):4045–4065
Booth DT, Samuel EC, Deena T (2010) Aerial assessment of
leafy spurge (Euphorbia esula L.) on Idaho’s Deep Fire
Burn. Native Plants J 11(3):327
Brabec MM, Germino MJ, Shinneman DJ, Pilliod DS, McIl-
roy SK, Arkle RS (2015) Challenges of establishing
big sagebrush (Artemisia tridentata) in rangeland res-
toration: effects of herbicide, mowing, whole-commu-
nity seeding, and sagebrush seed sources. Rangel Ecol
Manage 68(5):432–435
Brabec MM, Germino MJ, Richardson BA, Tulloch A (2017)
Climate adaption and post-fire restoration of a founda-
tional perennial in cold desert: insights from intraspe-
cific variation in response to weather. J Appl Ecol
54(1):293–302
Briske DD, Bestelmeyer BT, Stringham TK, Shaver PL
(2008) Recommendations for development of resil-
ience-based state-and-transition models. Rangel Ecol
Manag 61(4):359–367
Brudvig LA (2011) The restoration of biodiversity: where
has research been and where does it need to go? Am J
Bot 98(3):549–558
Burke IC, Reiners WA, Olson RK (1989) Topographic con-
trol of vegetation in a mountain big sagebrush steppe.
Vegetatio 84(2):77–86
Cardinale BJ, Duffy JE, Gonzalez A et al (2012) Bio-
diversity loss and its impact on humanity. Nature
486(7401):59–67
Chambers JC, Roundy BA, Blank RR, Meyer SE, Whittaker
A (2007) What makes Great Basin sagebrush eco-
systems invasible by Bromus tectorum? Ecol Monogr
77(1):117–145
Chambers JC, Bradley BA, Brown CS etal (2014a) Resilience
to stress and disturbance, and resistance to Bromus tec-
torum L. invasion in cold desert shrublands of Western
North America. Ecosystems 17(2):360–375
Chambers JC, Miller RF, Board DI et al (2014b) Resilience
and resistance of sagebrush ecosystems: implications for
state and transition models and management treatments.
Rangel Ecol Manag 67(5):440–454
Chambers JC, Beck JL, Bradford JB et al (2017a) Science
framework for conservation and restoration of the sage-
brush biome: linking the Department of the Interior’s
Integrated Rangeland Fire Management Strategy to
long-term strategic conservation actions: Part 1. Science
basis and applications. General Technical Report RMRS-
GTR-360. U.S. Department of Agriculture, Forest Ser-
vice, Rocky Mountain Research Station, Fort Collins,
Colorado, p 213
Chambers JC, Board DI, Roundy BA, Weisberg PJ (2017b)
Removal of perennial herbaceous species affects
response of Cold Desert shrublands to fire. J Veg Sci
28(5):975–984
Chambers JC, Maestas JD, Pyke DA, Boyd CS, Pellant M,
Wuenschel A (2017c) Using resilience and resistance
concepts to manage persistent threats to sagebrush eco-
systems and greater sage-grouse. Rangel Ecol Manag
70(2):149–164
Chambers JC, Urza AK, Board DI et al (2021) Sagebrush
recovery patterns after fuel treatments mediated by dis-
turbance type and plant function group interactions. Eco-
sphere 12(4):e03450
Chambers JC, Brown JL, Bradford JB etal (2023) New indi-
cators of ecological resilience and invasion resistance to
support prioritization and management in the sagebrush
biome, United States. Front Ecol Evol 10:1009268
Chaney L, Richardson BA, Germino MJ (2017) Climate
drives adaptive genetic responses associated with sur-
vival in big sagebrush (Artemisia tridentata). Evol Appl
10(4):313–322
Clark PE, Lee J, Ko K etal (2014) Prescribed fire effects on
resource selection by cattle in mesic sagebrush steppe.
