A group model-building intervention to support wildlife
management decisions
William F. Siemer Peter Otto
Cornell University, Dept. Natural Resources Dowling College, School of Business
119 Femow Hall, Ithaca NY 14853 Oakdale, NY 11769
Tel: (607) 255-2828, Fax: (607) 254-2299 Tel: (631) 244-3192, Fax: (631) 244-5098;
wfs1@comell.edu ottop@ dowling.edu
Abstract
A variety of approaches are being developed to elicit knowledge from clients and develop that
knowledge into conceptual maps and formal simulation models. We completed a project that
provides a case example where the “standard method” was adapted for use in a group model-
building intervention. We worked with a group of 10 wildlife managers to support system
conceptualization, model formulation, and management response to an increase in negative
human-black bear interactions in residential areas of New Y ork State. This article discusses the
procedural and conceptual steps, insights, and lessons leamed from our model building
intervention. Our paper focuses on model-building process and learning outcomes, rather than
quantitative validation of a simulation model.
Keywords: Group model building, Scripts, Standard Method, wildlife management
Introduction
Building system dynamics models with client groups has a long tradition in our field and is well
documented (cf. Morecroft and Sterman 1994; Richardson and Andersen 1995; Vennix 1996). In
the literature several approaches to group model building are discussed (cf. Richardson and Pugh
1981; Roberts et al. 1983; Vennix 1994) with varying stages on how the process of constructing
a computer simulation model involves a number of conceptual activities.
Richardson and Pugh (1981) define seven stages in building a system dynamics model:
problem identification and definition, system conceptualization, model formulation, analysis of
model behavior, model evaluation, policy analysis, and model use or implementation. Roberts et
al. (1983) suggest a similar approach to construct a simulation model. Otto and Struben (2004)
adapted the “standard method” (Hines 2001) for use to increase understanding about policies to
revitalize the fishing industry in Gloucester, Massachusetts. While the standard method consists
of similar iterative activities to elite knowledge from a client and then conceptualizes a model, as
the approaches noted above, the standard method specifies steps that a system dynamics modeler
should consider in a consulting environment. For example, it emphasizes the importance of
identifying key variables, which usually involves in-depth discussion with the client, a reference
mode to express a “hope” and “fear” scenario, and in-depth analysis of the different loops in the
system. The standard method thus uses the lens through which a consultant would look at a
model building initiative. While we do not consider our involvement in the project as consultants
but facilitators we applied the standard method driven by the previous experience gained in a
group modeling initiative for the Gloucester Fishing community.
This article discusses the procedural and conceptual steps, insights, and lessons learned
from our model building intervention. Our paper focuses on model-building process and
learning outcomes, rather than quantitative validation of a simulation model.
The Bureau of Wildlife in the New York State Department of Environmental
Conservation (DEC) is responsible for black bear (Ursus americanus) management in New Y ork
State. Black bears occur throughout New Y ork State, with primary populations inhabiting three
core ranges. Complaints to wildlife agencies about problems with black bears have been
increasing over the past decade in New York and throughout the northeastern United States
(IAFWA 2004). Complaints to wildlife agencies serve as an indicator that people are
experiencing a range of negative economic, psychological, and physiological effects. Of
particular concem in New York State is the increasing level of complaints about negative
human-bear interactions in residential areas (Schusler and Siemer 2004). Residential
development in New Y ork’s bear ranges is increasing and so is the frequency of human-bear
interactions in those areas (NY SDEC 2003a).
DEC and other wildlife management agencies in the northeastern U.S. are responding to
bear-related problems in residential areas, but doing so is costly (IAFWA 2004). It puts a strain
on the small staffs of wildlife agencies and reduces the level of funding available for other
management activities. Managers and others also worry that negative interactions with bear may
lower tolerance for bears and reduce public support for black bear conservation.
Wildlife agency documents provide a source of information on what wildlife managers
perceive as the source of these problems. Managers believe that increases in a bear population
lead to increases in complaints (and New Y ork’s black bear population is growing and becoming
more widely distributed, especially in southem New York [NYSDEC 2003a]). They believe
that availability of anthropogenic food sources (e.g., garbage, bird seed, pet food, gardens, crop
fields) influences the frequency and severity of negative human-bear interactions. In New Y ork
and other states, complaint records do reveal a strong association between problem interactions
and bear attraction to human food sources. Managers believe that problems with black bears can
be reduced through interventions such as regulated bear hunting and problem prevention
education. However, some of the assumptions underlying these policies have not been
rigorously evaluated and uncertainty remains on the nature of this problem and possible
solutions. Research designed to reduce some of this uncertainty could help wildlife managers
develop better problem definitions and better evaluate how proposed management actions are
likely to affect clearly defined problems.
