Articles

Are We Conserving What We

Say We Are? Measuring
Ecological Integrity within
Protected Areas
JEFFREY D. PARRISH, DAVID P. BRAUN, AND ROBERT S. UNNASCH

Managers of protected areas are under increasing pressure to measure their effectiveness in conserving native biological diversity in ways that are
scientifically sound, practical, and comparable among protected areas over time. The Nature Conservancy and its partners have developed a
“Measures of Success” framework with four core components: (1) identifying a limited number of focal conservation targets, (2) identifying key
ecological attributes for these targets, (3) identifying an acceptable range of variation for each attribute as measured by properly selected indicators,
and (4) rating target status based on whether or not the target’s key attributes are within their acceptable ranges of variation. A target cannot be
considered “conserved” if any of its key ecological attributes exceeds its acceptable range of variation. The framework provides a rigorous basis not
only for measuring success but for setting conservation objectives, assessing threats to biodiversity, identifying monitoring and research needs, and
communicating management information to nonspecialists.

Keywords: monitoring, ecological integrity, protected area effectiveness, measures of success

A re we conserving what we say we are? This

question is increasingly asked of and by protected area
managers worldwide. The answers, unfortunately, remain
a growing skepticism among policymakers and funding agen-
cies about the long-term value of these conservation efforts
(Senge 1994, Salafsky and Margoluis 1999a, Salafsky et al.
ambiguous at best. Conservationists and protected area man- 2001).
agers around the world spend millions of dollars each year to In response, several institutions have developed systems for
conserve biodiversity (Castro and Locker 2000, WRI 2000). measuring the efficiency and efficacy of protected area man-
Although efforts aimed at measuring the amount of conser- agement (e.g., Hockings 1998, TNC 1998, Courrau 1999,
vation activity are increasing, the ability to measure the con- Dudley et al. 1999, Hockings 2003). Most of these systems fit
servation impact of these investments and to document the within an overall framework promoted globally by the IUCN
true effectiveness of conservation actions has not greatly im- (World Conservation Union) World Commission on Pro-
proved (Hockings et al. 2000, Salafsky et al. 2002). Without tected Areas (Hockings et al. 2000, Hockings 2003). This
objective measurement, conservationists cannot claim suc- management effectiveness framework provides a system for
cesses, learn from failures, or work effectively and efficiently identifying the information protected area managers should
toward the conservation of the remaining biological diversity evaluate to determine whether management processes and
of the planet (Redford and Taber 2000, Salafsky et al. 2002). conservation impacts are progressing as desired. The frame-
For protected areas that focus on the conservation of bi-
ological diversity, the impact of conservation investment on
biodiversity status is being questioned by donors and policy- Jeffrey D. Parrish (e-mail: [email protected]) is the director of conservation plan-
makers alike. Yet few parks have established systems to eval- ning for the Global Priorities Group, The Nature Conservancy, 3368 West 37th
uate management effectiveness or to determine whether they Avenue, Denver, CO 80211. David P. Braun is a senior biohydrologist for the
are conserving the biodiversity they say they are (Brandon et Freshwater Initiative, The Nature Conservancy, 570 Seventh Avenue, Suite 601,
al. 1998, Hockings et al. 2000). This widespread inability to New York, NY 10018. Robert S. Unnasch is a senior ecologist for the Adaptive
measure progress, to learn through adaptive management, and Management Program, The Nature Conservancy, 1109 Main Street, Suite 303,
to hold organizations accountable for conservation has led to Boise, ID 83702. © 2003 American Institute of Biological Sciences.