Part 1. Spring grazing. J Arid Environ 100–101:78–88
Cook CW, Lewis CE (1963) Competition between big sage-
brush and seeded grasses on foothill ranges in Utah. J
Range Manag 16(5):245–250
Cooke SJ, Bennett JR, Jones HP (2019) We have a long way to
go if we want to realize the promise of the “Decade on
Ecosystem Restoration.” Conserv Sci Pract 1(12):e129
Connelly JW, Schroeder MA, Sands AR, Braun CE (2000)
Guidelines to manage sage grouse populations and their
habitats. Wildl Soc Bull 28(4):967–985
Copeland SM, Munson SM, Pilliod DS, Welty JL, Bradford JB,
Butterfield BJ (2018) Long-term trends in restoration and
associated land treatments in the southwestern United
States. Restor Ecol 26(2):311–322
Crausbay SD, Sofaer HR, Cravens AE etal (2021) A science
agenda to inform natural resource management deci-
sions in an era of ecological transformation. Bioscience
72(1):71–90
Davidson BE, Germino MJ, Richardson B, Barnard DM (2019)
Landscape and organismal factors affecting sagebrush-
seedling transplant survival after megafire restoration.
Restor Ecol 27(5):1008–1020
Davies KW, Bates JD (2017) Restoring big sagebrush after
controlling encroaching western juniper with fire: aspect
and subspecies effects. Restor Ecol 25(1):33–41
Davies KW, Boyd CS (2021) Seeding locally sourced native
compared to introduced bunchgrasses post-wildfire in
frigid Wyoming big sagebrush communities. Restor Ecol
29(5):e13397
Davies KW, Johnson DD (2017) Established perennial veg-
etation provides high resistance to reinvasion by exotic
annual grasses. Rangel Ecol Manag 70(6):748–754
Davies KW, Boyd CS, Beck JL, Bates JD, Svejcar TJ, Gregg
MA (2011) Saving the sagebrush sea: an ecosystem con-
servation plan for big sagebrush plant communities. Biol
Conserv 144(11):2573–2584
Davies GM, Bakker JD, Dettweiler-Robinson E et al (2012)
Trajectories of change in sagebrush steppe vegetation
communities in relation to multiple wildfires. Ecol Appl
22(5):1562–1577
Davies KW, Bates JD, Madsen MD, Nafus AM (2014) Res-
toration of mountain big sagebrush steppe following
prescribed burning to control Western Juniper. Environ
Manag 53(5):1015–1022
Davies KW, Bates JD, Clenet D (2020a) Improving restoration
success through microsite selection: an example with
Landsc Ecol (2024) 39:184 184
Page 19 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
planting sagebrush seedlings after wildfire. Restor Ecol
28(4):859–868
Davies KW, Boyd CS, Bates JD, Hamerlynck EP, Copeland
SM (2020b) Restoration of sagebrush in crested wheat-
grass communities: longer-term evaluation in northern
great basin. Rangel Ecol Manag 73(1):1–8
Doherty K, Theobald DM, Bradford JB et al (2022) A sage-
brush conservation design to proactively restore Ameri-
ca’s sagebrush biome. U.S. Geological Survey Open-File
Report 2022-1081. Reston, p 38
Duchardt CJ, Monroe AP, Heinrichs JA, O’Donnell MS,
Edmunds DR, Aldridge CL (2021) Prioritizing restora-
tion areas to conserve multiple sagebrush-associated
wildlife species. Biol Conserv 260:109212
Eiswerth ME, Krauter K, Swanson SR, Zielinski M (2009)
Post-fire seeding on Wyoming big sagebrush ecological
sites: regression analyses of seeded nonnative and native
species densities. J Environ Manag 90:1320–1325
Gann GD, McDonald T, Walder B et al (2019) International
principles and standards for the practice of ecological
restoration. Restor Ecol 27(S1):S1–S46
Germino MJ, Barnard DM, Davidson BE etal (2018) Thresh-
olds and hotspots for shrub restoration following a het-
erogeneous megafire. Landsc Ecol 33(7):1177–1194
Germino MJ, Brunson MW, Chambers JC etal (2021) Chap-
ter R. Restoration. In: Remington T, Deibert P, Hanser
S, Davis D, Robb L, Welty J (eds) Sagebrush conserva-
tion strategy—challenges to sagebrush conservation.