Understanding the factors driving increased complaints about problems with black bears
has broad practical importance to wildlife management agencies. Understanding the systems
that generate negative impacts (the important negative effects that people contact an agency to
complain about [Riley et al. 2002]) in this context is likely to help managers understand a suite
of contemporary wildlife management issues unfolding on the urban-rural interface. The
dynamic interactions occurring between people and black bears in residential areas may be
similar to those occurring between people and coyote, white-tailed deer, and mountain lion living
in the same ecozones. Thus, wildlife managers in New York have a keen interest in
understanding the system of factors that explain why problems with black bears are increasing in
residential areas.
In 2004, we began working with a team of 10 wildlife managers to support system
conceptualization, model formulation, and management response to an increase in negative
human-black bear interactions in residential areas of New Y ork State. This specific project was
part of a larger effort to inform decisions within a new bear management planning framework
(NYSDEC 2003b). Our work culminated in a system dynamics model, which served to improve
wildlife managers’ understanding of the complex interactions occurring between community
residents, wildlife agencies, hunters, and black bears.
Project description
Wildlife managers on the project team were concemed about managing a wide array of negative
human-bear interactions. They decided to focus this project on understanding how to manage
problems with black bears in residential areas. Human-bear interactions are also common in
more rural areas and can result in a range of negative effects that managers wish to limit (e.g.,
damage to row crops, apiaries, and fruit trees). However, residential interactions were of special
concem to the project team because they pose a very low, but real risk to human safety. Human
injuries are expected to remain uncommon, but incidences of human injury and unsafe
encounters with black bears in the northeast have increased in the past decade and managers are
concemed that any continued increase in the rate of threatening incidents could lead to the
devaluing of black bears to pest status in the eyes of wildlife management stakeholders. In order
to continue managing bears as a valued resource (instead of a pest), they believe it is imperative
to manage the number and severity of negative human-bear encounters in residential areas.
Expectations and challenges of the project
After a set of two workshops with the project team and independent working sessions by the
modelers, we defined the following policy (or management action) questions to answer through a
system dynamics modeling process. These questions are shown below.
e How would changes in hunting policies (i.e., the amount of land open to hunting or the
date of season opening) influence the frequency and severity of human-bear interactions
in residential areas?
e How would changes in the level of agency effort devoted to prevention education (i-.,
agency resources expended on information/education actions, providing staff time for on-
site technical assistance to residents) influence the frequency and severity of human-bear
interactions in residential areas?
e Are there leverage points in the system creating residential problems where a wildlife
management agency could reduce the frequency or severity of human-bear interactions in
residential areas through novel or innovative management actions?
These are practical management questions that may appear simple at first glance.
However, each is embedded in a complex set of social and ecological relationships containing
uncertainty, unrecognized parameters, and nonlinear feedback structures. The project team
viewed the group model building intervention as a means to understand (and later communicate
with management stakeholders about) the dynamic complexity of managing a subset of negative
human-bear interactions. In addition to answering the management policy questions outlined
above, the objectives of the project were to: (1) explore wildlife managers’ mental models of
how complaints about residential problems with black bears are generated; (2) explore wildlife
managers’ assumptions about how bear population, housing density, availability of human food
sources, and other key variables influence the rate of complaints about residential problems with
bears, tolerance for bears, and attitudes towards bears; and (3) identify priorities for additional
research (i.e., identify the most important variables in the system that we know the least about).
Achieving these objectives was important to the project team because doing so would enable the
team to implement subsequent steps in a cycle of adaptive impact management (Riley et al.
2003).
Wildlife management agencies have relied on recreational hunting as the primary means
to control the size of black bear populations. There are several uncertainties and challenges
associated with using that management tool to address the problem at hand. Hunting pressure is
difficult to exert on bears living in or near human residential areas; availability of bears to
hunters is influenced by annual variations in food availability and access to private lands;
hunting targets the population of bears, not specific individual bears habituated to human food
sources; and interest in bear hunting and willingness to take bears is not known.
