13
Elements of
Spatial Data Analysis
in Ecological Assessments
Patrick S. Bourgeron, Marie-Jose Fortin, and Hope C. Humphries
13.1 Introduction
Virtually any aspect of an ecological assessment
(EA) is likely to involve the topic of space, including its striking effect on landscapes and the distribution of human populations. For example, a spatially explicit approach is needed to address two
policy questions common to many EAs. What is
required for maintaining the long-term productivity of ecosystems? What is the impact of maintaining current management scenarios on, for example, major social issues or the maintenance of
rural communities and their economies in a given
area? Essential tasks of EAs also involve the explicit consideration of space, such as in combining
information from various geographic areas and
multiple scales. Most measurements of large-scale
phenomena, such as the effect of regional carbon
?Dd nitrogen cycles, hydrologic regimes, changes
In land-use patterns, and demographics, among
many others, carry the imprint of spatial variability and scaling. Therefore, explicit consideration of
all aspects of space (e.g., spatial variability and its
corollary, spatial scaling) is increasingly a central
concern in the design and implementation of EAs,
whether in map creation or incorporation in predictive modeling (see Chapters 3 and 18; also,
Haining, 1990; Ritchie, 1997).
EAs focus on environmental and social systems
in which spatial heterogeneity is functional and not
Patrick S. Bourgeron and Hope C. Humphries wish to
acknowledge partial funding provided by a Science to
Achieve Results grant from the U.S. Environmental Protection Agency ("Multi-scaled Assessment Methods:
Prototype Development within the Interior Columbia
Basin").
the result of random, noise-generating processes
(see Chapter 2; also, Kolasa and Rollo, 1991;
O'Neill et al., 1991; Legendre and Legendre,
1998). EA data are therefore inherently spatially
and temporally structured, likely at more than one
scale (see Chapter 2; also, Davis et al., 1991;
Schneider, 1994; Bradshaw, 1998; Pahl-Wostl,
1998). The characterization of physical, ecological,
and social systems depends on the particular spatial, temporal, and organizational perspectives resulting from addressing specific objectives (see
Chapter 3). Therefore, it is essential to understand
how patterns and processes vary in space and across
geographic scales (King, 1997; Hammer, 1998; Legendre and Legendre, 1998; Pahl-Wostl, 1998).
Spatial variability and scaling interact with each
other to produce the spatial arrangement of patterns
and processes (see Chapters 2 and 3). As a result,
three questions must be addressed during the design phase of an EA (Haining, 1990). (1) What expressions of spatial variability in the area of interest should be examined? (2) What components of
the assessment must be spatially explicit? (3) At
what scale do processes act on phenomena of interest?
The main purpose of this chapter is to address
the first two questions concerning quantification of
spatial variability. The third question is treated in
Chapters 13 and 14, in which several methods for
examining scaling properties in spatial data analysis (fractal geometry, percolation theory, and fuzzy
statistical and modeling approaches) are presented.
Here we discuss and review concepts, issues, and
approaches to spatial analysis and thereby provide
a basis for developing a sound inductive approach
to data analysis in a spatial context during the different stages of an EA. This chapter is organized
187
Elements of
Spatial Data Analysis
in Ecological Assessments
Patrick S. Bourgeron, Marie-Jose Fortin, and Hope C. Humphries
13.1 Introduction
Virtually any aspect of an ecological assessment
(EA) is likely to involve the topic of space, including its striking effect on landscapes and the distribution of human populations. For example, a spatially explicit approach is needed to address two
policy questions common to many EAs. What is
required for maintaining the long-term productivity of ecosystems? What is the impact of maintaining current management scenarios on, for example, major social issues or the maintenance of
rural communities and their economies in a given
area? Essential tasks of EAs also involve the explicit consideration of space, such as in combining
information from various geographic areas and
multiple scales. Most measurements of large-scale
phenomena, such as the effect of regional carbon
?Dd nitrogen cycles, hydrologic regimes, changes
In land-use patterns, and demographics, among
many others, carry the imprint of spatial variability and scaling. Therefore, explicit consideration of
all aspects of space (e.g., spatial variability and its
corollary, spatial scaling) is increasingly a central
concern in the design and implementation of EAs,
whether in map creation or incorporation in predictive modeling (see Chapters 3 and 18; also,
Haining, 1990; Ritchie, 1997).
EAs focus on environmental and social systems
in which spatial heterogeneity is functional and not
Patrick S. Bourgeron and Hope C. Humphries wish to
acknowledge partial funding provided by a Science to
Achieve Results grant from the U.S. Environmental Protection Agency ("Multi-scaled Assessment Methods:
Prototype Development within the Interior Columbia
Basin").
the result of random, noise-generating processes
(see Chapter 2; also, Kolasa and Rollo, 1991;
O'Neill et al., 1991; Legendre and Legendre,
1998). EA data are therefore inherently spatially
and temporally structured, likely at more than one
scale (see Chapter 2; also, Davis et al., 1991;
Schneider, 1994; Bradshaw, 1998; Pahl-Wostl,
1998). The characterization of physical, ecological,
and social systems depends on the particular spatial, temporal, and organizational perspectives resulting from addressing specific objectives (see
Chapter 3). Therefore, it is essential to understand
how patterns and processes vary in space and across
geographic scales (King, 1997; Hammer, 1998; Legendre and Legendre, 1998; Pahl-Wostl, 1998).
Spatial variability and scaling interact with each
other to produce the spatial arrangement of patterns
and processes (see Chapters 2 and 3). As a result,
three questions must be addressed during the design phase of an EA (Haining, 1990). (1) What expressions of spatial variability in the area of interest should be examined? (2) What components of
the assessment must be spatially explicit? (3) At
what scale do processes act on phenomena of interest?
The main purpose of this chapter is to address
the first two questions concerning quantification of
spatial variability. The third question is treated in
Chapters 13 and 14, in which several methods for
examining scaling properties in spatial data analysis (fractal geometry, percolation theory, and fuzzy
statistical and modeling approaches) are presented.
Here we discuss and review concepts, issues, and
approaches to spatial analysis and thereby provide
a basis for developing a sound inductive approach
to data analysis in a spatial context during the different stages of an EA. This chapter is organized
187