Participatory Modeling of Endangered Wildlife Systems:
Simulating the Sage-grouse and Land Use in Central Washington
Allyson Beall, Len Zeoli and Andrew Ford
Program in Environmental Science and Regional Planning
Washington State University
Pullman WA 99164-4430
206-427-0962, 509-339-3473, 509-335-7846
abeall @mail.wsu.edu, lzeoli@mail.wsu.edu, forda@mail.wsu.edu
Abstract: The Greater sage-grouse (Centrocercus urophasianus) occupies the sage brush
habitats of Western North America. Large population declines in the last several decades have
made it a candidate for possible listing under the Endangered Species Act. Listing was recently
avoided in part because seventy local working groups are developing long-range management
plans in conjunction with federal and local agencies. The Foster Creek Conservation District, a
working group in Douglas County, Washington, saw the potential for system dynamics to
synthesize known sage-grouse dynamics and local land use patterns to support development of
their Habitat Conservation Plan and subsequent land management decisions. The resulting
model is providing insights into the cropland and shrub steppe ecosystems of Douglas County
and the management scenarios which may prevent the sage-grouse from an endangered status.
The model is designed to facilitate and support land use management decisions through the
collaborative exploration of model parameters and simulated scenarios.
Key words: participatory modeling, wildlife modeling, sage-grouse, collaboration, system
dynamics, land use modeling, endangered species.
Introduction
The Sage-grouse
The Greater sage-grouse (Centrocercus urophasianus) is a unique western North American
gallinaceous species that lives in the sagebrush (Artemisia) habitats of the western United States
and adjacent Canada (Figure 1). Sage-grouse are known as a sage brush obligate species because
they depend on sagebrush for food, shelter and nesting. The sagebrush areas of Douglas
County, Washington (Figure 2) and other North American locales have been greatly changed by
agricultural conversion, fire, invasion of exotic annuals, fragmentation, urbanization and
inappropriate livestock management (Connelly et al. 2004, Schroeder et al. 1999) to the extent
that sagebrush habitat is now found in patches of varying size and condition (Quigley and
Arbelbide 1997). In Douglas County alone, about 75% of the natural habitat has been converted
to agricultural land (Douglas County Draft MSHCP 2005). Anthropogenic change and
fragmentation of habitat have been the major driving forces in the decline of sage-grouse
populations which no longer exist across their former extensive historical range (Connelly et al.
2004, Stinson et al. 2004). Concern across the US West about this decline has caused the sage-
grouse to be considered for inclusion in the U.S. federal threatened and endangered species list
by the US Department of Fish and Wildlife (USFW). Listing will likely result in changes in the
management of the remaining sagebrush lands that harbor populations of sage-grouse and
consequently affect the activities and livelihoods of those dependent on sage-grouse lands
(Wambolt et al. 2002). Due to the controversy over this potential listing, and in lieu of listing at
this time, federal land management agencies have agreed to participate with local working
groups to develop long-range management plans that address sage-grouse population declines
and habitat needs.
Figure 1.Greater sage-grouse (Centrocercus urophasianus)
WHATCOM
BB nistoric RANGE
HB oCurRENT RANGE
a OUTUER OBSERVATIONS
Figure 2. Current and historic range of the Greater sage-grouse in Washington State. (Douglas County Multi
Species Habitat Conservation Plan (MSHCP)).
The Foster Creek Conservation District
Aside from the potential of federal listing, the sage grouse is listed as threatened in the State of
Washington (Washington Department of Fish and Wildlife 2005). To help address the sage-
grouse and other species of concern, the Foster Creek Conservation District (FCCD), Douglas
County, Washington is presently developing a Multiple Species Habitat Conservation Plan
(MSHCP, hereafter referred to as HCP). A large percentage of the land in Douglas County is
privately owned. Federal and state listings both have the potential to regulate private property
management. “It is the expressed desire of the private agricultural land owners in Douglas
County to reverse the declining population trends of FESA Species as well as other key fish and
wildlife species within the County”(Douglas County Draft MSHCP 2005). As part of this effort,
the District saw the potential for system dynamics to synthesize sage-grouse biology with land
use patterns to form a system-wide perspective of local impacts on the sage-grouse population.
The subsequent model has been named the Integrated Sage-Grouse and Human Systems Model
to indicate the broad scope of the FCCD’s management agenda.