Management agencies have also relied on educational intervention as a tool to control the
number and severity of problems that people experience with bears. However, there is much
uncertainty about the effectiveness of educational interventions as a tool to change behaviors that
create food attractants for bears (e.g., behaviors related to bird feeding, garbage disposal, use of
barbeque grills, etc.). Scholars have questioned the knowledge-attitude-behavior link assumed in
many educational intervention strategies. The specific educational strategies employed by
wildlife management agencies have received little evaluation (Gore 2004, Lackey and Ham
2003). Agencies have typically devoted very limited resources to these programs and it is not
known what level of investment in educational intervention would produce the desired level of
reduction in negative human-bear interactions.
Group model building intervention
As previously stated, we adapted Hines’ (2001) standard method as a framework for
guiding the project team through a model building effort. Our project included a set of on-line
activities (i.e., 4 workshops with the project team) and off-line activities ( i.e., the modeling team
met a number times and worked independently on model development), completed intermittently
over an 18-month period (February 2004 - July 2005). In the following sections we describe the
steps undertaken and insights gained from our modeling process.
Diagnosing the problem
In the first step of our project we convened the project team for a half-day workshop and we
used a set of “scripts” to elicit information about the clients’ perceptions about variables and
relationships of concem. We asked our group of wildlife managers (members of a DEC
management plan team) to think about negative impacts they wanted to manage, and then create
graphs over time for any variable they thought played a role in creating or managing negative
impacts. Team members created 45 different graphs during a scripted exercise called graphs
over time. On these graphs were 33 different variables that we clustered into 10 broad
categories. These variables became the key variables (and boundary for) our model. Figure 1
displays these variables in three sectors familiar to the project team.
Bear Sector Food Sector
Be ae woo
oa Birth rate s 77 Precipitation > .
/ Natural mortality \ 1 Availabilty of natural food
Hunting mortality 1 \ ein of residential food Zz
~ Earing canacty pea
----7 wee eee”
Human-Bear
—_
c samt Sector
Concem about bears»
” Tolerance of bears
in Complaints about bears
Support for hunting
t Prevention behavior 1
Hunting behavior
Experience with bears
\ Knowledge about bears
XN Household density 7
N\ © Agency activities | 4
~~ a
We eo veel
Figure 1. A summary of key variables wildlife managers believed were influencing the level of
negative human-bear interactions in residential areas of New York State.
We would later go on to integrate nine of the ten key variables listed below into a
simulation with six sectors (i.e. bear population, hunters, food, bear-human interactions,
knowledge/interest, and agency resources).
1. Tolerance: Human tolerance of bears.
2. Hunting pressure on bears: Days of hunting; public interest in hunting/participation in
hunting; number of big game hunters; public support for hunting.
3. Bear population: Statewide bear population, partitioned into 3 age and 2 sex cohorts.
4. Sources of bear mortality: Bear harvest; non-hunting mortality.
5. Positive attitudes about bears: Public interest in having bears present; level of public
knowledge about bears.
6. Concerns about bears: Public concern about bears and bear management; perception that
bear problems will occur.
7. Problems with bears: Number of complaints; timing of bear-related problems; number of
negative human-bear interactions; severity of negative interactions.
8. Habitat variables: Natural food availability; human food availability; amount or quality of
bear habitat.
9. Level of DEC activities: Customer service; level of public education provided by DEC; state
research programs; DEC staff working on bear management program.
10. Outdoor recreation in areas occupied by bears: Hiker days in bear range; recreation days
spent in wildemess.
After the first workshop the modelers met independently to synthesize information
gained through several elicitation scripts. Our synthesis of information from workshop #1
allowed us to construct the following problem statement, focused specifically on complaints
about problems with bears in residential areas.
Problem statement: Negative human-bear interaction is increasing in New York,
contributing to an increase in negative impacts. Rise in the number of complaints to
DEC is one indication that negative impacts are increasing.
In the minds of these managers, some set of interventions is necessary to achieve a
desired future (managing negative interactions such that the increase in complaints levels off at a
socially acceptable level).
Without intervention, managers fear that negative interactions and complaints will
continue to rise (Figure 2), with negative consequences for bear management stakeholders, black
bear conservation, and the wildlife management agency.