U.S. Geological Survey Open-File Report 2020–1125, pp
203–222
Gómez-Aparicio L (2009) The role of plant interactions
in the restoration of degraded ecosystems: a meta-
analysis across life-forms and ecosystems. J Ecol
97(6):1202–1214
Grant-Hoffman MN, Plank HL (2021) Practical postfire sage-
brush shrub restoration techniques. Rangel Ecol Manag
74(1):1–8
Gunnell KL, Monaco TA, Call CA, Ransom CV (2010) Seed-
ling interference and niche differentiation between
crested wheatgrass and contrasting native great basin
species. Rangel Ecol Manag 63(4):443–449
Hagen CA, Connelly JW, Schroeder MA (2007) A meta-
analysis of greater sage-grouse Centrocercus uropha-
sianus nesting and brood-rearing habitats. Wildl Biol
13(sp1):42–50
Hansen LD, Farnsworth LK, Itoga NK etal (2008) Two sub-
species and a hybrid of big sagebrush: comparison of
respiration and growth characteristics. J Arid Environ
72(5):643–651
Hardegree SP, Jones TA, Roundy BA, Shaw NL, Monaco TA
(2016) Assessment of range planting as a conservation
practice. Rangel Ecol Manag 69(4):237–247
Hartig F (2022) DHARMA: residual diagnostics for hierarchi-
cal (multi-level/mixed) regression models. R package
version 0.4.5
Hawbaker TJ, Vanderhoof MK, Schmidt GL etal (2020) The
Landsat Burned Area algorithm and products for the
conterminous United States. Remote Sens Environ
244:111801
Heller MM, Welty JL, Wiechman LA (2017) The conservation
efforts database: improving our knowledge of landscape
conservation actions. https:// pubs. usgs. gov/ publi cation/
70188 556
Hobbs RJ, Norton DA (1996) Towards a conceptual framework
for restoration ecology. Restor Ecol 4(2):93–110
Holl KD, Aide TM (2011) When and where to actively restore
ecosystems? For Ecol Manag 261(10):1558–1563
Holl KD, Crone EE, Schultz CB (2003) Landscape restoration:
moving from generalities to methodologies. Bioscience
53:491–502
Holling CS (1973) Resilience and stability of ecological sys-
tems. Annu Rev Ecol Syst 4:1–23
Jeffries MI, Finn SP (2019) The sagebrush biome range extent,
as derived from classified Landsat imagery. U.S. Geolog-
ical Survey data release. https:// doi. org/ 10. 5066/ P950H
8HS
Johnson JR, Payne GF (1968) Sagebrush reinvasion as affected
by some environmental influences. J Range Manag
21(4):209–213
Jones HP, Jones PC, Barbier EB etal (2018) Restoration and
repair of Earth’s damaged ecosystems. Proc R Soc B
285(1873):20172577
Knutson KC, Pyke DA, Wirth TA et al (2014) Long-
term effects of seeding after wildfire on vegetation
in Great Basin shrubland ecosystems. J Appl Ecol
51(5):1414–1424
Kolb KJ, Sperry JS (1999) Differences in drought adaptation
between subspecies of sagebrush (Artemisia tridentata).
Ecology 80(7):2373–2384
Lambert SM (2005) Seeding considerations in restoring big
sagebrush habitat. In: Shaw NL, Pellant M, Monsen SB
(eds) Sage grouse habitat restoration symposium pro-
ceedings. Proc. RMRS-P-38. U.S. Department of Agri-
culture, Forest Service, Rocky Mountain Research Sta-
tion, Fort Collins, pp 75–80
Lesica P, Cooper SV, Kudray G (2007) Recovery of big sage-
brush following fire in Southwest Montana. Rangel Ecol
Manag 60(3):261–269
Maestre FT, Salguero-Gómez R, Quero JL (2012) It is getting
hotter in here: determining and projecting the impacts of
global environmental change on drylands. Philos Trans R
Soc B 367(1606):3062–3075
Mahood AL, Balch JK (2019) Repeated fires reduce plant
diversity in low-elevation Wyoming big sagebrush eco-
systems (1984–2014). Ecosphere 10(2):e02591
Mata-González R, Reed-Dustin CM, Rodhouse TJ (2018) Con-
trasting effects of long-term fire on sagebrush steppe
shrubs mediated by topography and plant community.
