322
Elements of Ecological Land Classifications for Ecological Assessments
tributes. However, ELCs are always used to generate maps of ecological land units (ELUs),
whereas many other types of classification (e.g.,
taxonomic) do not include map generation. The basic characteristic that all ELCs share is their use in
the environmental stratification of geographic
space (Bunce et aI., 1996a, 1996b; also see Chapter 3): environmental space is stratified by ELC
classes derived from the variables used in their construction. The purpose of environmental stratification is to (1) quantify and minimize variation in
physical, chemical, biological, and ecological patterns and processes, or any other measure of interest (Hughes et aI., 1990; Bunce et al., 1996a) and
(2) represent the distributions of these patterns and
processes. Furthermore, it is hypothesized that this
reduction in variation allows better quantification
of the responses of patterns and processes to global
changes and land-use practices. ELCs are the product of a formal process in which the results of local ecosystem characterization (KUchler, 1973; also
see Chapter 3) are regionalized within a spatial
framework. This chapter does not discuss mapping
and stratification per se or the relationship between
classification and mapping. Hereafter, we refer to
an ELC class defined at any scale as an ELU. Reviews of the topics of ecosystem characterization,
ecological classification and mapping, and the various terms employed can be found in KUchler and
Zonneveld (1988), Sims et aI. (1996), and Grossman et al. (1999); also see Chapter 3.
Although they may differ in nature and scope,
ELCs share the goal of representing some or all
ecosystem components and their interactions. To
achieve this goal, some ELCs emphasize abiotic
(physical) ecosystem components, whereas others
emphasize biotic components or attempt to represent both. Figure 22.1 a represents a simplified general model of the ecosystem patterns and processes
and their hypothesized relationships that ELCs are
intended to represent. The model has four major
compartments: regional climate (A), the physical
environment (B), disturbances and land use (C),
and existing biotic characteristics (D). In this simplified ecosystem model, potential vegetation (PV)
(compartment B) is defined as the vegetation that
an area can potentially support in the absence of
disturbances and human activities (TUxen, 1959;
Pfister and Arno, 1980). Therefore, according to
this definition, PV is an expression of the physical
attributes of the environment and is rarely actually
observed. As defined, PV is useful for the characterization of ELUs (KUchler, 1964; Jensen et aI.,
1997; Zerbe, 1998); in some ELCs, ELUs are defined solely on the basis of PV (e.g., Daubenmire,
1966; Pfister and Arno, 1980). In this chapter, we
refer to PV as an attribute of the physical environment. The attributes of compartment C are rarely
incorporated in the definition of ELUs; they are often analyzed using ELUs as templates for deriving
other ecosystem properties (e.g., the historic range
of variability, HR V; see Section 22.3 and Chapter
19). Attributes of compartments A, B, and D have
been used alone or in combination to defme ELUs.
Only some of the factors identified in the model in
Figure 22.1 may be relevant to the definition of an
ELC, based on specific goals and spatial scales.
Three main approaches to ELCs have been followed. The first approach employs the recognition
of biotic communities (e.g., assemblages of species
or species attributes, such as physiognomy) as a
surrogate for the environment (Figure 22.1 b). The
most widely used method delineates vegetation
types, based on the assumption that vegetation is a
faithful expression of site characteristics (Kiichler,
1988). This method has a venerable history (McIntosh, 1985). A recent example of this approach in
the United States is the national Gap Analysis Project (Scott et aI., 1993). However, when testing the
value of an ecological relationship, arguments
about assessing the correlations between biotic
units and environmental factors can become circular, with the focus being switched between biotic
and abiotic factors (Mackey et aI., 1988; Short and
Hestbeck, 1995).
The second approach utilizes broad environmental patterns alone (Figure 22.1c) or broad correlations between the biota and indirect environmental variables (Figure 22.1d) to describe and
delineate ELUs. Indirect variables (Figure 22.1c)
are defined as those factors that do not have a direct influence on the ecosystem components of interest (definition adapted from Austin et aI., 1984;
Austin, 1985; Austin and Smith, 1989). For example, elevation is an indirect variable in determining
the distribution of the vegetation (Austin et al.,
1984). The definition of an indirect variable is specific to the pattern of interest. A variety of systems
based on combinations of soils, lithology, and landform have been used alone or combined with biotic data to produce classifications of ecological regions (ecoregions) or natural landscape units
-
FIGURE 22.1. Simplified general model of relationships
among ecosystem components used in ecological land
classifications. (a) all patterns and processes. See text for
explanation of compartments A, B, C, and D; (b) existing biotic characteristics alone; (c) indirect variables
alone (see text for definition); (d) existing biotic characteristics and indirect variables; (e) direct variables alone
(see text for definition).