Development of this model has been guided by a belief that sound ecological management
happens only with respectful coordination and communication between land management
agencies and land owners. Part of this coordination includes data sharing. The management
agency-land owner relationship may be strained, perhaps not so much by the data itself, but by
the data collection processes and types of data which either party considers important when
making decisions. Agency biologists are bound to scientific protocol, such as Population
Viability Anal (PVA), and peer review that produces reports that may be difficult to
understand or hard to access by the general public. Land owners use historical information
handed down through generations, personal observations, and instinct developed by their
intimate knowledge of the land. Local knowledge is typically not offered in a format that will
withstand scientific peer review. However they differ, both scientific and local knowledge are
valid and useful. Conversely, both types of knowledge may contain inconsistencies brought
about by lack of information, misinformation, or an inadequate understanding of system
complexity and dynamics. Furthermore, landowners may feel that issues concerning their
livelihoods should be more prominently included in the process. Melding local and scientific
information into a system dynamics model offers a unique venue for data verification, shared
learning, and improvements in communication and trust (van den Belt 2005, Stave 2002 p.139).
This in turn increases the likelihood that land use decisions will lead to improved stewardship.
The role of system dynamics
We believe system dynamics modeling can help the sage-grouse working groups better
understand land management challenges posed by declining sage-grouse populations. This
method has the greatest potential when used in a participatory fashion by scientists and managers
working together with others who also have a stake in land management decisions. Using group
system dynamics modeling for participatory environmental problem solving is a relatively new
process which has been used on a variety of environmental problems such as air quality, water
quality and quantity, and biological conservation management. Marjan van den Belt, describes
five case studies, their models and the lessons learned from the processes (van den Belt 2004).
Other published case studies include that of Kris Stave who used group system dynamics
modeling to help the citizens of Las Vegas explore remedies to air quality problems (Stave 2002,
p. 139). Tidwell et. al. used system dynamics modeling to assist citizens with watershed planning
in the Middle Rio Grand River valley (Tidwell et al 2004, p.357). Wildlife models have been
developed for bear management (Siemer and Otto 2005, Faust et al. 2004) and fishery
management (Otto and Struben 2004 p. 287). Videira et al. modeled “tourism, eco-tourism,
aquaculture, fishing, wildlife protection and nature conservation, effluent discharge and
navigation of fishing and recreation boats” (Videira et al. 2004, Videira 2005 p. 27). Spatial-
Dynamics were used by BenDor et al. in a decision support tool for ash borer eradication
(BenDor et al. 2005).
The Integrated Sage-Grouse and Human Systems Model was developed in collaboration with
land owners, agency representatives’ which included scientific experts, and representation from
The Nature Conservancy. The FCCD group had worked together for many years prior to this
modeling effort. The landowner committee and the technical committee had been working on the
HCP prior to the federal mandate to develop sage grouse plans due to the fact that sage-grouse
are listed as threatened in Washington State. This group was well into a participatory process
and they have developed respectful relationships and a sense of cohesiveness. After USFW
approval of the HCP, the next challenge that awaits the group is convincing land owners to sign
on to the plan. One of the goals of the model was to help FCCD get the “best bang for the buck”
when targeting landowners for HCP sign-on. Signing on to an HCP requires that landowners
develop a farm or ranch plan which must be approved as being within the guidelines of the HCP.
Land owners also agree to be monitored for compliance with their plan. In return, the land owner
is covered under the blanket of the HCP’s, ESA section 10, incidental take permit” and assured a
level of regulatory certainty.
The intention of the Integrated Sage-Grouse and Human Systems Model is to further develop
insights into the cropland and shrub steppe ecosystems of Douglas County to facilitate and
support land use management decisions affecting the Greater sage-grouse and other obligate
species. The model was designed to help FCCD understand which land types and conservation
efforts are most important for the recovery of the sage-grouse and within reasonable reach of
local landowners. Additionally, because sage-grouse are sage brush obligates, improving sage-
grouse habitat should also assist with the conservation of other species of concern.
t Washington Fish and Wildlife (WDFW), US Fish and Wildlife (USFW), Bureau of Land Management (BLM),
Natural Resources Conservation Service (NRCS), WA Department of Natural Resources (WDNR), Douglas County
Farm Service Agency, and an “at large” range expert.
2th 1982, Congress adopted Section 10 of the Endangered Species Act [(16 U.S.C. §1539(a)(2A)] as a way to
promote, "creative partnerships between the public and private sectors...in the interest of species and habitat
conservation." Section 10 authorized Habitat Conservation Plans (HCPs) to give landowners a means by which they
could "incidentally take" listed species or their habitats only after the landowners have identified what will be done
to "minimize and mitigate" the impact of the permitted take on the listed species.” www.fostercreek.net/esa.html
The Group Modeling Process
The model was constructed with Vensim PLE plus (Ventana Systems) over 12 weeks. The
FCCD participants provided all of the data and insights as to the operation of the sage-grouse
system in Douglas County. After the initial meeting, FCCD met as a group with the modelers on
two other occasions to provide feedback on model structure. Frequent contacts via phone and
email were also used to confirm important relationships within the model. Additional contact
with Mike Schroeder, Washington Department of Fish and Wildlife (WDFW) was needed to
confirm sage-grouse parameters. FCCD is very fortunate to have Schroeder as their official state
biologist. He has spent of good deal of his life devoted to the study of shrub steppe species and is
one of the most highly published experts on sage-grouse biology.