September 2003 / Vol. 53 No. 9 • BioScience 851

Articles

work identifies six aspects of protected area management A framework for measuring
for evaluation: context, planning, inputs, process, outputs, and conservation effectiveness
outcomes. This article focuses on the last of these six, the mea- The native biological diversity of any area includes innu-
surement of conservation outcomes. merable species unknown or at best poorly known to science,
embedded in numerous ecological systems whose webs of bi-
Measuring what matters most in biodiversity parks otic and abiotic interactions are only poorly understood.
Though protected areas throughout the world have different Where a protected area addresses only one or a few iconic
purposes, we focus our attention on those areas with the species, the goals and criteria for success may be relatively easy
principal purpose of biodiversity conservation. We propose to define. Where the focus of the protected area is native
that to assess conservation impact, management in these biological diversity itself, however, the challenge to setting goals
areas must address two primary outcome measures, namely, and measuring success becomes more difficult, particularly
threat status and ecological integrity. if the measures are to be scientifically defensible, practical,
comparable from one protected area to another, and replicable
Threat status. Are the most critical threats that confront bio- over time.
logical resources at a park changing in their severity or geo- Tracking biological diversity in an area using species cen-
graphic scope as a result of conservation strategies (or lack sus data provides one potential avenue for measuring success;
thereof)? For example, has bushmeat poaching declined as a another lies in the use of indexes of biotic integrity that in-
result of efforts to develop and improve contained domestic corporate information on both taxonomic and functional
animal husbandry as a protein source for local communities? composition of sampled communities (e.g., Noss 1990,
Measurement of threat status has gained increasing attention O’Connell et al. 1998, Karr and Chu 1999, Sayre et al 2000,
among practitioners and students of conservation (e.g., Salaf- Stein et al. 2000). Such approaches face many challenges in pro-
sky and Margoluis 1999b, Hockings et al. 2001, Margoluis and tected areas, especially those that span large areas or incor-
Salafsky 2001, Ervin 2002). Clearly, without reduction in the porate combinations of terrestrial, freshwater, and coastal
threats to biodiversity, those species and ecosystems that are marine ecosystems. The costs of repeated, comprehensive
biological censuses can be unsustainable. In addition, biotic
the focus of conservation investments will rapidly degrade and
responses to threats may lag behind the pace of the threats or
disappear. Yet, however important, measuring threat status is
be difficult to detect with sparse monitoring data. Further, dif-
insufficient on its own for several reasons. Most significantly,
ferent biotic measures may be difficult to compare or stan-
a focus on threat status alone must assume that there is a clear,
dardize within the same protected area over time, let alone
often linear, relationship between a threat and the status of bio-
across multiple protected areas. Finally, different biotic mea-
diversity. This runs counter to recent evidence of the nonlinear
sures may be difficult to interpret for people who are not spe-
dynamics of ecosystems and threshold effects (e.g., Scheffer
cialists in the particular taxa involved, and many conservation
et al. 2001). Also, a singular focus on threats can lead to a zero-
managers are, in fact, nonspecialists (e.g., Salafsky and Mar-
tolerance approach to threat activities in human-influenced goluis 1999b, Dale and Beyeler 2001).
landscapes, and under most circumstances such an approach An alternative approach to measuring conservation success
is unrealistic. pursued by a growing number of organizations, including The
Nature Conservancy, involves the use of some form of eco-
Ecological integrity. Do the ecological systems, communities, logical scorecard. Such scorecards tabulate and synthesize
and species that are the focus of conservation efforts occur with diverse scientific information about the focal biodiversity of
sufficient size, with appropriately functioning ecological an area into a small number of measurement categories,
processes, and with sufficiently natural composition, struc- which are standardized for use across multiple areas and
ture, and function to persist over the long term? For exam- conservation projects. Examples include the frameworks de-
ple, is the riparian forest maintaining its natural range of veloped by The Nature Conservancy (TNC 2000a) and the
species and patch composition, and is it resilient despite an River Health Programme in South Africa (e.g., Angliss et al.
increase in major flood events? Adapting the definition from 2001) and the framework advocated by Harwell, Young, and
Karr and Dudley (1981), we define ecological integrity as the others (Harwell et al. 1999, Young and Sanzone 2002).
ability of an ecological system to support and maintain a The Nature Conservancy and its global conservation part-
community of organisms that has species composition, di- ners have devoted significant resources to developing a prac-
versity, and functional organization comparable to those of tical framework for assessing conservation impact (Poiani et
natural habitats within a region. An ecological system or al. 1998, TNC 2000a, 2000b, Salafsky et al. 2002). This frame-
species has integrity or is viable when its dominant ecologi- work, called “Measures of Conservation Success,” is being
cal characteristics (e.g., elements of composition, structure, implemented by The Nature Conservancy and its partners in
function, and ecological processes) occur within their natural hundreds of large-scale conservation areas across the Amer-
ranges of variation and can withstand and recover from most icas, Asia, the Pacific Islands, and Africa. The Nature Con-
perturbations imposed by natural environmental dynamics servancy has recently refined this framework to strengthen its
or human disruptions. scientific rigor and improve its use in adaptive management,

