Integrating Socioeconomic Factors into Gap Analysis Products
Applications of Urban Growth Models
and Wildlife Habitat Models to Assess
Biodiversity Losses
FINAL REPORT
30 October 2001
Christopher B. Cogan, Researcher
Ph.D. candidate, Environmental Studies,
University of California-Santa Cruz
Frank W. Davis, PI
Donald Bren School of Environmental Science and Management
University of California-Santa Barbara
Keith C. Clarke, Co-PI
Dept. of Geography, University of California-Santa Barbara
Contract Administration Through:
University of California-Santa Barbara
Institute for Computational Earth System Science
Submitted by:
Christopher B. Cogan
Research Performed Under:
Cooperative Agreement No. 00HQAG0009 as part of the GAP Socioeconomic Seed
Grant Program, managed by Dr. Gary Machlis
U.S. Department of the Interior
U.S. Geological Survey
Biological Resources Division
Gap Analysis Program
Applications of Urban Growth Models and Wildlife Habitat Models to
Assess Biodiversity Losses
ABSTRACT
Habitat loss and subsequent fragmentation due to urban development is part of a
larger suite of anthropogenic impacts on biodiversity, but it now ranks among the
principle causes of species endangerment in the United States. Several types of urban
growth simulation models have been developed which can supply useful information for
biodiversity planning. In many cases however, the data required for biodiversity planning
may not be compatible with the urban models, leading to analytical inaccuracies and
misleading conclusions. Here, we examine several lines of logic likely to be employed in
biodiversity assessment and show how assumptions built into the data influence model
outcome.
Introduction
Biodiversity can be described conceptually as a collection of indices, ranging from
diversity of compositional, structural, and functional biotic elements (Noss 1990), to
ecosystem, species, and genetic diversity (Soulé and Mills 1998). Representing such a
broad ecological concept requires the combination of several models (Cogan in press),
and central amongst these are predictions of species habitat quality and quantity. Habitat
loss or fragmentation due to urban development is only one of many anthropogenic
impacts on biodiversity today (Tilman and Lehman 2001), but it now ranks among the
principle causes of species endangerment in the United States (Dobson et al. 1997,
Vitousek et al. 1997, Wilcove et al. 1998).
Urban growth models developed recently, incorporate a broad range of variables and
spatial dependencies (Makse et al. 1995, Couclelis 1997, White et al. 1997, Batty 1998,
Clarke and Gaydos 1998, Landis and Zhang 1998, Makse et al. 1998). These models do
not themselves attempt to determine the environmental consequences of future
urbanization. However, by providing a localized abstraction of the direction and
magnitude of land use change, they offer a starting point for assessing future impacts on
biodiversity. Planning efforts are not currently making full use of basic biodiversity
information (Press et al. 1996, Crist et al. 2000), and improved linkages between urban
growth models, biodiversity models, and land use planning is urgently needed. By using
predictive models of urbanization and its effects on biodiversity, county planners and
other stakeholders will be able to visualize and evaluate different future growth scenarios
as an effective way to lessen the impact of urbanization on biodiversity (e.g., Landis
2000) Biodiversity assessments often rely on ecological models that simplify and
abstract the biophysical world based on a series of assumptions about ecosystem
functions. The set of assumptions may not be clearly understood by the model user and
this can lead to inappropriate applications of the models. Problems of generalization and
appropriate model selection are further compounded as more models are used in
combination. When planning for both urban growth and biodiversity conservation today,
a planner may be faced with the task of using two potentially complex types of models:
one of urbanization processes and another of ecological processes.
Many urban growth models use a simple binary classification of land use into urban
or non-urban, although in fact urbanization includes a wide range of settlement patterns
and human densities. Perhaps most significant for biodiversity conservation is the
phenomenon of urban and suburban sprawl at the margins of existing metropolitan areas.