With the information and knowledge elicited in the first workshop, the modelers drew the
first causal feedback map (Figure 3). We used the causal loop diagram to discuss the model
boundaries and scope of the project with the project team as well as with wildlife experts from
the Department of Natural Resources at Comell University. After a few iterations, the project
team agreed that the causal feedback map reflected their understanding of how the variables are
interconnected.
In workshops with the project team we have not shown the whole map as seen in Figure 3
on one slide, but unfolded the map loop by loop. The advantage of this layered approach was that
we were able to capture individual loops and discuss the meaning and implications of balancing
vis-a-vis reinforcing loops. This script proved to be very useful in getting agreement among the
project team about the cause and effect of complaints from human-bear interactions.
Furthermore, the causal loop diagram also improved the team’s understanding of the complex
interactions in this system.
a Feared
2000 future
1}
a
.
22
#2
£ =
3 g 1000) < My Desired
i Na future
8 Ee
>
1980 1990 2004 2015 2030
Year
Figure 2. Problem statement (desired and feared future states) identified in 2004 by the project
team.
umber of bigl * a
[game hunters| , |
Opportunites for
pear hunting 81
a savetetion
é Perceived P
umber of ound
Fraction of big pial =
game hunters . ‘Awareness of es
Srooting bears ? bear hunting if
opportunities
Theses
Prevention
meberanees pepe)
to harvest. =
Hunting :
Adjustment |
time
B4 @y
effort
# Availability of
@ ieee meen
Number of
bears *
n
‘
; +. Fraction of bears (@s .
pana iesestrlgey vasiton Personal
a Veg andBen80 operons alls
josey
Used to resi
Availability of leone fc
natural food ms
Figure 3. Original causal feedback map developed with the black bear project team.
System boundary and dynamic hypotheses
A synthesis of information obtained through two group model building workshops with the
project team allowed us to develop a set of interrelated dynamic hypotheses. The dynamic
hypothesis (DH) statements shown below were extracted from a conceptual map developed with
the group. In our first workshop, we asked the team to select a number of variables and then
draw the expected behavior of the variable over time (we also asked them to identify the value
for the Y and X axis). Each DH includes hopes and fears about future behavior. Management
interventions are proposed as means to achieve desired future conditions.
Number of bears (DH;): Increase in the bear population leads to an increase in number
of bears attracted to human food, more negative bear-human encounters, and complaints to DEC.
The bear population eventually stabilizes at a level far above social carrying capacity. Reducing
the bear population reduces negative interactions and complaints to DEC.
Hunting (DH): Managers hope opening new hunting areas will increase hunter days,
which would reduce: the number of bears, the fraction of bears attracted to food, negative bear-
human interactions, and complaints to DEC. Inability to expand hunting areas or change season
dates will mean stable or declining hunting pressure (and ultimately, increased complaints).
Negative Interactions (DH3): As human density increases, availability of human foods
increases, the proportion of the bear population attracted to human food sources increases,
negative bear-human interactions increase, and complaints to DEC about residential problems
increase. Controlling access to human food sources reduces the fraction of bears that are
attracted to and then habituated to those foods. That reduces the number of negative interactions
and complaints to DEC.
Public Concerns (DH4): As bears occupy new areas, poor garbage handling attracts
bears to people, residential problems increase, and problems with bears contribute to increased
public concerms about bears. Word of mouth about bear problems can elevate public concem.
Concem levels off after people live with bears and gain personal experience. Personal
experience and effective education interventions raise knowledge of bears, increase rate of
proper garbage handling, reduce number of problems, and reduce public concerns about bears.
Education prevents the elevation in concem from taking place if people only leam from direct
experience.
Tolerance (DH;): As bears occupy new areas, poor garbage handling attracts bears to
people, and residential problems increase. Concern increases due to direct experience and word
of mouth about problems, leading more residents to become intolerant of bears. Managers hope
intolerance will level off as people gain knowledge about preventing problems (or they leam to
live with problems).
Education (DH): Effective education interventions raise knowledge of bear behavior
and increase problem prevention behavior. These things reduce negative interactions and the
fraction of people who complain, which leads to reduction in complaints. Managers’ fear that
education will not occur or will be ineffective, which would mean less change in behavior, less
knowledge gain, and increase in complaints.
