7.2 Spatial Structures and Spatial Analysis: A Short
Overview
7.2.1 Introduction
As mentioned in Chap. 6, spatial structures play a very important role in the analysis
of ecological data. Living communities are spatially structured at many scales, and
these structures are the result of several classes of processes. On the other hand, beta
diversity is the spatial variation in community composition; so, a study of the factors
that can explain the spatial variation of community composition is in every respect
an analysis of beta diversity (see Chap. 8). The environmental control model
advocates that external forces (climatic, physical, chemical) control living communities. If these factors are spatially structured, their patterns will be reflected on the
living communities (examples: patches of desert where the soil is humid enough to
support vegetation; gradient of successive communities through an intertidal zone).
The biotic control model predicts that intra- and interspecific interactions within
communities (examples: social groups of animals; top-down or bottom-up processes), as well as neutral processes such as ecological drift and limited dispersal,
may result in spatial patterns that are the cause of spatial autocorrelation in the strict
sense. Finally, historical events (e.g. past disturbances like fire or human settlements) may have structured the environment in a way that still influences presentday communities.
In all, ecological data reflect a combination of many structures, spatial or not:
• The overall mean of each response variable (species).
• If the whole sampling area is under the influence of an all-encompassing process
that changes the mean in a gradient across the area, then a trend is present. The
trend may be due to a process operating at a scale larger than the study area.
• Spatial structures at regional scales: ecological processes of various kinds (biotic
or abiotic) and neutral processes influence the data at scales finer than the overall
sampling area, producing identifiable spatial patterns.
• Local deterministic structures with no recognizable spatial component because
the sampling design is not fine enough to identify such fine-scale patches.
• Random noise (error): this is the residual (stochastic) component of the variation.
It can be attributed to local effects operating independently at each sampling site
and to sampling variation.
One of the aims of spatial analysis is to discriminate between these sources of
variation and model the relevant ones separately.
300
7 Spatial Analysis of Ecological Data
Overview
7.2.1 Introduction
As mentioned in Chap. 6, spatial structures play a very important role in the analysis
of ecological data. Living communities are spatially structured at many scales, and
these structures are the result of several classes of processes. On the other hand, beta
diversity is the spatial variation in community composition; so, a study of the factors
that can explain the spatial variation of community composition is in every respect
an analysis of beta diversity (see Chap. 8). The environmental control model
advocates that external forces (climatic, physical, chemical) control living communities. If these factors are spatially structured, their patterns will be reflected on the
living communities (examples: patches of desert where the soil is humid enough to
support vegetation; gradient of successive communities through an intertidal zone).
The biotic control model predicts that intra- and interspecific interactions within
communities (examples: social groups of animals; top-down or bottom-up processes), as well as neutral processes such as ecological drift and limited dispersal,
may result in spatial patterns that are the cause of spatial autocorrelation in the strict
sense. Finally, historical events (e.g. past disturbances like fire or human settlements) may have structured the environment in a way that still influences presentday communities.
In all, ecological data reflect a combination of many structures, spatial or not:
• The overall mean of each response variable (species).
• If the whole sampling area is under the influence of an all-encompassing process
that changes the mean in a gradient across the area, then a trend is present. The
trend may be due to a process operating at a scale larger than the study area.
• Spatial structures at regional scales: ecological processes of various kinds (biotic
or abiotic) and neutral processes influence the data at scales finer than the overall
sampling area, producing identifiable spatial patterns.
• Local deterministic structures with no recognizable spatial component because
the sampling design is not fine enough to identify such fine-scale patches.
• Random noise (error): this is the residual (stochastic) component of the variation.
It can be attributed to local effects operating independently at each sampling site
and to sampling variation.
One of the aims of spatial analysis is to discriminate between these sources of
variation and model the relevant ones separately.
300
7 Spatial Analysis of Ecological Data