The model was developed in an iterative fashion. The land use portion went through two major
iterations, and the sage-grouse life history sector was developed in three major iterations. A
unique feature of the model is the integration of the sage-grouse life cycle with the land use
sector. The feedbacks between land use and sage-grouse were perhaps the most important part of
the project. These feedback relationships were developed through four major iterations. And
finally, the interface views went through multiple modifications as model building progressed.
Many insights into the model were gained in these iterations, and others were gained by
experimentation with the completed model. This led to another major iteration of the interface
(that occurred after the contract period but before final presentation to the group) in the interests
of making these insights available and the model more accessible and usable by all stakeholders.
The sources of the model constants and reasoning behind the quantitative links have been
thoroughly documented in the comment window of each variable.
Parameter values were estimated in one-at-a-time fashion using a wide range of information
from the FCCD participants. Ford (1999 p.174) encourages modelers to take advantage of
information from across the information spectrum shown in Figure 3, and the FCCD participants
were encouraged when we explained that we would follow this approach. They were especially
encouraged by our willingness to make use of the expert judgment and personal intuition of the
members of the FCCD.
Physical Controlled Uncontrolled | Social Social Expert Personal
laws physical physical system. system judgment intuition
experiments | experiments | data cases
Figure 3. The information spectrum. (Ford 1999 p. 175).
The participants provided a great deal of data and statistical estimates that had been acquired
through peer reviewable processes. The most current data available on land cover had been
translated from aerial photos to GIS which FCCD used to delineate Douglas County into 16 land
types as part of their HCP. The range of each species of concern was then overlaid on the GIS.
Figure 4 illustrates the Douglas County sage-grouse area which covers 292,030 hectares and
includes 11 of the 16 land categories. Suitability indices were assigned to each category and
include consideration for sage-grouse density in any particular land type.
MAP LEGEND
fe
[1 evsts county border
Roads
Landscape classification
Continuous shrub steppe; gertle slope
Continuous shrub stoppe, steep (210 degrees)
Moderately trapmerted shrub ste
Moerately fragmented sh
steep (10 degrees)
ight fragmented shrub steppe
slope
CRP lands. lees fragmented landscape context
‘CRP lands - more fragmented landscape context
Cropland less ragmented landscape contest
Cropland more fragmented landscape context
Forest & Shrub
Paustone Wetland
Rocky cliffs barren & slope > 25 degrees)
Baren & Roads
Open Water
Orchards & Vineyards
Urban
Total Area
108.000 he
58,000 ha
19000 ha
6000 hs
20,000 ha
37000 ha
3300 ha
53000 ha
111,000 ha
9,700 he
130 ha
1,800 ha
2700 ba
7.400 ha
7.400 he
wha
Figure 4. Douglas County GIS with sage-grouse range delineated. Legend indicates land category totals for the
entire county.
Sage-grouse population estimates had been acquired by Schroeder through the use of peer
reviewed sage-grouse estimation procedures. Sage-grouse life history parameters were derived
from data provided by Schroeder, and from peer reviewed journal articles and reports about
sage-grouse, many authored or co-authored by Schroeder. The structure of the life history
portion of the model follows a life-stage transition form familiar to population biologists.
The parameters that linked sage-grouse density with land use were based on assumptions made
by the modelers, that are easy to follow, and are similar to those used in other types of biological
models. The parameters that are most uncertain are the relative values of habitat suitability. The
debate about these parameters and more importantly, what land use change will improve
suitability, is presently being discussed outside of the model. The model simply condenses data
into one format so that questions about original data assumptions may be tested. It has taken the
best science available to FCCD and placed it in a form that is accessible to both scientists and
non scientists. It has also made the concerns of landowners accessible to agency representatives
and scientists. FCCD realizes that ecosystem management is a holistic endeavor and must
include the concerns of the ecosystem’s human occupants as well as concern for the non human
occupants.
The Model Structure
Clearly, the model is a work in progress. As of March 2006, the model is comprised of 10 stocks
and 18 flows. The diagrams and results are spread across 26 views. Time is simulated in
months, and a typical simulation runs for 50 years to encompass the habitat planning horizon.
There are over 200 input parameters which are specified in the model. Spatial aspects have been
included through the aggregation of land categories and through the suitability indices which
include consideration for sage-grouse density (Schroeder per. com. 2005).