Rangel Ecol Manag 71(3):336–344
McAdoo JK, Boyd CS, Sheley RL (2013) Site, competition,
and plant stock influence transplant success of Wyoming
big sagebrush. Rangel Ecol Manag 66(3):305–312
Miller DL, Wood SN (2014) Finite area smoothing with gener-
alized distance splines. Environ Ecol Stat 21(4):715–731
Miller RF, Knick ST, Pyke DA etal (2011) Characteristics of
sagebrush habitats and limitations to long-term conserva-
tion. Greater sage-grouse: ecology and conservation of a
landscape species and its habitats. University of Califor-
nia Press, Berkeley, pp 145–184
Minnick TJ, Alward RD (2012) Soil moisture enhancement
techniques aid shrub transplant success in an arid shrub-
land restoration. Rangel Ecol Manag 65(3):232–240
Landsc Ecol (2024) 39:184
Page 20 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
Monroe AP, Aldridge CL, O’Donnell MS, Manier DJ, Homer
CG, Anderson PJ (2020) Using remote sensing products
to predict recovery of vegetation across space and time
following energy development. Ecol Indic 110:105872
Monroe AP, Nauman TW, Aldridge CL, O’Donnell MS, Duni-
way MC, Cade BS, Manier DJ, Anderson PJ (2022)
Assessing vegetation recovery from energy develop-
ment using a dynamic reference approach. Ecol Evol
12(2):e8508
Monroe AP, Tarbox BC, Jeffries MI, Welty JL, O’Donnell MS,
Arkle RS, Pilliod DS, Coates PS, Heinrichs JA, Manier
DJ, Aldridge CL (2024) Sagebrush recovery projections
across the biome, 30 years after two seeding treatment
applications, and associated model data (1986–2021).
USGS Data Release. https:// doi. org/ 10. 5066/ P13HM
GUZ
Morris LR, Monaco TA, Sheley RL (2014) Impact of cultiva-
tion legacies on rehabilitation seedings and native spe-
cies re-establishment in Great Basin shrublands. Rangel
Ecol Manag 67(3):285–291
Nelson ZJ, Weisberg PJ, Kitchen SG (2014) Influence of cli-
mate and environment on post-fire recovery of mountain
big sagebrush. Int J Wildland Fire 23(1):131
Newbold T, Hudson LN, Hill SL etal (2015) Global effects
of land use on local terrestrial biodiversity. Nature
520(7545):45–50
Newhall F (1972) Calculation of soil moisture regimes from
climatic record. Rev. 4 Mimeographed. U.S. Department
of Agriculture, Soil Conservation Service, Washington,
p 17
Noss RF, LaRoe III ET, Scott JM (1995) Endangered ecosys-
tems of the United States: A preliminary assessment of
loss and degradation. National Biological Service Report
28. U.S. Department of the Interior, U.S. Fish and Wild-
life Service, Washington, p 60
Noy-Meir I (1973) Desert ecosystems: environment and pro-
ducers. Annu Rev Ecol Syst 4:25–51
O’Donnell MS, Manier DJ (2022a) Soil-climate estimates in
the western United States: climate averages (1981–2010):
U.S. Geological Survey data release. https:// doi. org/ 10.
5066/ P9ULG C03
O’Donnell MS, Manier DJ (2022b) Spatial estimates of soil
moisture for understanding ecological potential and risk:
a case study for arid and semi-arid ecosystems. Land
11(10):1856
O’Connor RC, Germino MJ, Barnard DM etal (2020) Small-
scale water deficits after wildfires create long-lasting
ecological impacts. Environ Res Lett 15(4):044001
O’Donnell MS, Edmunds DR, Aldridge CL etal (2022) Defin-
ing biologically relevant and hierarchically nested popu-
lation units to inform wildlife management. Ecol Evol
12(12):e9565
Orning EK, Heinrichs JA, Pyke DA, Coates PS, Aldridge CL
(2023) Using state-and-transition simulation models to
scope post-fire success in restoring greater sage-grouse
habitat. Ecol Modeling 483:110396
Ott JE, McArthur ED, Roundy BA (2003) Vegetation of
chained and non-chained seedings after wildfire in Utah.