Because the sprawl is incremental, the loss of habitats in these “front line” areas can be
difficult to control. The boundary between urban and non-urban is fuzzy at best. Similar
challenges are involved in predicting the impact of urbanization on biological species,
involving simple habitat classification schemes and crude habitat suitability rating
systems to predict whether a given land use or land cover class is or is not suitable habitat
for a species.
As a first step, urban growth models sensitive to increasing human population
density and urban expansion in rural regions can be usefully combined with biodiversity
models. Since even relatively sparse development on a parcel-by-parcel basis can
dramatically affect biodiversity, the growth models should be able to detect change over
fairly small spatial areas (e.g. 100 meter grids). The biodiversity models should likewise
be sensitive to land use change at a similar spatial grain. Even with the constraints of
generalization and spatial grain compatibility, information gained from such a union of
predictive urban and biodiversity models will be valuable in helping to anticipate and
avoid biodiversity erosion caused by habitat loss and fragmentation.
In this report, we examine several lines of logic likely to be employed in the
combination of models used to plan for urbanization and biodiversity conservation. We
also show how assumptions inherent to the data can influence model outcome. Examples
are presented from a case study in Santa Cruz County, California, where multiple urban
growth scenarios are combined with land cover data, and wildlife habitat relationship
(WHR) models.
Methods
Habitat quality and quantity aspects of biodiversity were examined using three
principle inputs: urbanization scenarios, wildlife habitat maps, and species habitat
models. Output from the analyses is reported as loss of habitat area, or in some cases, in
terms of impact to the vertebrate species associated with degraded habitats. Limited data
availability did not permit an impact analysis for invertebrate species.
A flow chart of the models and analyses provides an overview of the biodiversity
sensitivity analysis (Figure 1). Three different previously developed models for
predicting patterns of urban expansion were tested. The three models included the “urban
buffer” (see below), “Landis” (Landis and Zhang 1998) and “Clarke” (Clarke and
Gaydos 1998) scenarios. Outputs from the different growth models were then used in
conjunction with coarse grain (100 ha minimum mapping unit) land cover maps from the
California Gap Analysis Project (GAP, Davis et al. 1998).
The Landis and Clarke models were also used with a finer grain (1 ha) land cover
data set. This map layer was commissioned by the Association of Monterey Bay Area
Governments (AMBAG) based on 30-meter Landsat Thematic Mapper (TM) imagery.
Spatial distributions of 21 individual vertebrate species predicted to occur in the study
area were made possible by applying wildlife habitat relationship (WHR) models (Airola
1988) to the courser grained GAP land cover data. Potential impacts of urban growth to
these species were explored by intersecting scenarios of future urban growth from each of
the three models with the WHR-based predicted distributions of the species. The onehectare finer grained land cover data was intersected with output from the three urban
growth models to generate impact assessments for generic habitat types, such as “coastal
oak woodland”, without evaluating potential to individual vertebrate species from WHR
(Fig. 1).
urban buffer growth
scenario
Landis growth scenario
GAP 100 ha land cover
Clarke growth scenario
County 1 ha land cover
WHR models
vertebrate species impacts
habitat impacts
Figure 1. Flow chart for biodiversity sensitivity analysis. Three urban growth scenarios
and two land cover models combine to evaluate vertebrate and habitat impacts in Santa
Cruz County, California.
Urban Growth Models:
As a case study, we have employed three different urban growth scenarios, each
portraying a possible future urbanization pattern in Santa Cruz County, California, USA.
All three scenarios are based on the same initial urbanization pattern from the California
Gap Analysis Project, derived largely from Landsat TM satellite imagery (Davis et al.
1998). The Gap urban data have a coarse (100 hectare) spatial grain, which can
potentially cause problems when analyzing some land cover effects (e.g. absolute
measures of fragmentation). In this study, we use the Gap urban data only for urban
model starting points and relative area comparisons.
The first growth scenario is based on a simple spatial buffer, which is generated by
expanding current urban land use areas outwards by a distance of 500 meters. The 500meter forecasted growth area appears as a narrow red band around the current urban areas
(Figure 2).