852 BioScience • September 2003 / Vol. 53 No. 9

 Articles

incorporating innovations and advice from several programs tion targets for an area when their health and population dy-
and partner organizations (TNC 2003). This refined frame- namics vary in response to the full range of critical environ-
work has been tested across dozens of field projects over the mental factors and biological processes of the ecosystem in
past 4 years, across a spectrum from data-rich to data-poor which they are embedded. Examples include species that
areas. The framework uses quantitative and qualitative data play critical trophic or landscape-shaping roles in an eco-
assembled by teams of partner institutions and experts to track system; require large ranges to sustain their populations or
important ecological characteristics and synthesize their accommodate migratory patterns; require a broad spectrum
status into a set of simple categorical ratings (e.g., poor, fair, of habitat conditions, from recently disturbed to long-
good, very good) of biodiversity status in an area. These rat- undisturbed successional stages; or are highly sensitive to
ings are scientifically credible and readily interpreted by pro- human interference. Yet the very qualities that make a species
tected area managers. Through repeated measurement, man- sensitive to changes in some features of an ecosystem often
agers can use the framework to determine whether the status make it insensitive to others and therefore less suitable as a con-
of biodiversity is responding to conservation investments servation target. For example, predators may eat either native
and strategies over time. The framework has the added ad- or exotic prey without harm, and riparian forests may thrive
vantages of providing a rigorous basis for setting conserva- even along rivers that have lost all their native fishes. Further,
tion objectives, assessing threats to biodiversity, identifying the dynamic environmental regimes and constraints that are
monitoring and research needs, and communicating man- critical for any one species may operate at different spatial and
agement information to nonspecialists. temporal scales than those that affect others (Holling 1992).
The proposed ecological scorecard for assessing ecosystem As a result, conserving the conditions that best suit only a few
integrity and species viability has four core components: native species will often not ensure the right conditions for
(1) selecting a limited suite of focal biodiversity targets, the the long-term survival of all native species and communities.
conservation of which is intended to serve as a coarse-filter/ We therefore recommend the selection of ecological com-
fine-filter framework for protecting the whole; (2) identify- munities or systems at the outset as “coarse-filter” conserva-
ing a limited suite of key ecological attributes for each target, tion targets (Noss and Cooperrider 1994, Poiani et al. 2000),
along with specific indicators for each, that provide the followed by the selection of species with unique ecological req-
information for measuring target status; (3) identifying an uisites not captured in the conservation of the communities
acceptable range of variation for each key ecological attribute or ecological systems in which they are embedded. The com-
of the focal conservation targets, defining the limits of vari- bined species, community, and ecological system targets—
ation within which the key ecological attribute must lie for the preferably a small and practical number—must create a safety
target to be considered conserved; and (4) assessing the net for the ecosystem as a whole, meaning that their conser-
current status of each target, based on the status of its key eco- vation will help ensure that suitable environmental conditions
logical attributes with respect to their acceptable ranges of vari- exist for the persistence of all native species within a conser-
ation, and integrating the assessments of target status into a vation landscape or protected area.
measure of the status of biodiversity overall. The Nature
Conservancy and other organizations have also developed Identifying the key ecological
scorecard frameworks for tracking institutional capacity, attributes for conservation targets
threats to conservation targets, and other inputs into protected To identify what is most important to manage for the con-
area management (e.g., Hockings et al. 2000, 2001, TNC servation of biodiversity in protected areas, we must first
2000a), but this article focuses on a scorecard specifically for synthesize our best understanding of the ecology of the con-
tracking the ecological outcomes of conservation activity. servation target, a process greatly aided by the development
of ecological models. An ecological model for a conservation
Identifying focal conservation targets target (a species, community, or ecological system) identifies
Biodiversity conservation targets (hereafter conservation tar- a limited number of biological characteristics, ecological
gets, sensu Noss 1996, TNC 2000b, Salafsky et al. 2002) are a processes, and interactions with the physical environment—
limited number of species, natural communities, or entire eco- along with the critical causal links among them—that dis-
logical systems that are chosen to represent the biodiversity tinguish the target from others, shape its natural variation over
of a conservation landscape or protected area. These targets time and space, and typify an exemplary reference occurrence
serve as the foci of conservation investment and measures of (Maddox et al. 2001). Some of these characteristics are espe-
conservation effectiveness. The reasoning behind such use of cially pivotal, influencing a host of other characteristics of the
reduced numbers of elements of biodiversity for conservation target and its long-term persistence. We label such defining
planning is richly addressed in the literature (e.g., Noss and characteristics of a target its key ecological attributes (see fig-
Cooperrider 1994, Christensen et al. 1996, Schwartz 1999, ure 1). For example, consider a riparian ecosystem located
Poiani et al. 2000, Carignan and Villard 2002, Sanderson et within the foothills of a montane ecoregion. It is possible to
al. 2002). identify enormous suites of species and describe numerous
What elements of biodiversity should be chosen as con- biotic and abiotic interactions that typify this system. Yet the
servation targets? Individual species work well as conserva- spring flooding regime would clearly qualify as a key ecological

September 2003 / Vol. 53 No. 9 • BioScience 853

Articles

Process

Key attribute
Adult
migration
upstream

Adult Spawning
survivorship and growth adult (upper floodplain
(ocean and delta) and river)

Migration passage Spawning

Smolt flows habitat size
outmigration

Female lays
Connectivity eggs in gravel
Physiological
transformation Spawning
for habitat structure
outmigration
Water chemistry
(pH, DO,
turbidity)

Eggs
Juvenile survivorship and (upper floodplain and
growth (floodplain) river)

Juvenile
migration
River–floodplain Rearing habitat downstream
connectivity size

Figure 1. Conceptual ecological model of the key ecological attributes for a species-level conservation
target, chinook salmon (Oncorhynchus tshawytscha), for the Cosumnes River Protected Area, Califor-
nia. (DO stands for dissolved oxygen.)