Land Use and Suitability Indices
Aggregate land categories include cropland, Conservation Reserve Program (CRP)* and shrub
steppe designations which are further delineated by their proximity to one another, steepness of
slope, and their degree of fragmentation.* Shrub Steppe designations cannot be changed through
modified land use; however, Wheatland and CRP land categories have been integrated into stock
and flow variables to allow the user to change percentages of CRP contracts in the sage-grouse
area (Figure 5). Additionally, the proximity of Wheatland and CRP to shrub steppe was also
considered.
This population of sage-grouse is migratory and uses different parts of its range for winter and
breeding, therefore each of the land use categories has an assigned suitability index for both
winter and breeding use. This effectively splits land designation into 22 categories each with a
unique suitability index. For example the highly fragmented shrub steppe has a suitability of 0.65
for breeding (out of 1) and 0.05 for winter habitat, while gentle and continuous shrub steppe has
a breeding suitability of 0.1 and a winter suitability of 0.6.
The habitat suitability index for any land category, multiplied by hectares of land in that
category, creates a parameter called “habitat units”. The assumption made is that, given a
positive rate” of increase in the sage-grouse population, more habitat units whether acquired by
expanding habitat through restoration, or by improving existing habitat, should be able to support
more birds. Due to the potential for controversy over the suitability indices, user interfaces
provides sliders to manipulate the indices.
3 “The Conservation Reserve Program (CRP) is a voluntary program for agricultural landowners. Through CRP, [a
farmer] can receive annual rental payments and cost-share assistance to establish long-term, resource conserving
covers on eligible farmland. The Commodity Credit Corporation (CCC) makes annual rental payments based on the
agriculture rental value of the land, and it provides cost-share assistance for up to 50 percent of the participant’s
costs in establishing approved conservation practices. Participants enroll in CRP contracts for 10 to 15 years.”
(Farm Service Agency)
“Land designations: cropland in crop landscape, cropland in shrub steppe (ss) landscape, CRP in crop landscape,
CRP in shrub steppe landscape, shrub steppe gentle and continuous, shrub steppe steep and continuous, shrub steppe
entle and moderate, shrub steppe steep and moderate, shrub steppe fragmented, palustrine wetland and barren.
A positive rate of increase is necessary for improved population numbers. A negative rate would indicate that the
population is in a long term decline and will not recover without intervention. Negative rates for sage-grouse could
be caused such things as inbreeding depression (lowers productivity) or a breakdown in the social structure of the
lek mating system. (also see foot note 6)
a
10 years SS
Conversion of CRP
‘Total Land
Conversion of CRP to Moorman to Wheat in crop
Wheat in SS landscape landscape
— ‘These values are used in the x
Cropland Management view.
‘Total contracte
d CRP
Cropland in SS
landseape that
will always be
CRP
Conversion of ‘Comension of wheat
wheat to CRP in audi taun ten to CRP in erop
cae t= a
SS landscape — ESiewae cere
Area of SS land: ie aL bsanhadonys
rea of SS landscape and [iniiseape with
with crops and CRP - RP ae
Figure 5. Cropland and CRP View of the systems model
Economics and Local Knowledge
Gathering local information in Douglas County was limited due to the time constraints of the
model building period; however, the model offers a starting point for discussion and continued
discovery of the potential impacts of both land and sage-grouse management upon land owners
and sage-grouse. Sliders are available that allow the user to estimate the cost of improvements
per acre that landowners may incur for inclusion in the HCP. Additional concerns include those
of wheat growers who are likely to be impacted by the 2007 Farm Bill. After a review of an early
version of the model, Wade Troutman (per. com. 2005) suggested a “switch” that would remove
CRP from the system (a potential outcome of a farm bill which removed conservation support).
This was subsequently added, and the model now simulates the impact of this potential loss of all
CRP upon sage-grouse populations and area economics.
A simple economic model that contrasts area-wide net CRP income with wheat income and
production costs has been included (Figure 6). Considerable time has been spent on a “cattle
economics and potential land impacts of grazing” component of the model. At present this has
not been integrated into the sage-grouse model due to the lack of verification from ranchers. This
is a complex system in of itself and there are many assumptions and parameters that require
consensus as to both their impacts and usefulness to the sage-grouse model. However, the model
presently offers ranchers a platform to discover potential impacts to sage-grouse through the
improvement of shrub steppe suitability. Ranchers will no doubt begin to describe how the
improvements should or will take place, and the benefits, costs and impacts to ranching
operations. As this discovery develops, FCCD may find it useful to include these items in a
future version of the model.