Elements of Ecological Land Classifications for Ecological Assessments
tributes. However, ELCs are always used to generate maps of ecological land units (ELUs),
whereas many other types of classification (e.g.,
taxonomic) do not include map generation. The basic characteristic that all ELCs share is their use in
the environmental stratification of geographic
space (Bunce et aI., 1996a, 1996b; also see Chapter 3): environmental space is stratified by ELC
classes derived from the variables used in their construction. The purpose of environmental stratification is to (1) quantify and minimize variation in
physical, chemical, biological, and ecological patterns and processes, or any other measure of interest (Hughes et aI., 1990; Bunce et al., 1996a) and
(2) represent the distributions of these patterns and
processes. Furthermore, it is hypothesized that this
reduction in variation allows better quantification
of the responses of patterns and processes to global
changes and land-use practices. ELCs are the product of a formal process in which the results of local ecosystem characterization (KUchler, 1973; also
see Chapter 3) are regionalized within a spatial
framework. This chapter does not discuss mapping
and stratification per se or the relationship between
classification and mapping. Hereafter, we refer to
an ELC class defined at any scale as an ELU. Reviews of the topics of ecosystem characterization,
ecological classification and mapping, and the various terms employed can be found in KUchler and
Zonneveld (1988), Sims et aI. (1996), and Grossman et al. (1999); also see Chapter 3.
Although they may differ in nature and scope,
ELCs share the goal of representing some or all
ecosystem components and their interactions. To
achieve this goal, some ELCs emphasize abiotic
(physical) ecosystem components, whereas others
emphasize biotic components or attempt to represent both. Figure 22.1 a represents a simplified general model of the ecosystem patterns and processes
and their hypothesized relationships that ELCs are
intended to represent. The model has four major
compartments: regional climate (A), the physical
environment (B), disturbances and land use (C),
and existing biotic characteristics (D). In this simplified ecosystem model, potential vegetation (PV)
(compartment B) is defined as the vegetation that
an area can potentially support in the absence of
disturbances and human activities (TUxen, 1959;
Pfister and Arno, 1980). Therefore, according to
this definition, PV is an expression of the physical
attributes of the environment and is rarely actually
observed. As defined, PV is useful for the characterization of ELUs (KUchler, 1964; Jensen et aI.,
1997; Zerbe, 1998); in some ELCs, ELUs are defined solely on the basis of PV (e.g., Daubenmire,
1966; Pfister and Arno, 1980). In this chapter, we
refer to PV as an attribute of the physical environment. The attributes of compartment C are rarely
incorporated in the definition of ELUs; they are often analyzed using ELUs as templates for deriving
other ecosystem properties (e.g., the historic range
of variability, HR V; see Section 22.3 and Chapter
19). Attributes of compartments A, B, and D have
been used alone or in combination to defme ELUs.
Only some of the factors identified in the model in
Figure 22.1 may be relevant to the definition of an
ELC, based on specific goals and spatial scales.
Three main approaches to ELCs have been followed. The first approach employs the recognition
of biotic communities (e.g., assemblages of species
or species attributes, such as physiognomy) as a
surrogate for the environment (Figure 22.1 b). The
most widely used method delineates vegetation
types, based on the assumption that vegetation is a
faithful expression of site characteristics (Kiichler,
1988). This method has a venerable history (McIntosh, 1985). A recent example of this approach in
the United States is the national Gap Analysis Project (Scott et aI., 1993). However, when testing the
value of an ecological relationship, arguments
about assessing the correlations between biotic
units and environmental factors can become circular, with the focus being switched between biotic
and abiotic factors (Mackey et aI., 1988; Short and
Hestbeck, 1995).
The second approach utilizes broad environmental patterns alone (Figure 22.1c) or broad correlations between the biota and indirect environmental variables (Figure 22.1d) to describe and
delineate ELUs. Indirect variables (Figure 22.1c)
are defined as those factors that do not have a direct influence on the ecosystem components of interest (definition adapted from Austin et aI., 1984;
Austin, 1985; Austin and Smith, 1989). For example, elevation is an indirect variable in determining
the distribution of the vegetation (Austin et al.,
1984). The definition of an indirect variable is specific to the pattern of interest. A variety of systems
based on combinations of soils, lithology, and landform have been used alone or combined with biotic data to produce classifications of ecological regions (ecoregions) or natural landscape units
-
FIGURE 22.1. Simplified general model of relationships
among ecosystem components used in ecological land
classifications. (a) all patterns and processes. See text for
explanation of compartments A, B, C, and D; (b) existing biotic characteristics alone; (c) indirect variables
alone (see text for definition); (d) existing biotic characteristics and indirect variables; (e) direct variables alone
(see text for definition).