attribute of this ecosystem, because its magnitude, spatial (again including disturbance dynamics), and attributes of
extent, timing, and duration can play a pivotal role in a landscape structure and architecture that sustain the target’s
cascade of biological dynamics, such as seed dispersal for composition and its natural dynamics (Noss 1990, 1996,
native riparian vegetation, variation in soil composition and Noss et al. 1995, Christensen et al. 1996, Schwartz 1999,
fertility, elimination of invasive species that compete with na- Poiani et al. 2000, TNC 2000a,Young and Sanzone 2002). Iden-
tive species, and patterns of succession. tifying key attributes that address more than just biotic com-
The Nature Conservancy’s Measures of Success frame- position is important for two reasons. First, the abundance and
work rests on the premise that key ecological attributes must composition of a target may lag in their responses to envi-
be managed and conserved to sustain each conservation tar- ronmental impairments, and data on biotic interactions, en-
get. By explicitly identifying such attributes, protected area vironmental regimes, and landscape structure can help ensure
managers can specify what is important to manage and mon- the early detection of threats and changes resulting from hu-
itor about individual conservation targets and, through mon- man activities. Second, conserving the focal targets is not
itoring these targets, can assess conservation success. Con- the ultimate goal but a means for conserving all native bio-
servation targets and their key ecological attributes, therefore, diversity in an area. Consideration of these additional types
are the essential currency for conservation management at any of key ecological attributes will further ensure that crucial as-
scale. pects of ecological integrity are managed for the conservation
The key ecological attributes of any conservation target in- of all native biodiversity.
clude not only its biological composition (and crucial pat- Key attributes of a target’s biological composition and
terns of variation in this composition over space) but also the spatial variation will differ, depending in part on whether the
biotic interactions and processes (including disturbance and suc- target is an individual species, an assemblage of species, or a
cession dynamics), environmental regimes and constraints natural community or ecological system. These attributes

854 BioScience • September 2003 / Vol. 53 No. 9

 Articles

include the abundance of species and the overall spatial ex- ulation within the protected area and make them the focus
tent (range) of the target. Noss (1990) and Karr and Chu of conservation management and measurement.
(1999) summarize the types of key attributes of composition
relevant to these different scales of biological organization. Defining when a target is conserved: The acceptable
Key biotic interactions and processes are those that sig- range of variation
nificantly shape the variation in a target’s biological compo- The proposed scorecard framework defines a conservation tar-
sition and its spatial structure over space and time. These may get as conserved when all of its key ecological attributes are
include not only interactions among specific species and maintained or restored within some explicitly delineated
functional groups but also broad ecological processes that range of variation over space and time, the limits of which con-
emerge from the interactions among biota and between biota stitute the minimum conditions for persistence of the target.
and the physical environment. Examples include productiv- We suggest calling this range of conditions the acceptable
ity, nutrient cycling, distribution of biomass among trophic range of variation for a target’s key ecological attributes.
levels, biological mediation of physical or chemical habitat, Species, natural communities, and ecological systems all
and potential for trophic cascades (e.g., Pace et al. 1999, evolve over time within dynamic environments, and most of
Scheffer et al. 2001). their ecological attributes have some temporal variation
Key environmental regimes and constraints (including (Landres et al. 1999). For example, there is natural variation
both normal and extreme variation) shape physical and in the age and species composition of a forest canopy and in
chemical habitat conditions and thereby significantly shape the frequency and intensity of fires or flooding regimes. This
the target’s biological composition and structure over space variation is not random; rather, it is limited to a particular
and time. Examples include attributes of (a) weather patterns; range that is recognized as either (a) natural and consistent
(b) soil moisture and surface and groundwater regimes; (c) with the long-term persistence of each target or (b) outside
fire regimes; (d) water circulation patterns in lakes, estuaries, the natural range because of human influences (e.g., fire
and marine environments; (e) soil erosion and accretion; suppression in fire-maintained systems) or other drastic en-
and (f) geology and geomorphology. vironmental change. Further, the natural variation of the
Key attributes of landscape structure and architecture physical environment and biotic interactions within that en-
comprise a special subset of environmental constraints, in- vironment create a dynamic template that shapes how species
cluding connectivity and proximity among biotic and abiotic evolve and what species may (or may not) be able to persist
features of the landscape at different spatial scales (e.g., in any given area. Managing conservation targets based on the
Holling 1992). Such constraints, for example, may affect the concept of an acceptable range of variation, therefore, is im-
ability of that landscape to sustain crucial habitat requirements portant both for ensuring the persistence and integrity of a
of individual species, sustain processes that transport habitat- protected area’s biological diversity and for safeguarding
forming matter (nutrients, sediment, plant litter) across the species’ evolutionary potential (Christensen et al. 1996,
landscape, and permit recolonization of disturbed locations Holling and Meffe 1996, Poff et al. 1997).
and demographic sinks. The distinction between an acceptable and a natural range
The identification of key ecological attributes also requires of variation is important. Although there has been some
the identification of the specific kinds of information, or in- theoretical and practical debate over the concept of a natural
dicators, that can be measured to inform managers of changes range of variation, it has proved a useful construct for setting
in the status of those attributes. Protected area managers benchmarks for conservation practice (Landres et al. 1999,
should select for each attribute one or more indicators that Swetnam et al. 1999, Allen et al. 2002). However, what is
meet several well-established criteria (Noss 1990, Margoluis “natural” is difficult to define, given the limited knowledge of
and Salafsky 1998, Dale and Beyeler 2001). many species and systems and the extent of human involve-
Figure 1 and table 1 identify key ecological attributes and ment in, and disturbance to, biodiversity around the globe
indicators for a sample conservation target, in this case the (Hunter 1996). Indeed, in some areas current ecological sys-
population of a species, chinook salmon (Oncorhynchus tems have no historic counterparts, because the ecological sys-
tshawytscha), in the Cosumnes River, California. The ecological tems in these areas have been so thoroughly transformed by
model for this species illustrates several critically important direct human alterations such as anthropogenic chemicals and
characteristics in the species’ life cycle. These include the size introduced species. Further, scientific knowledge of most
of unfragmented floodplain habitat for rearing young of the ecological systems and species has a relatively short history,
year; the connectivity between ocean, delta, floodplain, and as does the preserved record of most environmental regimes
river; and the river bottom (riffle) structure during spawn- (e.g., weather, fire, hydrology). As researchers begin to un-
ing. Research and expert opinion confirm that these ecolog- derstand the ways in which ecosystems can be naturally dy-
ical attributes, if degraded, would rapidly result in the loss of namic on time scales not only of years and decades but also
ecological integrity to an extent that could make recovery sub- of centuries (e.g., Holling 1992, Swetnam et al. 1999) and mil-
stantially more difficult. With explicit documentation of data lennia, it becomes apparent that human knowledge of the nat-
sources and decisions, managers can identify these charac- ural range of variation in populations, communities, and
teristics as the key ecological attributes for the salmon pop- ecological systems arises from only a small sample of time.