Fraction of
nhc ERP SH
Hectares of
cropland contracted
into CRP this year
nd change?
Net income from
wheat per year
‘Wheat price per
bushel slider
Bushels per gg"
ere slider A
Wheat production
cost per acre slider
Total land jn wheat
wreatibd andl
‘Yes or No CRP?
Commodity support or
~ # Conservation support?
Actual CRP
faction contracted
“Total available er
po! opland land area
Area of deste
<dwheat land
Current
Cinpland
Month for nd
change to happen
Current
"RP ‘Chunge in hectares one
‘of Wheathind from
2005 values
Change in net income
fiom wheat per year fom
2005 valies
Net change in
Net Wheat cone total CRP and
pe year per hectare ‘Whiat Income
from 2005 values
crys per
a contin
Change in CRP income p me
‘er year ftom 2005 vahies
‘This view is connected to the Wheatland
and CRP view through several
variables. Income and production costs
are to be considered area averages and
may be adjusted with sliders on the
Land Use Control Panel.
Go to Wheatland and CRP |” Go to Land Use Graphs
‘Total contracted CRP
comes from Crophind
(Change in CRP Ha
fiom 2005 values
income from
CRP contracts
\ /
— CRP peracre
Figure 6. Cropland Management View
Sage-grouse Demographics
Important life stages and demographic rates of the greater sage-grouse have been identified from
current literature (Schroeder 1997, Schroeder et al. 1999, Stinson et al. 2004) and with
information provided by Schroeder. Life stages appear in the model (Figure 7) as stocks for eggs
in nests, chicks, adult female birds, and adult male birds. Demographic rates appear as the flows
between the stocks and include variables such as “successful hatches”.
It is not known how density dependence limits sage-grouse populations; however, it does occur.
Biologists have indicated that the current Douglas County population of approximately 650 birds
is most likely at carrying capacity (Connelly et al. 2004). Schroeder (per. com.), notes that
breeding and winter habitat may both be equally limiting at this time. Therefore, population
limits have been set in the model by dividing the current estimated population into the total
breeding and winter habitat units. This establishes a density dependent relationship between the
birds and both types of habitat units. Another density dependant relationship is established
through the major feedback loop highlighted with the darkened blue line on the “Female Life
History” view of the model. This relationship limits the number of nests in available breeding
habitat to a default amount based on the current conditions. In other words, to establish more
nests, more habitat units are needed. Other feedback loops operate through mortality rates of
both chicks and adults and appear on their associated life history views.
Female Sage Grouse Life History
migaitly death of monthly deaths of
elective hatch female sex ratio. wi SESH
. rate
C ~
ae een Chicks ‘Adult Female Birds
eee iying success hatches suceessiul Female ‘monly female deaths
. Chicks fiom senescence
—
This i the
‘dependent
feedback oop
responsi or
population sin,
indated number of
a ieee jee ‘vs
eggs per hen hs
ane Friction of hens
nesting total BH units
required for hens
BH avallabilty cao
for hens
Figure 7. Female Bird Life History and Reproduction
Illustrative Results
Exploring Sage-grouse Demographics
Current demographic rates in Douglas County are favorable for the continued persistence and
growth of the sage-grouse population. Productivity is high and female mortality is low
compared to other studies (Connelly et al. 2004). The model population responds positively to
increases in habitat quantity or quality. However, demographic parameters are not givens, but
rather they are estimates with an associated standard deviation and may change over time with
changing environmental or other conditions, such as an Allee effect®. Sensitivity analysis has
shown that the model is sensitive to changes in certain demographic rates. Sensitive parameters
are chick survival, female mortality, percent of successful females (i.e., females that have more
than 1 chick) and eggs/nest. The systems analysis of these parameters was conducted by
discovering a critical or threshold value for each, defined as the value at which the sage-grouse
° Allee effects are an inverse density dependence that influences reproduction when a population falls below a
critical density that is required for the stimulation of breeding (Akcakaya et al. 1999). It can be very important in
species that have a socially structured mating system (Ebenhard 2000). It is also inclusive of any combination of
factors that cause the growth rate of a population to decline as it gets smaller. Several things can occur in small
populations including an inability to find mates or not enough animals to stimulate a breeding response. Also, if
predation is constant, species in a declining population are subjected to greater and greater risk. At some point,
reduced fitness and inbreeding depression can also occur. Declines in breeding success can cause further declines in
population size and coupled with demographic stochasticity, leading to what is termed the extinction vortex. In
sage-grouse, the social structure associated with the lek mating system might begin to break down in a declining
population causing a lower percentage of females to be successful.
10
population began to decline. Each sensitive parameter was explored independently of the others.