J Range Manag 56(1):81–91
Ott JE, Cox RD, Shaw NL etal (2016) Postfire drill-seeding
of Great Basin plants: effects of contrasting drills on
seeded and nonseeded species. Rangel Ecol Manag
69(5):373–385
Ott JE, Cox RD, Shaw NL (2017) Comparison of postfire seed-
ing practices for Wyoming big sagebrush. Rangel Ecol
Manag 70(5):625–632
Ott JE, Kilkenny FF, Summers DD, Thompson TW (2019)
Long-term vegetation recovery and invasive annual sup-
pression in native and introduced postfire seeding treat-
ments. Rangel Ecol Manag 72(4):640–653
Pedersen EJ, Miller DL, Simpson GL, Ross N (2019) Hierar-
chical generalized additive models in ecology: an intro-
duction with mgcv. PeerJ 7:e6876
Pellant M (1990) The cheatgrass-wildfire cycle—Are there any
solutions? In: Monsen SB, Kitchen SG (eds) Symposium
on cheatgrass invasion, shrub die-off, and other aspects
of shrub biology and management. General Technical
Report INT-276. USDA Forest Service, Las Vegas, pp
11–18
Pellant M (1994) History and applications of the Intermoun-
tain greenstripping program. In: Monsen SB, Kitchen SG
(eds) Ecology and Management of Annual Rangelands.
General Technical Report INT-313. USDA Forest Ser-
vice, Boise, pp 63–68
Perring MP, Standish RJ, Price JN et al (2015) Advances in
restoration ecology: rising to the challenges of the com-
ing decades. Ecosphere 6(8):131
Pilliod DS, Welty JL, Toevs GR (2017) Seventy-five years of
vegetation treatments on public rangelands in the Great
Basin of North America. Rangelands 39(1):1–9
Pilliod DS, Welty JL, Jeffries MI (2019) USGS Land Treat-
ment Digital Library data release: a centralized archive
for land treatment tabular and spatial data (ver. 3.0,
November 2020) U.S. Geological Survey data release.
https:// doi. org/ 10. 5066/ P98OB OLS
Pilliod DS, Jeffries MI, Welty JL, Arkle RS (2021) Protecting
restoration investments from the cheatgrass-fire cycle in
sagebrush steppe. Conserv Sci Pract 3(10):e508
Pyke DA (2011) Restoring and rehabilitating sagebrush
habitats. In: Knick ST, Connelly JW (eds) Greater
Sage-Grouse: ecology and conservation of a landscape
species and its habitats, Studies in Avian Biology. Uni-
versity of California Press, Berkeley, pp 531–548
Pyke DA, Wirth TA, Beyers JL (2013) Does seeding after
wildfires in rangelands reduce erosion or invasive spe-
cies? Restor Ecol 21(4):415–421
Pyke DA, Shriver RK, Arkle RS etal (2020) Postfire growth
of seeded and planted big sagebrush—strategic designs
for restoring greater sage-grouse nesting habitat. Restor
Ecol 28(6):1495–1504
R Core Team (2022) R: a language and environment for sta-
tistical computing. R Foundation for Statistical Com-
puting, Vienna
Reichenberger G, Pyke DA (1990) Impact of early root com-
petition on fitness components of four semiarid species.
Oecologia 85(2):159–166
Remington T, Tart DL, Manning ME, Welty JL, Pilliod DS
(2021) Chapter A. Introduction to the sagebrush biome.
In: Remington T, Deibert P, Hanser S, Davis D, Robb
L, Welty J (eds) Sagebrush conservation strategy—
Challenges to sagebrush conservation, U.S. Geological
Survey Open-File Report 2020–1125, pp 3–9
Landsc Ecol (2024) 39:184 184
Page 21 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol:. (1234567890)
Rey Benayas JM, Newton AC, Diaz A, Bullock JM (2009)
Enhancement of biodiversity and ecosystem services
by ecological restoration: a meta-analysis. Science
325:1121–1124
Rigge M, Homer C, Cleeves L et al (2020) Quantifying
Western U.S. rangelands as fractional components
with multi-resolution remote sensing and in situ data.