September 2003 / Vol. 53 No. 9 • BioScience 855

Articles

Table 1. Format for the assessment of key ecological attributes for a conservation target, in this case a species target,
chinook salmon (Oncorhynchus tshawytscha), for the Cosumnes River Protected Area, California, as illustrated in figure 1.
Rating
Key ecological Very Current
attribute Indicator Poor Fair Good good rating
Habitat size: Areas of floodplain 0 acres of 0–100 acres 101–1000 acres > 1000 acres Good
rearing rearing habitat floodplain habitat of habitat of habitat of habitat

Habitat size: River miles of spawning Latrobe Falls to Latrobe Falls to Latrobe Falls to Latrobe Falls to Fair
spawning habitat with at least one Granlees Dam Highway 16 Schneider property Meiss Road
functional spawning riffle (5.5 miles) (7 miles) (8 miles) (14 miles)
(with young-of-year production)

Migration: Magnitude and timing of No connectivity between Periods of flow Periods of flow Periods of flow Fair
passage flows fall flows the delta and spawning of 60 cfs at 60–200 cfs during > 200 cfs during
habitat Michigan Bar migration season; migration season;
during migration at least 25 days > 25 days of
season; at least of duration duration
10 days of duration
Habitat structure: Substrate composition > 50% fine sediment 10%–50% fine sedi- Approximately 80% 80% gravel, 20% Fair
spawning of riffles ment, 50%–90% gravel, 20% cobble, cobble, no fine
gravel and cobble some fine sediment sediment

Recruitment: Abundance of juveniles 0–0.10 catch per hour 0.11–0.25 catch 0.26–1 catch per > 1 catch per hour Fair
juvenile in a rotary screw trap per hour hour
abundance

cfs, cubic feet per second; DFG, Department of Fish and Game; TNC, The Nature Conservancy; USFWS, US Fish and Wildlife Service.