For example, the annual chick survival rate is 0.167. Chick survival is an adjustable slider set at
the default position of 0.167 on the Sage Grouse Life History interface. By decreasing chick
survival in small increments, it was found that the trajectory of the grouse population began to
decline at a value of 0.128 (see Figure 8). This is a very small downward change in the survival
rate (approximately .04) and shows the sensitive nature of this parameter. The specified amount
of change is well within the standard deviation assigned to that rate, and therefore, this threshold
value should be considered as a possible annual rate for any year. Chick survival is the only
parameter with a threshold value that falls within the standard deviation of its rate, although
eggs/nest is very close. The model indicates that chick survival is the most critical parameter,
which coincides with other analysis, with the literature cited above and expert opinion
(Schroeder, per.com. 2005). The same type of analysis was done for the other 3 sensitive
parameters. Results are in Table 1.
Sage Grouse Population
800
600
400
0 120 240 360 480 600 720 840 960 1080 1200
Time (Month)
smoothed total bird population : baseline
smoothed total bird population : reduced chick survival
Figure 8. Response of the Douglas County Greater sage-grouse population to a downward change in the annual
chick survival rate from the baseline model value of 0.167 to a critical value of 0.128. The critical or threshold
value was found by experimentally decreasing the survival rate until the population trajectory began to go negative.
11
Parameter Model Value Threshold Difference Standard
Value Deviation
Chick Survival 0.167 0.128 ~0.04 0.10
Female Mortality 0.25 0.36 09 0.068
Fraction of Successful 0.59 0.45 0.14 0.10
Females
Eggs/nest 9.1 7.0 2.1 2.
Table 1. Threshold values for sensitive parameters in the systems model of the Douglas County Greater sage-grouse
population.
There are also synergistic effects when more than one demographic rate changes at the same
time. Small (and declining) populations are subject to an inverse density dependence known as
the Allee effect which is any factor, or more likely, combination of factors that cause the growth
rate of a population to decline as it gets smaller. The systems model allows the user to change
several demographic rates at the same time to simulate the decrease in productivity and survival
normally associated with declining populations. For example, the 4 sensitive demographic rates
were reduced concurrently by 10%. Results are in Figure 9 and show the population in serious
decline.
Sage Grouse Population
800
600
400
360 480
600
720
Time (Month)
smoothed total bird population : baseline
smoothed total bird population : negative change in 4 parameters
-53, and eggs/nest at 8.2.
840 960
1080 1200
istic response of the Douglas County Greater sage-grouse population to a negative change of 10%
e parameters, chick survival at 0.15, female mortality at 0.275, fraction of successful females at
12
Exploring Land Use Changes
The systems model allows users to investigate the effects of potential land use changes on the
sage-grouse population. Model users have been supplied with sliders for many parameters that
can be used alone or in combination to investigate possible future scenarios. Population
projections resulting from the inclusion of land in the HCP are shown in Figure 10.
Sage Grouse Population
1,200
1,000
800
600
400
(0) 60 120 180 240 300 360 420 480 540 600
Time (Month)
smoothed total bird population : baseline
smoothed total bird population : hcp
smoothed total bird population : 100% hcp
Figure 10. The effect of HCP inclusion on sage-grouse population showing a potential population of up to 1,000
birds. (HCP 0%, is equivalent to a baseline sometime in the future’).
One of the biggest concerns voiced by FCCD was related to losses in the amount of land under
CRP contract. A specific 8% of CRP land has become critical nesting habitat for sage-grouse and
is under threat of removal from the CRP program. We used the model to simulate these losses. It
also simulates potential mitigation for these losses with inclusion of land in the HCP. The blue
line is again the baseline or current condition in Figure 11. The green line illustrates an 8% loss
range wide (includes CRP in cropland). The grey line illustrates an 8% loss just in shrub steppe
landscape CRP. And, finally the red line is total loss of CRP. Figure 12 illustrates that a loss of
100% of the CRP can be more than offset by inclusion of half the land in HCP.
THCP 0% is equivalent to a baseline sometime in the future for two reasons. 1) Recently improved management
practices in the shrub steppe will provide for better habitat suitability. And 2), CRP contracts were initiated in 1997
and 1998. Additional time will allow the habitat in these areas to mature, again improving suitability.
13
Sage Grouse Population
800
700
600
500
400
0 60 120 180 240 300 360 420 480 540 600
Time (Month)
smoothed total bird population : baseline
smoothed total bird population : loss of erp
smoothed total bird population : loss of 8%crp rw
smoothed total bird population : loss of 8%crp ss
Figure 11. An illustration of the effect of CRP loss on sage-grouse population. RW is rangewide, SS is shrub steppe.