Scope of the project
We set a time horizon of 50 years, a value we felt was reasonable to test a number of policies
considering the scope of our project. A time horizon of 5 decades allowed us to simulate: birth
and death of several generations of black bears (average lifespan for bears in New Y ork is 6-8
years); turnover in hunters as they mature and drop out; public reaction to changes in hunting or
education policies; and changes in food availability associated with changes in household
density. Our simulation was not designed to explore the effects of broad habitat changes. The
time horizon we selected is too short to examine such questions.
We used two sets of historic data as reference modes. We have data on complaints about
problems with black bears over the past 15 years. We also have data on annual hunting-related
mortality for bears since 1955. We used these two sources of historical data to evaluate the
model structure and build the project team’s confidence in the model. When we presented the
model to the team, we initialized our model with settings that approximate conditions like those
found in the Adirondack region of New Y ork State.
The problem under investigation is rooted in bear population changes, human behavior
patterns, and residential development that go back several decades. As noted above, we
constructed a simulation that generally reflects historical data on bear take and bear-related
complaints. We were trying to develop a simulation that would generate plausible behavior in
terms of the frequency and amplitude of change in bear take and bear-related complaints (i.e., we
wanted the simulation to produce realistic behavior). However, we were not attempting to build
a simulation that would closely replicate historic data patterns or forecast precise levels of
complaints in any future year (i.e, we were not attempting to develop a model with high
precision).
We developed a stock-flow model with 199 variables (including 16 stocks) and six model
views. We labeled the sectors: bear population, hunters, food, bear-human interactions,
knowledge/interest, and agency resources. We partitioned the model into sectors to better
visualize the structural components of the model, following George Richardson's advice on good
modeling practice. Furthermore, having individual sectors made it easier to communicate with
the project team, since the individual model sectors reflected process maps we used in our
model-building workshops.
Model building process
Introduction to system dynamics
Before we presented first stock-and-flow diagrams to the client group, we introduced the
methodologies of quantitative system dynamics simulation in very broad terms. We felt this short
introduction was necessary to help the client understand the diagrams, which we presented in the
meetings. One argument for using a direct and straightforward approach, presenting relatively
detailed stock-and-flow diagrams, is time efficiency. The disadvantage is that this approach does
not involve the project team in detailed conceptualization of the model structure. However, in an
iterative process the team will always be able to reflect back to the model and make suggestions
for changes.
Model sector views
As previously stated, this paper does not focus on the technical details of the model but on scripts
to communicate with the project team and thus create ownership in the model. The sector views
shown below were used to discuss structural issues with the project team and to get agreement on
the level of detail. We include two sector views as examples, to illustrate the use of stock-and-
flow maps in discussion with the project team. Four other sector views are omitted from this
paper.
We used sector maps as scripts to get agreement on the structural representation of the
model and to create ownership of the team by involving them at an early stage in the model
building process. Group discussion about the sector maps exposed underlying assumptions for
critical reflection. Group discussions helped us clarify connections and make refinements to the
sector views in an iterative process.
Bear_population sector. The model structure for the bear population is rather
aggregated because we only use three age cohorts; cubs, sub adults, and adults. The structure
omits immigration emigration, two variables that can influence the stock labeled total bears.
However, after discussions with the project team it was decided that for the purpose of testing
effects from different harvest levels, the structure as shown in figure 5 would be sufficiently
detailed.
Hunter sector. The hunter sector depicts how the hunting rate is influenced by a number
of variables. We are still at an early stage of the model conceptualization thus some of the
assumptions in this sector need further investigation. For example, the variable “opportunities for
bear hunting” is currently an aggregated view on how DEC regulates hunting rates through
opening up new hunting ground or extending the hunting season. However, adding complexity to
the model at the current stage of the project may not provide more insights into the fundamental
dynamics.
Discussion and reflection
Population modeling is a well used and accepted practice in the field of wildlife management
(Conroy 1993). Dynamic systems models have been developed to explore questions about
populations of grizzly bears (Faust et al. 2004) and black bears (Patton 1997, Freedman et al.
2003, Dobey et al. 2005). Holling (1978), Walters (1986), and others have established a
practice called adaptive resource management, which relies heavily on a systems approach.