Therefore, defining what is natural may be infeasible for gradations are possible, helping to set more precise conser-
many, if not all, conservation targets. vation objectives and measure their progress over time. For
Direct knowledge of the natural range of variation, there- example, a target for which considerable effort is needed to
fore, is merely one source of information for developing hy- keep one or more key attributes within their acceptable ranges
potheses about the desired status for key ecological attributes. of variation should be considered less well conserved than a
Other sources of such information include ecological mod- target for which less active management is required. Similarly,
els, expert knowledge, and comparisons among other exam- among targets that do not meet the minimum criteria for be-
ples of the same or similar species, communities, or ecolog- ing conserved, those for which one or more key attributes are
ical systems. Particularly where such examples have been trending further away from, rather than back toward, their ac-
affected by human impacts of varying types and magnitudes, ceptable ranges of variation would typically be considered
comparisons can be especially informative about where the more imperiled than those for which this is not the case.
limits may lie beyond which the persistence of the target And targets for which one or more key attributes are ap-
may be at risk. proaching or have exceeded some threshold of recoverabil-
The term acceptable range, unlike natural range, draws at- ity altogether—or require a massive effort to prevent such fail-
tention to the idea that it is not necessary to characterize all ure—would be considered more imperiled than those whose
the details of an attribute’s hypothesized range of variation. key attributes are all within the recoverable range. We rec-
Instead, researchers need only to characterize certain outer lim- ommend using a hierarchical, four-part rating scale to cap-
its that define the envelope of conditions within which the tar- ture these distinctions in conservation status and to represent
get should be expected and allowed to vary over time (e.g., the quantitative ecological data behind the assessment in a way
Christensen et al. 1996, Holling and Meffe 1996). The resulting that is intelligible to nonspecialists (figure 2).
acceptable range of variation, while most likely not replicat-
ing prehuman conditions, will ensure the long-term persis- Assessing conservation target status
tence of the target (Swetnam et al. 1999, Egan and Howell Assessing the status of each conservation target involves
2001). assembling information on the indicators for all of its key eco-
The concept of an acceptable range of variation for key logical attributes and determining the appropriate status
ecological attributes establishes the minimum criteria for rating for each attribute (figure 2). The ratings for the indi-
identifying a conservation target as conserved. Additional vidual attributes can be combined to generate an overall

856 BioScience • September 2003 / Vol. 53 No. 9

 Articles

Table 1. (continued)

Basis for Management Current status Basis for

Goal indicator rating objective (date) current rating Comments
Good TNC knowledge of floodplain Rearing habitat provided Approximately 750 acres TNC data Reflects ecological health.
connectivity; area based on for the young-of-year pro- (November 2001) Monitoring information
run size data (Yoshiyama et al. duced in the river (given needs can be determined
1998; Peter Moyle, University the current low numbers, during the course of other
of California–Davis, personal 100–1000 acres is duties.
communication, 2000) adequate)
Good TNC comparison of historical Some spawning occurrences Seven miles of good TNC data Reflects ecological health.
spawning range to current spawn- in full historical range spawning habitat (2000) Monitoring information
ing range (Keith Whitener, Con- needs can be determined
sumes River Preserve, Galt, CA, during the course of other
personal communication, 2000) duties.
Good TNC ground-truthing of passage No passage restrictions Good (1999), fair (2000); TNC data Reflects threat abate-
conditions on an annual basis during migration season variable conditions year ment. Approximately 4
(Keith Whitener Consumes River to year half-days to ground truth
Preserve, Galt, CA, personal for both migration factors.
communication, 2000)

Good Evidence that fine sediment 80% gravel, 20% cobble, Fine sediment dominant; TNC and DFG data Reflects ecological health
deposition causes damage to some fine sediment (in waiting for DFG 2000 data and threat abatement.
incubating eggs and disrupts an undammed river some report
the food web (USFWS species fine sediments are expected,
profiles) as there is no trap)
Very Numbers calculated from Number of juveniles 0.12 catch per hour DFG data Reflects ecological health.
good limited DFG data for the increased to at least 1 per (1998); waiting for recent DFG has not committed
Cosumnes River; calculations hour by 2010 data to monitoring beyond
based on target numbers 2001.
for adults

conservation status rating for each target. The

ratings for all targets, in turn, can be combined
to generate an overall rating for the protected
area. At each step in this tabulation, the defini-
tion of what it means for a target to be con- All key ecological
served must be maintained. That is, if any at- attributes within acceptable
tribute for a target lies outside its acceptable ranges of variation?
range of variation, then that target itself cannot
be considered conserved. If any target is con-
sidered not to be conserved, then challenges still
persist for the adequate management of the pro- Target is not conserved Target is conserved
tected area, and resources should be redirected
as feasible toward restoring the conservation
target to conserved status.
Table 1 shows the rating information for a Major intervention Major intervention
sample of the key ecological attributes of the required to restore all required to maintain all
salmon target presented in figure 1. It illustrates attributes to acceptable attributes within acceptable
a simple format for recording the information ranges of variation? ranges of variation?
used to develop a scorecard. Maintaining such a
record is invaluable for documenting manage-
ment outcomes and maintaining institutional
memory as conditions change and monitoring
Target rating: Target rating: Target rating: Target rating:
data accumulate over time. Poor Fair Good Very good

Why use this framework?

The measurement framework presented here
provides not only a standardized basis for Figure 2. Basic decision tree for rating the status of a conservation target.