Sage Grouse Population
800
700
600
500
400
0 60 120 180 240 300 360 420 480 540 600
Time (Month)
smoothed total bird population : baseline
smoothed total bird population : 50% HCP, 100% CRP loss
Figure 12. 50% inclusion in HCP can offset the effect of 100% loss in CRP on sage-grouse population.
14
Environmental Variation
Biologists, managers and people who make their living from the land know that productivity of
land and therefore its carrying capacity changes from year to year, in large part due to variations
in local weather patterns. Sage-grouse populations tend to fluctuate in approximately 10 year
cycles (Connelly et al. 2004). The FCCD group requested that this type of variability be
included to add more realism to the model. Due to the lack of data on these cycles, variability
was determined by a loose correlation with historic rainfall which did not specifically show a ten
year cycle. We took the high and low values over a 100 year period, and added a random
function to select a value within that range for any given year. The rainfall value was linked to
habitat by causing the number of habitat units to expand when rainfall was above the long-term
average and to shrink when it was below that average. The resulting output did show a cyclic
behavior in the sage-grouse population, but not at the expected interval (Figure 13). It was
decided by the group that such added complexity did not really aid in understanding their system
at this time. Other forms of adding variability are being considered for future work.
Female Birds React to Environmental Variability
0 60 120 180 240 300 360 420 480 540 600
Time (Month)
Adult Female Birds : habitat variability
Environmental Variation for BH : habitat variability
Environmental Variation for WH : habitat variability
Figure 13. Environmental variability was added to both breeding habitat (BH) and winter habitat (WH).
Concluding Discussion
When we arrived at our first meeting with the FCCD, we had very little knowledge of Douglas
County or the problems faced by landowners. We only knew that FCCD was concerned about
their sage-grouse. Throughout the next 12 weeks, group members educated us about the nuances
of their concerns, provided expert data and the context in which it was gathered and analyzed,
and they gave us a look inside the lives of the people who call the sage-grouse country of
15
Douglas County home. They invited us to use our skills as model builders to help them
synthesize a variety of data types into one context which would allow them to investigate the
connections and feedbacks in their system. Furthermore, they understood that a variety of
concerns must be addressed for ecosystem management to be successful.
The model was delivered in its present form to the Foster Creek Conservation District in
November of 2005. It has been designed to help develop instincts for adaptive management of
the complex sage brush steppe and cropland ecosystems that are home to the sage-grouse of
Douglas County. It combines sage-grouse biology with land use patterns to simulate the response
of the birds to current and future management decisions. It was designed to be self contained and
available to all interested parties in the sense that users do not need the modelers present to
operate the model. The FCCD participants appreciate that the model is not designed to predict
any specific values or points in time. Rather, the goal is to indicate potential trends in the sage-
grouse population based on past and present data. The model is designed and should be used for
combining and computing data into an illustrative representation of a complex system. Models
are valuable because they simultaneously compute a myriad of parameters and relationships and
can simulate system trajectories under various conditions. Models also offer a step up from
decisions made on “gut feeling” by allowing those feelings to be tested and evaluated.
As we assimilated FCCD data into the systems model we tried to anticipate the questions that
would be asked of the model as well as the reactions of the working group. Experience tells us
that as users explore a model they will inevitably want to start making improvements. Potential
improvements suggest that we have encouraged people to better understand a system that is
important to them. Adaptive management suggests that humanity take a holistic view of its
surroundings, to learn about the parts as they are related to the whole, and to understand the
dynamic interactions of those parts as they work together in a system. It also suggests that when
new information is gained, and better options discovered, that better management should follow.
This well tested model structure offers a base on which additional concerns may be layered; it
may be easily updated as new information becomes available. We have discussed adding
additional economic components, more ranching concerns, and a “more methodical process to
quantify suitability indices” (Dudek pers. com.). FCCD is also working on a remote sensing
project to help them monitor habitat. This data may be used to update the model as habitat
changes occur. Additionally, the land use template of the model may be adapted and applied to
other species of concern in Douglas County.
We will soon be conducting interviews with FCCD members to assess the model and modeling
process to date. We will continue to document the model, the process and its usefulness to the
management objectives of the Foster Creek Conservation District Multi Species Habitat
Conservation Plan.
16
References
Akcakaya, H. R., Burgman, M. A. and Ginzburg, L. R. 1999. Applied Population Ecology.
Sunderland, Massachussettes, Sinauer Associates.
BenDor, T. Metcalk, S., Fentenot, L. and Sangunett, B. 2005. A decision Support System for
Emerald Ash Borer Eradication Using Spatial-Dynamic Modeling. Paper submitted to the 23"
International Conference of the System Dynamics Society, July 17-21, 2005, Boston, USA.
