7.3 Use of Existing Data
Existing data layers, such as maps of vegetation,
soils, geology, climate, and terrain, can be used as
initial surrogates for field data. Problems can arise
if data layers have different scales, adding to the
initial uncertainty and error rate inherent in each
map. For example, inconsistencies in existing subregional vegetation maps acquired from multiple
sources rendered them unsuitable for use in the
ICBEMP (Hann et aI., 1997). However, this approach has been used successfully by various investigators. Maps of land systems in Australia
(Christian and Stewart, 1953) were incorporated
into a database to describe the range of natural environments in the study region. The database was
later used to construct a stratified sampling scheme
to describe the range of natural variability contained in various areas in the region (Pressey and
Nicholls, 1991).
7.3.2 Remotely Sensed Data
Remotely sensed imagery (see Chapter 10) is readily available over most of North America (e.g., the
National Satellite Land Remote Sensing Data
Archive at the USGS EROS Data Center, Holm
1997, Larsen et aI., 1998). Use of remotely sensed
data is a cost-effective approach for broad-scale
mapping of land cover (e.g., the National Land
Cover Characterization program at the USGS
EROS Data Center, Van Driel and Loveland, 1996)
and vegetation (e.g., the USGS GAP Analysis Program, or GAP, Scott et aI., 1990, 1993; Scott and
Jennings, 1997, 1998). In addition, remotely sensed
imagery can provide a host of broad-scale measurements on the structure and function of regional
ecosystems (Wessman and Asner, 1998). However,
there are significant challenges in using information and producing maps from remotely sensed data
(Estes and Mooneyhan, 1994) that pertain to the
scale of measurement of the data (Davis et aI.,
1991; Wilkinson, 1996; Bian, 1997; Wessman and
Asner, 1998).
There has been a trend toward acquiring ground
and remotely sensed data over a range of scales
(e.g., Running et aI., 1989; Davis et al., 1991; Cao
and Lam, 1997). Davis et aI. (1991) concluded that
some of the problems in the use of remotely sensed
data relate to understanding the spatial scale dependence of patterns and processes. The search for
multiscaled patterns and processes (Levin, 1992;
Milne, 1994; Minshall, 1994; Turner et al., 1994)
in tum requires the collection of ground data at the
appropriate intensity and frequency. The spatial
and temporal scales of remotely sensed data may
differ from those of interest in particular cases (Tre95
week, 1999). Special attention is needed to understand the scaling relations of the phenomena of interest (e.g., see Mack et aI., 1997, for deriving
species-area relationships and Mladenoff et aI.,
1997, for discriminating among ecoregions, also
see the wide-ranging series of approaches on the
subject in Quattrochi and Goodchild, 1997) and to
integrate remotely sensed and field data (e.g., Lobo
et aI., 1998; Wessman and Asner, 1998; Treweek,
1999).
7.3.3 Spatially Referenced
Species Records
Another source of existing data is spatially referenced species records, such as herbarium and museum records. Databases that compile this kind of
information are maintained over most of North
America by state- and province-based Natural Heritage Programs or Conservation Data Centers and
The Nature Conservancy (Jenkins 1985) using Biological Conservation Database (BCD) software.
Herbarium and museum records for specific plant
and animal species, plant communities, and/or
ecosystems (including information on identification, location, ownership, and any relevant biological and management information) are collected
and entered into the BCD for each state or province.
Although the information in the database is descriptive, cost-effective uses can be made of the
data for many purposes. The GAP programs (Scott
et aI., 1990, 1993; Scott and Jennings, 1997, 1998)
combine such distribution information for selected
animal species with vegetation maps to determine
species-habitat relations. This exploratory analysis
defines geographic areas that have suitable habitats
for the species.
As part of SNEP, records of rare and endemic
plant species were collected from a variety of
sources, such as the California Native Plant Society's Inventory of Rare and Endangered Vascular
Plants of California (Skinner and Pavlik, 1994),
rare plant data maintained by the California Natural Diversity Database (California Department of
Fish and Game, 1995), and rare plant information
from the Nevada Heritage Program (Morefield and
Knight, 1991; Morefield, 1994). Plant species distributions in the Sierra Nevada were attributed to
coarse-scale river basins, counties, and topographic
quadrangles for analyses (Shevock, 1996). The
analysis was constrained by the fact that distribution information for many taxa was found to be incomplete due to limited field studies and collection
of vouchered specimens.