Their ideas have been adopted as the basis for a national system (adaptive harvest management)
10
Fractional cub
Camying Capacity
Fractional SAF Average natural
death rate <a fe ceaae Relative bear
<Subadult raction of hunting deat ity
panes effortto hanest
o Per: 2 efectotacnsiy
Liters per bear A <Effectof density cn lifespan
oe on lifespan: Effectof density wo
on lifes pan>
sual teat oP eae Sarria AF natural Total
ee ae Tombaniog hen death rate population:
‘Average cubs Breeding rate bs
liter
pe initial subadult
females
jon per Subaautt -——~, —
Setoton see, eS eae Ee > sciit Female: Hunting Onvort
bin rate female cub SAF Sunil rate ‘AF death from
shnbalne. ~~ bunting rate ‘Searhunting
initial cubs \ Seasons
Total inital AF me
Effectirom density proportionot gy Noa. 3 population fires oi ‘Hunting effortto
on birth rate male cubs ni subadutt inital AM hanestadultbears
initial subadult pe
‘males Fraction of hunting
Subadult 2) perstivates cfforts to harvest adult
Density effect Relative bear Males ‘AM death from bears
ae mie aid fF Sastre sarc
sunivel rate
<SAF death rat
avanatural from hurtin
SAMdeath rate Sivreiral A death rate
from hunting death rate Effectof density ;
con lifespan> sete eesti
‘Sear hunting Fraction of hunting oO
asons> = ———9& effort to harvest <Effectof density ¥ death
subadult on lifespan: Average AM hunt
Reeetione| Sai ratural lifetime
death rate
<tHunting Ovo This rate is oniyto calibrate the model, ie.
compute the harvest rate peryear
<Fracion of hunting
efforts to hanestadult
bears
Se
Figure 5. Sector Map for Bear Population.
aime:
11
Fraction of pote hunters
wg gairingn interest per
rrerth
New hunters eegee
DEC efforts
=Total population>
Policy lever (suggested range
from 0 - 1) adjustable over 600
rronths
scare ine
von = al
— | eee, samo
seis
ous
aN ma Proce ie
Potential Bg | soar achstirent eonate
garme hunters Gnigneee hunters Losing interest in
hic
—_
Contacts: Potential hunter “aworenees of o rice Time to change
er contacts New hunters from nos policy
potential word of mouth 2
fied pena
i aan —s
ae,
Contacts win | __ wan renter FINAL TIME Fractional perceived hurting
Flin Hind, ORES reo vs ea ict lente
felons eres Serkan
Shows row
sehen
wctleaea
cnbeem ati ' -
sens hee emcisematse toma f cmp Raitt e =
SET cnn breatgeaseae
cs ww None, oat
soe
ce
Nc
suepot perme pare
ae
ectceezot
stop Sees we
we Ss
<tousehoids complaining ——— Willingness. pe es
sonore Pe sll
aoe a we
<Households complaining Threshold for Normal effect from Sc ones
about severe interactions p_ hunting support
Figure 6. Model Sector to capture number of hunters and hunting rate.
12
to set harvest regulations for species of migratory waterfowl in the United States (Johnson and
Case 2000).
However, the wildlife management profession has not embraced the full range of
potential applications to which systems thinking and systems modeling might be applied. Most
research and applications to date have focused exclusively on population dynamics, largely
ignoring social variables, such as hunter effort (Jensen 2002). Several scholars have argued that
society stands to gain as much or more benefit from modeling efforts that go beyond population
dynamics, to integrate social or economic variables (Johnson and Case 2000, McKoy 2003) and
involve community members and decision makers in the modeling process (Grant 1998, Sage et
al. 2003). Sage et al. (2003) provide an example model (using STELLA software) with a “flight
simulator” that allows the operator to learn about white-tailed deer population dynamics with
simulations for different levels of habitat quality, predation, and hunting pressure. Our work is
intended to provide another demonstration project, showing wildlife managers how they can use
systems thinking and systems modeling to increase understanding about the dynamic complexity
of a wildlife management issue. In a 1998 manuscript, W. E. Grant challenged natural resource
management professionals to “be more assertive” in promoting use of a systems approach to
environmental education, ecological research, and natural resources management (Grant
1998:74). He states, “I firmly believe the only way to deliver the systems message effectively is
through apprenticeships in which participants are involved actively in the process of developing
and using system dynamics models under supervision of experienced modelers” (Grant 1998:74).