Remote Sens 12(3):412
Rigge M, Homer C, Shi H etal (2021) Rangeland fractional
components across the western United States from
1985 to 2018. Remote Sens 13(4):813
Rigge MB, Bunde B, Postma K, Shi H (2022) Rangeland
Condition Monitoring Assessment and Projection
(RCMAP) fractional component time-series across the
western U.S. 1985–2021: U.S. Geological Survey data
release. https:// doi. org/ 10. 5066/ P9ODA ZHC
Rodhouse TJ, Irvine KM, Sheley RL etal (2014) Predicting
foundation bunchgrass species abundances: model-
assisted decision-making in protected-area sagebrush
steppe. Ecosphere 5(9):at108
Roundy BA, Chambers JC (2021) Effects of elevation and
selective disturbance on soil climate and vegetation in
big sagebrush communities. Ecosphere 12(3):e03377
Roundy BA, Chambers JC, Pyke DA etal (2018) Resilience
and resistance in sagebrush ecosystems are associated
with seasonal soil temperature and water availability.
Ecosphere 9:e02417
Sage RF (2020) Global change biology: a primer. Glob
Chang Biol 26(1):3–30
Shaw NL, DeBolt AM, Rosentreter R (2005) Reseeding big
sagebrush: Techniques and issues. In: Shaw NL, Pel-
lant M, Monsen SB (eds) Sage-grouse habitat restora-
tion symposium proceedings. Proc. RMRS-P-38. U.S.
Department of Agriculture, Forest Service, Rocky
Mountain Research Station, Fort Collins, pp 99–108
Shinneman DJ, McIlroy SK (2016) Identifying key climate
and environmental factors affecting rates of post-fire
big sagebrush (Artemisia tridentata) recovery in the
northern Columbia Basin, USA. Int J Wildland Fire
25(9):933–945
Shinneman DJ, Germino MJ, Pilliod DS, Aldridge CL, Vail-
lant NM, Coates PS (2019) The ecological uncertainty of
wildfire fuel breaks: examples from the sagebrush steppe.
Front Ecol Environ 17(5):279–288
Shinneman DJ, Strand EK, Pellant M etal (2023) Future direc-
tion of fuels management in sagebrush rangelands. Ran-
gel Ecol Manag 86:50–63
Shriver RK, Andrews CM, Pilliod DS etal (2018) Adapting
management to a changing world: warm temperatures,
dry soil, and interannual variability limit restoration suc-
cess of a dominant woody shrub in temperate drylands.
Glob Change Biol 24(10):4972–4982
Simler-Williamson AB, Germino MJ (2022) Statistical con-
siderations of nonrandom treatment applications reveal
region-wide benefits of widespread post-fire restoration
action. Nat Commun 13(1):3472
Simler-Williamson AB, Applestein C, Germino MJ (2022)
Interannual variation in climate contributes to contin-
gency in post-fire restoration outcomes in seeded sage-
brush steppe. Conserv Sci Pract 4:e12737
Smith JT, Allred BW, Boyd CS etal (2020) Are sage-grouse
fine-scale specialists or shrub-steppe generalists? J Wildl
Manag 84(4):759–774
Soil Survey Staff (1999) Soil taxonomy: a basic system of soil
classification for making and interpreting soil surveys,
second edition. Agricultural handbook 436, United States
Department of Agriculture, Natural Resources Conserva-
tion Service, Washington, p 886
Soil Survey Staff (2014) Keys to soil taxonomy, 12th edn.
United States Department of Agriculture, Natural
Resources Conservation Service, Washington, p 372
Stevens R, Monsen SB (2004) Chapter17. Guidelines for res-
toration and rehabilitation of principal plant communi-
ties. In: Monsen SB, Stevens R, Shaw NL (eds) Restor-
ing western ranges and wildlands, vol. 1. Gen. Tech. Rep.