A second growth scenario is based on a model of urbanization developed by Landis
and Zhang (1998), which incorporates socioeconomic and physical data to predict areas
of future urbanization. Using logit (natural log of the odds ratio) models of historical land
use change, 100-meter grid cells are predicted to be urban or non-urban in a future time
period. Probability and magnitude of land-use change is predicted from projected
population growth individually modeled for each city and county in the study area. With
population growth specifying the demand levels for urbanization, a series of six
independent variables is used to forecast which areas will be developed. The variables
are: 1) initial site use; 2) demand factors, i.e. employment data; 3) accessibility, i.e.
commute distance; 4) cost constraints, i.e. development costs for the site, availability of
services; 5) policy constraints, i.e. zoning categories; 6) adjacent-use effects and
proximities, i.e. industrial neighbors or nearby shopping centers.
Current Urban Landuse
Forecast Urban Landuse
Other Landuse Areas in Santa Cruz County, California
KILOMETERS
0
10
10
Data Source: CA GAP
5
0
5
MILES
10
20
15
Figure 2. Urban growth forecast using an urban buffer model. Santa Cruz County,
California.
The combination of population growth models with site-specific spatial allocation
models uses large amounts of data to characterize the spatial, political, economic, and
historic circumstances of the local study area. Output from the model can be in table form
or in map form. The Landis model results for Santa Cruz County were transformed into a
geographic information system (GIS) map (Figure 3), depicting the pattern of
urbanization used in this case study. The Landis model can generate alternative outcomes
depending on adjustments to the input parameters, though variations are not temporally
explicit. For a detailed discussion of the model, see Landis and Zhang (1998).
The third urban growth scenario used a cellular automata approach developed by
Clarke (1998) to predict a year-by-year sequence of growth based upon physical
landform and land use data. Model inputs included road locations, presence of existing
urban areas, slope, and protected exclusion areas. The Clarke model uses a series of precalibrated control values, which are self modified with each theoretical year of growth.
Initial calibration is based on historic trends. Output of the model is an image of forecast
growth areas. The digital image was converted to GIS grid map format (Figure 4). As
with the Landis model, a 100-meter grid template was used to map the presence or
absence of urban land cover. In contrast to the Landis model, the physical data inputs
were largely derived from available remote sensing products, using a combination of
aerial photography and satellite imagery. The Clarke model is spatially and temporally
explicit, scaleable to any particular spatial grain or extent, and does not utilize
socioeconomic data.
Current Urban Landuse
Forecast Urban Landuse - Landis Model
Other Landuse Areas in Santa Cruz County, California
KILOMETERS
0
10
10
Data Source: CA GAP
5
0
5
MILES
10
20
15
Figure 3. Urban growth forecast using the Landis model. Santa Cruz County, California.
Current Urban Landuse
Forecast Urban Landuse - Clarke Model
Santa Cruz County, California
KILOMETERS
0
10
10
Data Source: CA GAP
5
0
5
MILES
10
20
15
Figure 4. Urban growth forecast based on the Clarke model. Santa Cruz County,
California.
Landover Data:
Two different land cover datasets provide an opportunity to assess changes in model
output resulting from differences in the land cover parameters. The coarser of the two
land cover datasets is the GAP land cover GIS map layer (Figure 5). This map layer was
originally created by human photo interpretation of Landsat TM imagery with a
resampled pixel size of 100 by 100 meters (1 ha). The GAP interpreters also used a
variety of ancillary data, including historic vegetation maps dating from the 1930’s. GAP
land use and land cover polygons were delineated with a 100-hectare minimum size
generally referred to as the minimum map unit (MMU). The 100 ha MMU is coarse
compared to many other map products, however it is the most detailed statewide land
cover product for California. For more information on the development and appropriate
use of the California GAP data, see Davis et al. (1998).
One aspect of the GAP land cover data is particularly useful for biodiversity
analysis. Each land cover polygon is linked to a list of vertebrate species, based on the
California Wildlife Habitats Relationship (WHR) system (Airola 1988, Davis et al.