Existing data layers, such as maps of vegetation,
soils, geology, climate, and terrain, can be used as
initial surrogates for field data. Problems can arise
if data layers have different scales, adding to the
initial uncertainty and error rate inherent in each
map. For example, inconsistencies in existing subregional vegetation maps acquired from multiple
sources rendered them unsuitable for use in the
ICBEMP (Hann et aI., 1997). However, this approach has been used successfully by various investigators. Maps of land systems in Australia
(Christian and Stewart, 1953) were incorporated
into a database to describe the range of natural environments in the study region. The database was
later used to construct a stratified sampling scheme
to describe the range of natural variability contained in various areas in the region (Pressey and
Nicholls, 1991).
7.3.2 Remotely Sensed Data
Remotely sensed imagery (see Chapter 10) is readily available over most of North America (e.g., the
National Satellite Land Remote Sensing Data
Archive at the USGS EROS Data Center, Holm
1997, Larsen et aI., 1998). Use of remotely sensed
data is a cost-effective approach for broad-scale
mapping of land cover (e.g., the National Land
Cover Characterization program at the USGS
EROS Data Center, Van Driel and Loveland, 1996)
and vegetation (e.g., the USGS GAP Analysis Program, or GAP, Scott et aI., 1990, 1993; Scott and
Jennings, 1997, 1998). In addition, remotely sensed
imagery can provide a host of broad-scale measurements on the structure and function of regional
ecosystems (Wessman and Asner, 1998). However,
there are significant challenges in using information and producing maps from remotely sensed data
(Estes and Mooneyhan, 1994) that pertain to the
scale of measurement of the data (Davis et aI.,
1991; Wilkinson, 1996; Bian, 1997; Wessman and
Asner, 1998).
There has been a trend toward acquiring ground
and remotely sensed data over a range of scales
(e.g., Running et aI., 1989; Davis et al., 1991; Cao
and Lam, 1997). Davis et aI. (1991) concluded that
some of the problems in the use of remotely sensed
data relate to understanding the spatial scale dependence of patterns and processes. The search for
multiscaled patterns and processes (Levin, 1992;
Milne, 1994; Minshall, 1994; Turner et al., 1994)
in tum requires the collection of ground data at the
appropriate intensity and frequency. The spatial
and temporal scales of remotely sensed data may
differ from those of interest in particular cases (Tre95
week, 1999). Special attention is needed to understand the scaling relations of the phenomena of interest (e.g., see Mack et aI., 1997, for deriving
species-area relationships and Mladenoff et aI.,
1997, for discriminating among ecoregions, also
see the wide-ranging series of approaches on the
subject in Quattrochi and Goodchild, 1997) and to
integrate remotely sensed and field data (e.g., Lobo
et aI., 1998; Wessman and Asner, 1998; Treweek,
1999).
7.3.3 Spatially Referenced
Species Records
Another source of existing data is spatially referenced species records, such as herbarium and museum records. Databases that compile this kind of
information are maintained over most of North
America by state- and province-based Natural Heritage Programs or Conservation Data Centers and
The Nature Conservancy (Jenkins 1985) using Biological Conservation Database (BCD) software.
Herbarium and museum records for specific plant
and animal species, plant communities, and/or
ecosystems (including information on identification, location, ownership, and any relevant biological and management information) are collected
and entered into the BCD for each state or province.
Although the information in the database is descriptive, cost-effective uses can be made of the
data for many purposes. The GAP programs (Scott
et aI., 1990, 1993; Scott and Jennings, 1997, 1998)
combine such distribution information for selected
animal species with vegetation maps to determine
species-habitat relations. This exploratory analysis
defines geographic areas that have suitable habitats
for the species.
As part of SNEP, records of rare and endemic
plant species were collected from a variety of
sources, such as the California Native Plant Society's Inventory of Rare and Endangered Vascular
Plants of California (Skinner and Pavlik, 1994),
rare plant data maintained by the California Natural Diversity Database (California Department of
Fish and Game, 1995), and rare plant information
from the Nevada Heritage Program (Morefield and
Knight, 1991; Morefield, 1994). Plant species distributions in the Sierra Nevada were attributed to
coarse-scale river basins, counties, and topographic
quadrangles for analyses (Shevock, 1996). The
analysis was constrained by the fact that distribution information for many taxa was found to be incomplete due to limited field studies and collection
of vouchered specimens.