September 2003 / Vol. 53 No. 9 • BioScience 857

Articles

measuring the effectiveness of protected area management but forest to industrial forest plantations of alien species may
several additional benefits for parks management. leave only remnants of the original forest that are too small
to support adequate populations of the pollinators and seed
It focuses strategy development along ecological, rather dispersers needed for successful regeneration. The identifi-
than jurisdictional, boundaries. Protected area management cation and assessment of key ecological attributes thus pro-
often focuses on implementing strategies only within the vides a means for identifying not only what threats require
protected area boundaries, failing to consider issues of land- abatement but what altered attributes require restoration in
scape connectivity and ecological processes that extend out- order to achieve conservation goals.
side the park. Such strategies can achieve only limited success
in reaching the park’s ultimate objectives of conserving its eco- It helps identify crucial research needs. Knowledge of the key
logical integrity. The proposed framework can help identify ecological attributes of conservation targets will always be lim-
the true spatial scales at which each target’s key ecological at- ited at best. This uncertainty may lead to the abandonment
tributes function, irrespective of jurisdictional boundaries, so of conservation strategies for poorly understood targets or to
that conservation investments can take these scales more a desperate quest for information on the target, without a clear
fully into account. sense of the questions that must be answered to improve
conservation management effectiveness (Feinsinger 2001). The
It provides consistency and specificity in setting conservation proposed framework helps identify the most critical questions
objectives. Setting clear, specific, measurable objectives for both and knowledge gaps for protected area management by specif-
threats and target integrity is essential for effective conservation ically challenging management teams to identify the key eco-
management (Noss 1990, Margoluis and Salafsky 1998). The logical attributes and their acceptable ranges of variation for
proposed framework helps set these benchmarks consistently each target (Sayre et al. 2000).
across all targets, establishing what attributes of the target need
to be managed, how much change is needed to achieve con- It promotes focused and efficient monitoring programs.
servation success, and how to measure progress toward that Monitoring programs to support the measurement of bio-
end. Without such a framework, it is only too easy to adopt diversity and threat status are globally recognized as crucial
vague objectives such as “conserve the southern sea otter, elements of any protected area management program (Noss
Lontra felina.” With the framework, managers can establish 1990, Brandon et al. 1998).Yet most monitoring programs are
a more specific, scientifically defensible objective, such as opportunistic rather than strategic, because it is unclear what
“by 2010, double prey quantity and diversity for L. felina as should be monitored, where, and why (Dale and Beyeler
compared to 2003 baselines in order to achieve a minimum 2001). The proposed framework helps identify exactly which
viable population size of 5 to 10 individuals per square kilo- of a target’s ecological characteristics need to be monitored,
meter of protected coastline.” encourages the identification of measurable indicators, and
helps park managers answer logistical questions concerning
It enhances the identification and anticipation of threats to where and when an indicator should be monitored, given its
biodiversity. For every key ecological attribute of a target that likely spatial distribution, natural fluctuations, and suscepti-
is altered beyond its acceptable range of variation (e.g., in- bility to change.
sufficiently frequent flooding cycles in a riparian system),
there is a cause that must be addressed. The proposed frame- Implementing the framework
work provides a means of systematically identifying and rat- Implementing the proposed framework may appear to be a
ing the status of possible threats. It therefore provides a means daunting task. Conservation planners are likely to have little
of ensuring that protected area management does not focus confidence in quantitative descriptions of the acceptable
only on highly visible but not necessarily significant threats range of variation for a key ecological attribute, and they
(such as conversion of native grassland to exurban develop- may possess no information at all on a target’s ecology, let alone
ment on a portion of the reserve) and overlook other less vis- specific data for rating a target’s status. Nevertheless, the
ible yet potentially more destructive threats (e.g., invasive framework challenges researchers and planners to advance
species that affect the composition of a grassland community their conservation work with hypotheses based on the best
across an entire park). available information and to guide conservation manage-
ment while documenting all assumptions and information
It promotes the development of comprehensive conservation gaps. Even where local scientific data are thin, it will be
strategies. Conservation strategies based on ad hoc percep- possible to find experts who are familiar with the general
tions of threats to protected areas almost inevitably focus all composition, structure, and function of the biodiversity in
resources on threat abatement (Bryant et al. 1998, Salafsky and question or with similar systems for comparison. Such expert
Margoluis 1999b).Yet frequently the abatement of a threat will knowledge can serve as the basis for moving forward with
not improve the status of a conservation target, because it ig- hypotheses toward the goals of a rigorous assessment of
nores the need to restore other crucial ecological attributes. ecological integrity and a thoughtful program of adaptive
For example, curbing the conversion of native temperate management. The data gaps, in turn, can serve as a menu for