1998). With this linkage, each land cover polygon in the GAP database can be associated
with habitats for particular species. Using standard GIS techniques, we used urban
growth model outputs in conjunction with GAP land cover to determine which habitats
and which vertebrate species may be impacted in the near future.
KILOMETERS
0
10
10
Data Source: CA GAP
5
0
5
MILES
10
20
15
Figure 5. California Gap Analysis Project (GAP) land cover polygons in Santa Cruz
County, California.
Another land cover map is also available for Santa Cruz County. This GIS map layer
was commissioned by the Association of Monterey Bay Area Governments (AMBAG)
based on the same 30 meter TM imagery the GAP project used. In this case however, the
land cover partitions were machine classified, resulting in a product with a 30 meter
MMU (1/9 ha), and approximately 40 land cover classes (Figure 6). The AMBAG cover
classes are not directly compatible with the WHR system. In parallel to the analysis using
the GAP products, the AMBAG data was also combined with the urban growth models,
for investigation of impacted habitats following urbanization.
The combination of urban growth data and land cover data was modeled in two
ways. As a measure of vertebrate species impact, when any portion of a land cover
polygon was predicted to become urbanized, the entire polygon was assumed to be
compromised for that species. As a measure of habitat impact, the exact proportion of the
land cover polygon predicted to become urbanized was calculated for the analysis of
habitat loss.
AMBAG landcover polygons
Santa Cruz County, California
KILOMETERS
0
10
10
Data Source: AMBAG
5
0
5
MILES
10
Figure 6. AMBAG land cover polygons in Santa Cruz County, California.
20
15
Results
Urban Growth Models:
From the broad county perspective, the Landis and Clarke urban growth model
scenarios had some similarities in their predictions of future urbanization in Santa Cruz
County, California. Both models avoided the steep rural areas in the northwest and
northern ridge areas, and typically selected sites closer to preexisting urban land use.
Several agricultural areas in the south were selected by both models. There were also
several key differences between the Landis and Clarke model results. The Clarke cellular
automata approach (Figure 4) usually forecast urbanization to occur in areas within 500
meters of current urban land use, as well as other regions near major roads. Very little
development was forecast for remote coastal areas, in spite of the presence of roads and
flat terrain. The Landis model did not predict correspondingly high levels of roadside
expansion, but did forecast large amounts of isolated new development in areas not
contiguous with existing urban land use. The buffer model may not appear to be realistic
as a growth scenario, however the land areas identified by this approach were
consistently targeted by the other two models. Thus, the buffer model represents a
simplified starting point for predicted urbanization, without using the model parameters
of more sophisticated approaches.
Land Cover Models:
The two different models of land cover; one from GAP and a second from AMBAG
produced useful data for comparison. A few simple statistics on these data models reveal
important differences that affect the biodiversity analysis. The GAP land cover data are
spatially coarser, with a minimum map unit of 100 ha, and a maximum polygon size of
approximately 4,000 ha. The GAP data describe Santa Cruz County using 208 polygons
in 14 primary classes. Secondary and tertiary classes describe areas of mixed land cover
smaller than the MMU.
In contrast, the AMBAG data uses a much finer spatial grain, derived directly from
30 meter Landsat Thematic Mapper Satellite imagery. These data have a minimum map
unit of 1/9 ha, and a maximum polygon size of 49,000 ha. The AMBAG data use
approximately 11,000 polygons to map Santa Cruz County in 40 classes, with no
secondary classification.
Habitat Impacts:
Habitat impacts were assessed in terms of land area lost to urbanization using the
three separate growth models. This comparison is based on the AMBAG 30 meter MMU
land cover data. For each of the 30 impacted habitat types, the sum of area lost is shown
as a bar graph (Figure 7). Maximum impact is predicted using the Clarke model, resulting
in over 10,000 ha of Redwood / Douglas-fir habitat converted to urban land cover.
