to make educated decisions regarding the appropriate use of data. As
Arnold et al. (2000) point out, “Researchers and educators must sometimes
walk a fine line between explaining such limitations, and undermining the
perceived value of the information by detailing a long list of technical
caveats.”
In addition to dangers associated with the wide availability of digital data,
new dangers have arisen with the advent of sophisticated software packages, including desktop GIS, models, and statistical packages. These packages allow users who may be unfamiliar with the underlying algorithms to
execute complicated analyses with the click of a mouse, often leading to
inappropriate results, which could lead to poor decision making. To fully
understand the results, users must not only be aware of the limitations of
the data, but must also determine what is being done by the software and
the underlying assumptions.
10.3 Data Collection/Acquisition Concerns
The primary constraints on gathering appropriate, high-quality data are the
time and resources available for data collection and acquisition efforts. In
many cases, ecological models are constructed to guide management and
policy decisions that must be made quickly. However, the ecological data
required for the construction of models needs to be collected over a longer
time scale. Available resources, such as money, equipment, and experienced
personnel, can be a deciding factor as to whether data are collected by the
modelers, contracted to another organization for collection, or acquired
from an existing source. Often, to meet time constraints, data are obtained
from existing sources, incurring a serious tradeoff among data availability,
appropriateness, and quality.
The time, resources, and existing data available for a modeling project
can influence decisions about the structure and function of the ecological
model being constructed. All models are generalizations of the target ecological system. It is hoped that they retain the key factors that drive the
system dynamics. However, the level of generalization can be constrained
by the time, resources, and existing data for the modeling effort. In cases
where management decisions must be made quickly, resources for data collection are limited, or the existing data are at a coarse temporal or spatial
scale, the ecological model that is constructed must be very simple. Simple
models that greatly generalize the system are useful as learning tools to
explore the ecological system and potential management actions, but should
not be relied upon for strict quantitative predictions of ecological performance. Simple models can also help focus future data collection efforts.
When additional time and resources are available for data collection or the
existing data closely match the desired extent and resolution, increasingly
rigorous models can be built to provide quantitative predictions. Remem10. Effective Ecological Modeling: Data Issues
187
Arnold et al. (2000) point out, “Researchers and educators must sometimes
walk a fine line between explaining such limitations, and undermining the
perceived value of the information by detailing a long list of technical
caveats.”
In addition to dangers associated with the wide availability of digital data,
new dangers have arisen with the advent of sophisticated software packages, including desktop GIS, models, and statistical packages. These packages allow users who may be unfamiliar with the underlying algorithms to
execute complicated analyses with the click of a mouse, often leading to
inappropriate results, which could lead to poor decision making. To fully
understand the results, users must not only be aware of the limitations of
the data, but must also determine what is being done by the software and
the underlying assumptions.
10.3 Data Collection/Acquisition Concerns
The primary constraints on gathering appropriate, high-quality data are the
time and resources available for data collection and acquisition efforts. In
many cases, ecological models are constructed to guide management and
policy decisions that must be made quickly. However, the ecological data
required for the construction of models needs to be collected over a longer
time scale. Available resources, such as money, equipment, and experienced
personnel, can be a deciding factor as to whether data are collected by the
modelers, contracted to another organization for collection, or acquired
from an existing source. Often, to meet time constraints, data are obtained
from existing sources, incurring a serious tradeoff among data availability,
appropriateness, and quality.
The time, resources, and existing data available for a modeling project
can influence decisions about the structure and function of the ecological
model being constructed. All models are generalizations of the target ecological system. It is hoped that they retain the key factors that drive the
system dynamics. However, the level of generalization can be constrained
by the time, resources, and existing data for the modeling effort. In cases
where management decisions must be made quickly, resources for data collection are limited, or the existing data are at a coarse temporal or spatial
scale, the ecological model that is constructed must be very simple. Simple
models that greatly generalize the system are useful as learning tools to
explore the ecological system and potential management actions, but should
not be relied upon for strict quantitative predictions of ecological performance. Simple models can also help focus future data collection efforts.
When additional time and resources are available for data collection or the
existing data closely match the desired extent and resolution, increasingly
rigorous models can be built to provide quantitative predictions. Remem10. Effective Ecological Modeling: Data Issues
187