General Introduction - Integrated Assessment of Health and Sustainability of Agroecosystems - page 12

Indicators must have a defined range in the healthy ecosystem (Schaeffer
et al., 1988). In some cases, these ranges—described as thresholds, standards, and
targets—depend on subjective value judgment (Gallopin, 1994a). The health status
of an ecosystem is indicated not only by the comparison of indicators to such ranges
but also by the pattern exhibited by the indicators over time and space and in relation
to each other. Relationships between indicators are key to understanding ecosystem
function and in relating cause to effect.
While the human and animal health diagnosis involves a limited set of param-
eters, ecosystems present a list that not only varies with level in the ecological hier-
archy, but also consists of virtually an infinite number of measurable parameters
(Schaeffer et al., 1988). Another difficulty has been that researchers from differ-
ent disciplines, conditioned by particular perceptions of this complex situation and
bound by habit to particular scales of research, arrive at different kinds of lists of
parameters to be measured (Waltner-Toews, 1996). How can indicators be selected
to ensure that the suite is at the same time parsimonious, covers all domains and all
important attributes of the system, provides managerially useful information, and is
cost-effective and timely to measure?
Since the aim of agroecosystem health research is the management of agroeco-
systems, it follows that the most useful suites of indicators are those that aid manag-
ers in their decision-making processes. Indicators must therefore be related to the
goals and objectives of the agroecosystems as well as the capacity and potential of
the system and the perceived risks and potential stresses. Although the goals and
objectives may be based—to a large extent—on value judgment and can be highly
subjective, measures of health (no matter what health is conceived to be) must be
technical, multidisciplinary, and objective (Waltner-Toews and Wall, 1997). Ideally,
indicators should be features of the agroecosystem that change with alteration in the
health status of the ecosystem. Because of this, indicators have been referred to as
the vital signs of the ecosystem (Rapport et al., 1985). Indicators may also be features
of an ecosystem that indicate the presence, absence, or magnitude of stress or risk.
These can be termed risk factors . A third category of indicators measures the poten-
tial, capabilities, or reserves of the ecosystem and can be termed health promoters .
1.4.5 s e l e C t i of n of f i n D i C A t of r s
Lightfoot and Noble (1993) and Izac and Swift (1994) suggested different suites
of indicators that can be used in the context of smallholder agroecosystems, while
Thompson and Pretty (1996) used a number of indicators to assess the impact of
a soil conservation program. Izac and Swift (1994) focused on the products, by-
products, and amenities of the agroecosystem at various levels of the agricultural
hierarchy. Although their list includes some elementary measures—such as soil pH,
soil exchangeable aluminum content, and stream turbidity and acidity—that can be
objectively assessed, most (such as nutritional status of households and communi-
ties, ratio of aggrading to degrading land area, and biodiversity and complexity) are
compound attributes that can only be assessed using other proxy variables.
Thompson and Pretty (1996) included sustained increases in productivity,
decreases in resource degradation, increases in local resilience, and decreases in