Minimum impact occurred with the 500 meter buffer model, though for a few habitats
this trend was inconsistent. Also shown in the graph (Figure 7), are the large differences
in land area and the range of inconsistencies among the three growth models. The large
range of variability necessitated the use of a log scale on the y-axis.
10000
Clarke Model
Landis Model
Area Urbanized (ha)
(log scale)
1000
Buffer Model
100
10
Re
dw
Gr ood
Co
as / D
s & ou
as
t al
Co glas
Oa
as F
k W Co
Co tal O ir D
a
o
oo sta
a
dl l O stal ak W min
an ak O
a
d & W ak ood nt
Re
Re ood Wo land
dw
dw la od
oo
oo nd lan
d/
d
d/ &
Do
Do G
ug
ug rass
las Spa
las
Fi rse Oth
F
Co r & ly V er A G ir
as Co eg g ra
tal as et ric ss
O ta at u
Or ak l O ed/ lture
ch W ak Fa
ar oo W llo
d
d
Kn s/O lan oodl w
ob utd d & and
co oo
ne r N Shr
u
Pa Pine urse b
Sh
lu
Vi rub
str Do ries
ne
m
&
i
n
Eu ya
e E ina
ca rds Coa
m nt
ly
pt /Bu stal Str erg
us sh O aw en
& es ak
b t
Co (eg W erri
as . R oo es
tal a dla
Oa spb nd
lo
lo
w Euc k W errie
w
de aly oo s)
de
ns
ns
ity ptus dlan
ity
ur & d
ur Po
ba G
ba nd
n & ras
n & er
Gr s
Sh C osa
as
ru oa Pi
s
b
s
n
W
t
Re & al e D a
dw Re Oa om ter
oo dw k W in
d/ oo
Do d ood ant
ug /Do lan
Eu las F ugla d
Co
ca i r s F
as
tal
Gr lypt & S ir
as us hru
Oa
s& &
b
kW
Eu Shr
oo
ca ub
dl
l
an
yp
d&
tu
s
Eu Shr
ca ub
l
y
Go pt
lf u s
C
Eu ou
ca rse
ly
pt
us
1
Figure 7. Comparison of three growth scenarios in Santa Cruz County California: 500
meter urban buffer, Landis growth model, and Clarke growth model. Habitat data are
from AMBAG 30 meter land use / land cover maps. Habitat classes are rank ordered
based on the results from the Clarke model.
Vertebrate Impacts:
Native vertebrate impacts were assessed using the GAP 100 ha MMU land cover
data. As with the habitat assessment, the maximum impact case is predicted when using
the Clarke model (Figure 8). Percent habitat loss is based on recent (1993), not historic
species habitat in the ecoregion. All three growth models predict vertebrate species
impacts in similar rank order, with Vaux’s swift (Chaetura vauxi) most heavily impacted
(38% under the Clarke model) followed by hermit warbler (Dendroica occidentalis) and
golden-crowned kinglet (Regulus satrapa). The Landis and Clarke models also have
overall similar magnitudes of predicted impact, while the 500 meter buffer model
predicts up to 30% less impact and does not clearly distinguish among the top 12 species.
40
35
30
Habitat
Loss
(%)
Clarke Model
Landis Model
500 m buffer
25
20
15
10
5
go
Va
u
lde her x's s
n-c mit wi
row wa ft
ne rble
dk
r
pin ingle
M
t
e
ac
gil shr siski
e
l
n
oli ivr wmo
a
ve
-si y's
le
Tr ded warb
ow fly le
bri ca r
Ca
tc
d
lif
orn blac ge's her
k
sh
i
no a gia salam rew
rth nt
ern sal and
e
a
sh allig man r
arp at
d
-sh or l er
inn iza
r
he ed h d
a
r
mi
w
wh
tt k
ite
-cr ru hrus
ow bbe h
ne
rb
d
o
so spar a
ch
l
r
i
est
tar ow
nu
t-b wi y vir
eo
a c nt
Am ked er w
r
c
e
e
Ca
ric hic n
lif
an kad
orn
g
ia
bro old ee
f
sle
w
nd
n c inch
er
r
sal eep
am er
an
de
r
0
Figure 8. Comparison of three growth scenarios in Santa Cruz County California: 500
meter urban buffer, Landis growth model, and Clarke growth model. Species and habitat
data are from the California Gap Analysis Project (GAP). Habitat classes are rank
ordered based on the results from the Landis model.