The project we report here is just that: an effort to provide wildlife management professionals an
apprenticeship experience with assistance from an experienced modeler.
While exercising the model and providing insights to the team is a means to an end,
modeling and its iterative process is a learning opportunity for the team as well as the modelers.
Using a well grounded framework to guide a team through a modeling process is imperative to
ensure that (a) the team takes ownership of the model and (b) it instills confidence when testing
policies.
Our final workshop took place in July 2005. We will soon complete the policy testing
phase of this project and then we will work with the project team to formulate project
conclusions and action recommendations. We are encouraged by our early efforts to build
system models with teams of wildlife management professionals. Our experience to date
suggests that group model building work, using the standard method, holds promise as a means
to help teams of wildlife managers gain a deeper understanding of the complex interactions in
the systems they strive to manage. We have found that scripts for group model building used in
this project also have communicated well with other wildlife managers not involved in the black
bear model building process. Concurrent with the black bear project, we have initiated
qualitative systems thinking exercises with wildlife managers focused on quality deer
management, beaver management in residential areas, and wildlife disease management in
national parks. Managers in each of these contexts have responded positively to scripts that
produce concept maps articulating long-held assumptions.
Before the third model building workshop, we asked members of the project team to
complete a questionnaire with questions about their assumptions and expectations related to the
13
project. Now that all model building workshops are complete, we will ask members of the
project team to complete a post-process questionnaire assessing key beliefs and attitudes about
the model-building intervention and about the most appropriate actions to manage increasing
complaints about black bears in New Y ork. We developed the pre- and post-process assessment
instruments based on the work of Rouwette (2003). We will use information from those written
responses and personal interviews completed with project participants to assess changes in
beliefs and attitudes, consensus about problem definition, and consensus about how the agency
should respond. More rigorous assessment of insights gained and learning outcomes will be
possible when that work is completed.
Acknowledgments
For their participation in the group model-building intervention, we thank the following
personnel in the New York State Department of Environmental Conservation, Bureau of
Wildlife: Lou Berchielli, Larry Bifaro, Chuck Dente, Greg Fuerst, Steven Heerkens, Dick Henry,
John Major, Matthew Merchant, John O’Pezio, Ed Reed, and Dave Riehlman.
For contributions to simulation construction and validation, we thank Shawn Riley
(Michigan State University), Chuck Nicholson (Cornell University), and Paul Newton (Comell
University), and George Richardson (University at Albany).
Several members of Comell University’s Human Dimensions Research Unit in the
Department of Natural Resources contributed to this study. Dan Decker and Jody Enck made
valuable contributions to model conceptualization. Meredith Gore and Julie Weber contributed
to identification of secondary data and literature sources for model calibration.
Funding for this project was provided by New Y ork Federal Aid in Wildlife Restoration
Grant WE-173-G Job 146-III-3b, the Cornell University Agricultural Experiment Station, and
the Comell System Dynamics Network (CSD Net).
References
Conroy MJ. 1993. The use of models in natural resource management: prediction not
prescription. Transactions of the North American Wildlife and Natural Resources
Conference 58: 509-519.
Dobey S, Masters DV, Scheick BK, Clark JD, Pelton MR, Sunquist ME. 2005. Ecology of
Florida black bears in the Okefenokee-Osceola ecosystem. J ournal of Wildlife
Management Monographs 158.
Faust LJ, Jackson, R, Ford A. 2002. Models for management of wildlife populations: lessons
from spectacled bears in zoos and grizzly bears in Y ellowstone. System Dynamics
Review 20(2): 163-178.
Freedman AH, Portier KM, Sunquist ME. 2003. Life history analysis for black bears (Ursus
americanus) in a changing demographic landscape. Ecological Modeling 167 (2003): 47-
64.
14
Gore ML. 2004. Comparison of intervention programs designed to reduce human black bear
conflict: a review of literature. HDRU Series Publication No. 04-4. Department of
Natural Resources, Comell University: Ithaca: New Y ork.
Grant WE. 1998. Ecology and natural resource management: reflections from a systems
perspective. Ecological Modeling 108 (1998): 67-76.
Holling CS (ed). 1978. Adaptive environmental assessment and management. John Wiley &
Sons: New Y ork.
Hines J. 2001. Course material: “The “standard method.” Sloan School of Management,
Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA
02139-4307.