RMRS-GTR-136-vol-1. U.S. Department of Agriculture,
Forest Service, Rocky Mountain Research Station, Fort
Collins, pp 199–294
Strassburg BBN, Beyer HL, Crouzeilles R etal (2019) Strate-
gic approaches to restoring ecosystems can triple conser-
vation gains and halve costs. Nat Ecol Evol 3(1):62–70
Suding KN (2011) Toward an era of restoration in ecology:
successes, failures, and opportunities ahead. Annu Rev
Ecol Evol Syst 42(1):465–487
Svejcar T, Boyd C, Davies K, Hamerlynck E, Svejcar L (2017)
Challenges and limitations to native species restoration
in the Great Basin, USA. Plant Ecol 218(1):81–94
Thompson TW, Roundy BA, Durant McArthur E, Jessop BD,
Waldron B, Davis JN (2006) Fire rehabilitation using
native and introduced species: a landscape trial. Rangel
Ecol Manag 59(3):237–248
United Nations General Assembly (2019) Resolution adopted
by the General Assembly on 1 March 2019. United
Nations Decade on Ecosystem Restoration (2021–2030).
United Nations, New York
U.S. Geological Survey (2005) Elevation derivatives for
national applications. U.S. Geological Survey Fact Sheet
2005–3049, p 2
Van Wambeke AR (2000) The Newhall Simulation Model for
estimating soil moisture & temperature regimes. Depart-
ment of Crop and Soil Sciences, Cornell University,
Ithaca, p 9
Walker B, Holling CS, Carpenter SR, Kinzig A (2004) Resil-
ience, adaptability and transformability in social–ecolog-
ical systems. Ecol Soc 9(2):5
Wambolt CL, Walhof KS, Frisina MR (2001) Recovery of big
sagebrush communities after burning in south-western
Montana. J Environ Manag 61(3):243–252
Wijayratne UC, Pyke DA (2009) Investigating Seed Longevity
of Big Sagebrush (Artemisia tridentata). U.S. Geological
Survey Open-File Report 2009–1146, p 26
Williams AP, Cook ER, Smerdon JE etal (2020) Large con-
tribution from anthropogenic warming to an emerging
North American megadrought. Science 368:314–318
Wood SN (2017) Generalized additive models: an introduction
with R, 2nd edn. CRC Press, Boca Raton
Wood SN (2023) mgcv: mixed GAM computation vehicle with
automatic smoothness estimation. R package version
1.8-42
Wood DJA, Seipel T, Irvine KM, Rew LJ, Stoy PC (2019) Fire
and development influences on sagebrush community
Landsc Ecol (2024) 39:184
Page 22 of 23
184
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Vol.: (0123456789)
plant groups across a climate gradient in northern
Nevada. Ecosphere 10(12):e02990
Wu J, Zhang Q, Li A, Liang C (2015) Historical landscape
dynamics of Inner Mongolia: patterns, drivers, and
impacts. Landsc Ecol 30(9):1579–1598
Yirdaw E, Tigabu M, Monge A (2017) Rehabilitation of
degraded dryland ecosystems—review. Silva Fennica
51:167
Publisher’s Note Springer Nature remains neutral with regard
to jurisdictional claims in published maps and institutional
affiliations.
Landsc Ecol (2024) 39:184 184
Page 23 of 23
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”),
for small-scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are
maintained. By accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use
(“Terms”). For these purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or
a personal subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or
a personal subscription (to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the
Creative Commons license used will apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data
internally within ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking,
analysis and reporting. We will not otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of
companies unless we have your permission as detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that
Users may not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to
circumvent access control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil
liability, or is otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by
Springer Nature in writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer
Nature journal content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates
revenue, royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain.
Springer Nature journal content cannot be used for inter-library loans and librarians may not upload Springer Nature journal
content on a large scale into their, or any other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any
information or content on this website and may remove it or features or functionality at our sole discretion, at any time with or
without notice. Springer Nature may revoke this licence to you at any time and remove access to any copies of the Springer Nature
journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express
or implied with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or
warranties imposed by law, including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be
licensed from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other
manner not expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com