Discussion
The species habitat analysis presented here is a close examination of one major
factor in the assessment of biodiversity. Other biodiversity elements such as ecoregional
analysis, restoration potential, special features, and habitat shape are also important,
though these are not specifically addressed in this study. We used three urban growth
models and two land cover maps in a relatively straightforward combination of data
(Figure 1) to compare measures of habitat and vertebrate impacts. Here, habitat impacts
are considered to be actual habitat areas converted to urban land use. For example, if a
1000 ha forest is reduced to 900 ha after urbanization the habitat loss is 10%. If the same
forest is reassessed in terms of native vertebrate habitat, it may be more important to
consider buffer distances from impacts, non-linear predation effects, and other complex
landscape metrics. These more specific approaches can be valuable in some instances,
however, when applied to a regional study with many species the results can be
misleading. Stated differently, it is challenging to model disturbance effects as
realistically as possible, while working with a group of dissimilar species over a broad
area.
The approach to vertebrate habitat assessment presented here assumes that if a highly
intrusive land use such as urbanization enters a habitat patch, then the entire patch is
likely to be compromised in terms of vertebrate species habitat quality. In some
instances, this assumption may overemphasize the impact of urbanization. On the other
hand, it is also likely that urbanization effects are underemphasized in cases where urban
expansion approaches (but not actually enters) a habitat area. An alternate model could
employ spatial buffers to model the neighborhood effects of urbanization, however this
approach introduces additional complexities such as splitting map polygons, and imposes
the need for species-specific analysis. Both the habitat and species types of impacts are
important, however it is necessary to clarify the conceptual differences between habitat
and vertebrate impacts when evaluating or discussing urban growth impacts. A
noteworthy example of this distinction was shown in a study of mesopredators and avian
prey in Southern California (Crooks and Soulé 1999).
The methods used in this analysis are based upon an underlying logical sequence
most simply presented as a flow chart (Figure 9). A central assumption here is that
different urban growth patterns should have measurably different biodiversity impacts.
As with any metamodel, it is also important to ensure that the data and various
component models are compatible for integrated analysis. It is often illuminating to
investigate where the logic of a scientific investigation might become unsound, as well as
where it is strong. The logical flowchart outlines key junctions where this type of
biodiversity assessment might face impediments and offers explanations and
recommendations for each situation.
Biodiversity Analysis
Variations in
urban growth
patterns are
not critical in
biodiversity
analysis.
Explanation:
Particular
species will
always be
impacted –
perhaps due to
their rarity in
the county vs.
the ecoregion.
Action: Treat
these species
and habitats as
special cases;
use the
biodiversity
model to
evaluate the
remaining
biodiversity
elements.
Variations in urban growth
patterns do impact biodiversity.
Model error prevents variation
in growth pattern from
producing a measurable
biodiversity response.
Urban growth
scenarios are
constrained
into similar
patterns.
Explanation:
Urban
models lack
sufficient
realistic
variation.
Action:
test with
different
or random
growth
scenarios.
Variation in growth
is measurable in
terms of
biodiversity.
Biodiversity data
are too coarse to
respond to fine
urbanization
differences.
Explanation:
Habitat models are
too coarse grained
for measurable
response to urban
change.
Action: use as is
for coarse grain
analysis, but use
finer grain habitat
model and new
WHR models for
fine biodiversity
analysis.
Explanation:
model is
working with
available
data.
Action: use
urban growth
scenarios and
existing
species
habitat data
to evaluate
biodiversity
impacts.
Figure 9. Logical flow chart for the evaluation of biodiversity analysis with urban growth
models.