International Association of Fish and Wildlife Agencies (IAFWA). 2004. Potential costs of
losing hunting and trapping as wildlife management tools. Animal Use Issues
Committee, IAFWA: Washington DC.
Jensen AL. 2002. Analysis of harvest and effort data for wild populations in fluctuating
environments. Ecological Modeling 157 (2002): 43-49.
Johnson, FA, Case DJ. 2000. Adaptive regulation of waterfowl harvests: lessons
learned and prospects for the future. Transactions of the North American Wildlife
and Natural Resources Conference 65: 94-108.
Lackey BK, Ham SH. 2003. Contextual analysis of interpretation focused on human-black bear
conflict in Yosemite National Park. Applied Environmental Education and
Communication 2: 11-21.
McKoy, NH. 2003. Behavioral externalities in natural resource production possibility frontiers:
integrating biology and economics to model human-wildlife interactions. Journal of
Environmental Management 69(2003): 105-113.
Morecroft, JDW, Sterman JD (eds). 1994. Modeling for learning Organizations. Productivity
Press: Portland, OR (Now available from Pegasus Communications, Waltham, MA).
NYSDEC. 2003a. Black bears in New York: natural history, range, and interactions with
people. New York State Department of Environmental Conservation: Albany, New
Y ork.
NYSDEC. 2003b. A framework for black bear management in New York. New Y ork State
Department of Environmental Conservation: Albany, New Y ork.
Otto P, Struben J. 2004. Gloucester Fishery: Insights from a group model building intervention.
System Dynamics Review 20(4): 287-312.
Patton, BC. 1997. Synthesis of chaos and sustainability in a nonstationary linear dynamic model
of the American black bear (Ursus americanus Pallus) in the Adirondack Mountains of
New York. Ecological Modeling 100 (2003): 11-42.
Richardson GP, Andersen DF. 1995. Teamwork in group model building. System Dynamics
Review 11(2): 113-137.
Richardson GP, Pugh AL III. 1981. Introduction to System Dynamics Modeling with DYNAMO.
MIT Press: Cambridge, MA (Now available form Pegasus Communications, Waltham,
MA).
Riley SJ, Decker DJ, Carpenter LH, Organ JF, Siemer WF, Mattfeld GF, Parsons G. 2002. The
essence of wildlife management. Wildlife Society Bulletin 30(2): 585-593.
15
Riley SJ, Siemer WF, Decker DJ, Carpenter LH, Organ JF, Berchielli LT. 2003. Adaptive
Impact Management: An Integrative Approach to Wildlife Management. Human
Dimensions of Wildlife 8; 81-95.
Roberts NH, Andersen DF, Deal RM, Grant MS, Shaffer WA. 1983. Introduction to Computer
Simulation: The System Dynamics Modeling Approach. Addison-Wesley: Reading, MA.
Rouwette EAJ. 2003. Group model building as mutual persuasion. Wolf Legal Publishers:
Nijmegen, The Netherlands.
Sage RW Jr., Patton BC, Salmon PA. 2003. Institutionalized model-making and ecosystem-
based management of exploited resource populations: a comparison with instrument
flight. Ecological Modeling 170 (2003):107-128.
Schusler T, Siemer WF. 2004. Report on stakeholder input groups for black bear impact
management in the Lower Catskills, Upper Catskills, and Westem New York. Comell
Cooperative Extension of Tompkins County: Ithaca, New Y ork.
Siemer WF, Decker DJ. 2003. 2002 New Y ork State Black Bear Management Survey: Study
Overview and Findings Highlights. HDRU Series Publication 03-6. Department of
Natural Resources, Comell University: Ithaca, New Y ork.
Vennix, JAM. 1994. Building consensus in strategic decision-making: insights from the process
of group model-building. Paper presented at the 1994 Intemational System Dynamics
Conference, Stirling, Scotland.
Vennix, JAM. 1996. Group model building: Facilitated Team Learning Using System
Dynamics, Wiley & Sons: New Y ork.
Vennix, JAM, Andersen DF, Richardson GP. 1997. Introduction: Group model-building— Art
and science. System Dynamics Review 13(2): 103-106.
Walters, C. 1986. Adaptive Management of Natural Resources. MacMillan: New Y ork.
16
CC BY-NC-SA 4.0