Connelly, J. W., S. T. Knick, M. A. Schroeder, and S. J. Stiver 2004. Conservation Assessment
of Greater Sage-grouse and Sagebrush Habitats. Cheyenne, Wyoming, Western Association of
Fish and Wildlife Agencies.
Douglas County Multi Species Habitat Conservation Plan (MSHCP) Foster Creek Conservation
District Douglas County, WA. Power Point
http://www.westgov.org/wga/initiatives/grouse/dudek.pdfouglas. Retrieved Feb 28, 2006.
Draft Douglas County Multiple Species Habitat Conservation Plan January 13, 2005.
Dudek, B. Foster Creek Conservation District Manager. Waterville WA.
Ebenhard, T. 2000. Population viability analyses in endangered species management: the wolf,
otter and peregrine falcon in Sweden. Ecological Bulletins 48:143-163.
Farm Service Agency http://www. fsa.usda.gov/dafp/cepd/crp.htm. Retrieved Feb 24, 2006.
Faust, L. Jackson R., Ford, A., Earnhardt, J., and Thompson, S.D.. 2004. System Dynamics
Review. Vol. 20. No. 2.
Ford A. 1999. Modeling the Environment. An Introduction to System Dynamics Modeling
of Environmental Systems. Island Press. Covelo, California.
Otto, P. and Struben, J.. 2004. Gloucester Fishery: insights form a group modeling intervention.
System Dynamics Review. Vol. 20.
Quigley, T. M. and Arbelbide, S. J.. 1997. An assessment of ecosystem components in the
interior Columbia River Basin and portions of the Klamath and Great Basins. Portland, Oregon:
U. S. Dept of Agriculture, Forest Service, Pacific Northwest Research Station, U.S. Dept of
Interior, Bureau of Land Management. Report nr PNW-GTR-405.
Richardson, G. and Anderson D. 1995. Teamwork in group model building. System Dynamics
Review. Vol. 11. No. 2.
Schroeder, M. A., Young, J.R. and Braun, C. E. 1999. Sage-grouse. The Birds of North
America. A. Poole and F. Gill, Cornell Laboratory of Ornithology.
17
Schroeder, M. A. 1997. Unusually high reproductive effort by sage-grouse in a fragmented
habitat in north-central Washington. Condor 99: 933-941.
Siemer, W. and Otto P. 2005. A group model-building intervention to support wildlife
management. Paper submitted to the 23” International Conference of the System Dynamics
Society, July 17-21, 2005, Boston, USA.
Stave, K. 2002.Using system dynamics to improve public participation in environmental
decisions. System Dynamics Review, Vol. 18 #2.
Stinson, D.W., Hays, D.W. and Schroeder, M.A 2004.Washington State Recovery Plan for the
Greater Sage-grouse. Washington Department of Fish and Wildlife, Olympia, Washington.
Tidwell, V.D., Passell H.D., Conrad, S.H., and Thomas, R.P. 2004. System dynamics
modeling for community-based water planning: Application to the Middle Rio Grande. Aquatic
Science 66.
Troutman, W. 2005. Foster Creek Conservation District Board Supervisor. Waterville WA.
van den Belt, M. 2004. Mediated Modeling. Island Press. Washington, D.C.; Covelo, California;
London.
Vennix, Jac A. M. 1996. Group Model Building. Facilitating Team Learning Using
System Dynamics. John Wiley and Sons. New York.
Ventana Systems INC., 60 Jacob Gates Road, Harvard, MA 01451.
Videira, N. 2005. Stakeholder Participation in Environmental Decision-Making: The Role of
Participatory Modeling. PhD dissertation. Universidade Nova de Lisboa, Faculdade de Ciéncias
e Tecnologia. Lisboa Portugal.
Videira, N., van den Belt, M., Antunes, P., Santos, R. and Gamito, S. 2004. Decision Support in
Coastal Zone Management in the Ria Formosa, Portugal. In Mediated Modeling. Island Press.
Washington, D.C.; Covelo, California; London.
Wambolt C.L., Aaron J. Harp, Bruce L. Welch, Nancy Shaw, John C. Connelly, Kerry P. Reese,
Clait E. Braun, Donald A. Klebenow, E. Durant McAuthur, James G. Thompson, L. Allen
Torell, and John A. Tanaka. 2002. Conservation of Greater Sage-Grouse on Public Lands in the
Western U.S.: Implications of Recovery and Management Policies. Caldwell, Idaho: Policy
Analysis Center for Western Public Lands. Report nr PACWPL Policy Paper SG-02-02.
Washington Department of Fish and Wildlife 2005
http://wdfw.wa.gov/wlm/diversty/soc/threaten.htm. retrieved July 1, 2005
18