Given perfectly accurate biodiversity and urban growth models, lack of biodiversity
response will still occur if the two models are not spatially or thematically compatible.
One indicator of this type of incompatibility can be seen in the comparison of vertebrate
habitat losses following different urbanization scenarios (Figure 8). One interpretation of
this result suggests that vertebrate impacts are much the same following either the Clarke
or the Landis models. Indeed, it seems remarkable that the rank order of species and even
habitat impacts is so similar under two independent and seemingly different growth
models. It would seem to require a radically different growth model like the simplistic
500-meter buffer to produce a significantly different outcome. Another, perhaps more
likely interpretation is also possible. If the GAP data on wildlife habitat relationships is
spatially coarser that the growth models, our ability to differentiate between the Landis
and Clarke models will be diminished. In support of this hypothesis, the appearance of
the map products, and (most importantly) the habitat impacts (Figure 7) indicate
substantial differences between each of the three urban models.
The balance of spatial grain and thematic detail is an important consideration when
producing and using maps of land cover for use in biodiversity analysis. Using the
AMBAG 30 meter MMU land cover map (Figure 6), the fine map grain results in
relatively large areas (up to 49,000 ha) to be mapped as contiguous albeit marginally
connected patches. At slightly coarser map grains, many of the corridors of connecting
habitat would merge into other classes resulting in a very different dataset for the habitat
modeler. This example illustrates how fine grain maps with coarse thematic detail can
overemphasize habitat connectivity. In this case, the assumption that urban disturbance
on the edge of a habitat patch impacts the entire patch becomes tenuous when using fine
spatial grain, coarse thematic grain data such as the AMBAG 30 meter land cover map.
As 100 meter or finer grain urban growth models gain acceptance as a reasonable spatial
scale to model the biodiversity land use complex, more research is needed to ascertain the
appropriate levels of thematic resolution in land use and land cover mapping.
There are several difficulties associated with measuring regional urban impacts on
vertebrate species. The model presented here uses polygons of habitat to represent
potential distributions of vertebrate species, and assumes that analysis of divided
polygons is not a valid application of the data. Detailed study of a specific divided habitat
polygon is possible given appropriate species-specific data, however this local approach
will not be effective regionally. Urban development is sometimes seen as a continuous
creeping of small steps whereby each development project in isolation is difficult to
assess for regional biodiversity impact. The species assessment method presented here
uses habitat polygons to model impacts, effectively dealing with the “urban creep” issue
while maintaining biologically meaningful area units. The complementary combination
of a discrete species metric along with a continuous habitat model is a powerful and
much needed approach.
As biodiversity models such as those discussed here evolve and build in complexity,
our land cover maps and wildlife habitat relationship models will be pressed to deliver
more information with higher quality standards. Some of our data sources have already
evolved from simple maps of predicted species location to become temporally dynamic
models of predicted species connectivity and spatial pattern. Unfortunately, most of our
current maps are not up to this advanced standard. Like most modelers, cartographers
have long known that the design constraints of producing the best habitat maps will
depend on the specific questions being asked of the data. This fundamental principal is
sometimes obscured or overlooked when we allow technological capabilities such as
satellite sensor resolution and radiometric spectral response to overly influence our
understanding of habitat classification and vertebrate distribution.
This paper has outlined a method for assessing habitat and species degradation given
different development scenarios. The inherent assumptions and complexities of this type
of analysis were also explored. Examples have been presented which show how
assessment of biodiversity following urbanization may be misleading. We have also
shown how seemingly significant differences in urban growth pattern may be obscured
by incompatible habitat data. These findings are presented to facilitate an improved
understanding of habitat and species impact models, and to provide direction for future
land use and land cover mapping. The specific models discussed here are important
elements of more generalized biodiversity assessments, which are continually improving
our understanding of biodiversity and promise to provide additional guidance to
minimize the disruptive impacts of urbanization and development.
Acknowledgments:
We thank Bob Johnston, Mike Jennings, Mary Anne Van Zuyle, and Uta Passow for
reviewing earlier versions of this paper and providing constructive comments. Their
insights were greatly appreciated. This work was partially funded by the United States
Geological Survey, Cooperative Agreement No. 00HQAG0009 as part of the GAP
Socioeconomic Seed Grant Program, managed by Dr. Gary Machlis.
References:
Airola, D. A. 1988. Guide to the California wildlife habitat relationships system. Jones
and Stokes Associates, Sacramento, CA.
Batty, M. 1998. Urban evolution on the desktop: simulation with the use of extended
cellular automata. Environment and Planning A 30: 1943-1967.
Clarke, K. C., and L. J. Gaydos. 1998. Loose-coupling a cellular automaton model and
GIS: long-term urban growth prediction for San Francisco and
Washington/Baltimore. International Journal of Geographical Information
Science 12: 699-714.
Cogan. in press. Biodiversity conflict analysis at multiple spatial scales. in J. M. Scott,
ed. Predicting Species Occurrences: Issues of Accuracy and Scale. Island Press,
Washington, D. C.
Couclelis, H. 1997. From cellular automata to urban models: New principles for model
development and implementation. Environment and Planning B-Planning &
Design 24: 165-174.
Crist, P. J., T. W. Kohley, and J. Oakleaf. 2000. Assessing land-use impacts on
biodiversity using an expert systems tool. Landscape Ecology 15: 47-62.
Crooks, K. R., and M. E. Soulé. 1999. Mesopredator release and avifaunal extinctions in
a fragmented system. Nature (London) 400: 563-566.
Davis, F. W., D. M. Stoms, A. D. Hollander, K. A. Thomas, P. A. Stine, D. Odion, M. I.
Borchert, J. H. Thorne, M. V. Gray, R. E. Walker, K. Warner, and J. Graae. 1998.
The California gap analysis project: Final report. University of California, Santa
Barbara, CA.
Dobson, A. P., A. D. Bradshaw, and A. J. M. Baker. 1997. Hopes for the future:
Restoration ecology and conservation biology. Science 277: 515-522.
Landis, J. 2000. CUF, CUF II, and CURBA: A family of spatially explicit urban growth
and land-use policy simulation models. Pages 157-200 in R. K. Brail and R. E.
Klosterman, eds. Planning Support Systems. ESRI Press, Redlands, CA.
Landis, J., and M. Zhang. 1998. The second generation of the California urban futures
model. Part 1: Model logic and theory. Environment and Planning B-Planning &
Design 25: 657-666.
Makse, H. A., J. S. Andrade, M. Batty, S. Havlin, and H. E. Stanley. 1998. Modeling
urban growth patterns with correlated percolation. Physical Review E 58: 70547062.
Makse, H. A., S. Havlin, and H. E. Stanley. 1995. Modelling urban growth patterns.
Nature 377: 608-612.
Noss, R. F. 1990. Indicators for monitoring biodiversity - a hierarchical approach.
Conservation Biology 4: 355-364.
Press, D., D. F. Doak, and P. Steinberg. 1996. The role of local government in the
conservation of rare species. Conservation Biology 10: 1538-1548.
Soulé, M. E., and L. S. Mills. 1998. No need to isolate genetics. Science 282: 1658-1659.
Tilman, D., and C. Lehman. 2001. Human-caused environmental change: Impacts on
plant diversity and evolution. Proceedings of the National Academy of Sciences of
the United States of America 98: 5433-5440.
Vitousek, P. M., H. A. Mooney, J. Lubchenco, and J. M. Melillo. 1997. Human
domination of earth's ecosystems. Science 277: 494-499.
White, R., G. Engelen, and I. Uljee. 1997. The use of constrained cellular automata for
high-resolution modelling of urban land-use dynamics. Environment and
Planning B-Planning & Design 24: 323-343.
Wilcove, D. S., D. Rothstein, J. Dubow, A. Phillips, and E. Losos. 1998. Quantifying
threats to imperiled species in the United States. Bioscience 48: 607-615.