858 BioScience • September 2003 / Vol. 53 No. 9

 Articles
Castro G, Locker I. 2000. Mapping Conservation Investments: An Assessment
motivating and directing research toward the most pressing
of Biodiversity Funding in Latin America and the Caribbean. Washing-
conservation questions. With such a framework in hand, ton (DC): Biodiversity Support Program.
park managers can ensure scientific rigor in their planning and Christensen NL, et al. 1996. The report of the Ecological Society of
management, promote consistency in the evaluation of man- America Committee on the Scientific Basis for Ecosystem Management.
agement effectiveness, and begin to track the impacts of Ecological Applications 6: 665–691.
Courrau J. 1999. Monitoring protected area management in Central
management actions. With sound measures of ecological in-
America: A regional approach. Parks 9: 56–60.
tegrity and species viability, managers of protected areas can Dale VH, Beyeler SC. 2001. Challenges in the development and use of
begin to say with greater confidence whether or not they are ecological indicators. Ecological Indicators 1: 3–10.
conserving what they say they are. Dudley N, Hockings M, Stolton S. 1999. Measuring the effectiveness of
protected area management. Pages 249–257 in Stolton S, Dudley N,
eds. Partnerships for Protection. London: Earthscan.
Acknowledgments Egan D, Howell EA. 2001. The Historical Ecology Handbook: A Restora-
This article reflects the innovative work and experience of tionist’s Guide to Reference Ecosystems. Washington (DC): Island Press.
thousands of conservation practitioners and protected area Ervin J. 2002. Rapid Assessment and Prioritization of Protected Area Man-
managers within The Nature Conservancy and its conserva- agement (RAPPAM) Methodology. Gland (Switzerland): World Wide
tion partners. It specifically rests on the work of The Nature Fund for Nature.
Feinsinger P. 2001. Designing Field Studies for Biodiversity Conservation.
Conservancy’s Ecosystem Integrity Working Group, whose
Washington (DC): Island Press.
members included Mike Beck, Jenny Brown, Pat Comer, Jeff Harwell MA, et al. 1999. A framework for an ecosystem integrity report card.
Hardesty, Shirley Keel, Ricardo Soto, Tim Whittier, and the BioScience 49: 543–556.
authors. We thank The Nature Conservancy’s Freshwater Hockings M. 1998. Evaluating management of protected areas: Integrating
Initiative for serving as a laboratory for the new framework. planning and evaluation. Environmental Management 22: 337–346.
———. 2003. Systems for assessing the effectiveness of management in
We are grateful to members of The Nature Conservancy’s protected areas. BioScience 53: 823–832.
Conservation Measures and Audit Team, including Doria Hockings M, Stolton S, Dudley N. 2000. Evaluating Effectiveness: A Frame-
Gordon, Alex Mas, Wendy Millet, Brad Northrup, Nick Salaf- work for Assessing the Management of Protected Areas. Cambridge
sky, Dan Salzer, M. Sanjayan, Stacey Solie, Tim Tear, and (United Kingdom): IUCN.
Steve Watkins, for their promotion of and improvements to Hockings M, Stolton S, Dudley N, Parrish JD. 2001. Enhancing Our Heritage
Toolkit: The Workbook on How to Build Assessment, Monitoring, and
the framework. We thank M. Sanjayan and Bradford Northrup Reporting Systems on Management Effectiveness of World Heritage
for their leadership in supporting the development of these Sites. (18 August 2003; www.enhancingheritage.net/docs_public.asp)
concepts within the organization. The conservation team at Holling CS. 1992. Cross-scale morphology, geometry, and dynamics of
the Cosumnes River landscape in California, under the di- ecosystems. Ecological Monographs 62: 447–502.
Holling CS, Meffe GK. 1996. Command and control and the pathology of
rection of Ramona Swenson, generously provided the salmon
natural resource management. Conservation Biology 10: 328–337.
example herein. Richard Margoluis, Nick Salafsky, Tim Tear, Hunter ML. 1996. Benchmarks for managing ecosystems: Are human activities
Doria Gordon, Peter Kareiva, John Wiens, Jeff Baumgartner, natural? Conservation Biology 10: 695–697.
Jamison Ervin, and our anonymous reviewers provided ex- Karr JR, Chu EW. 1999. Restoring Life in Running Waters: Better Biologi-
traordinarily helpful reviews of this paper. We greatly ap- cal Monitoring. Washington (DC): Island Press.
Karr JR, Dudley DR. 1981. Ecological perspective on water quality goals.
preciate Nicole Rousmaniere’s assistance in preparing the fi-
Environmental Management 5: 55–68.
nal figures for this paper, and we are grateful for the support Landres PB, Morgan P, Swanson FJ. 1999. Overview of the use of natural vari-
of the Colorado chapter of The Nature Conservancy (J. D. P.), ability concepts in managing ecological systems. Ecological Applica-
Lisa Mickley (J. D. P.), and The Nature Conservancy’s Fresh- tions 9: 1179–1188.
water Initiative (D. P. B.) and Rocky Mountain Division Maddox D, Poiani K, Unnasch R. 2001. Evaluating management success:
Using ecological models to ask the right monitoring questions. Pages
(R. S. U.).
563–584 in Sexton WT, Malk AJ, Szaro RC, Johnson NC, eds. Ecologi-
cal Stewardship. Oxford (United Kingdom): Elsevier Science.
References cited Margoluis R, Salafsky N. 1998. Measures of Success: Designing, Managing,
Allen CD, Savage M, Falk DA, Suckling KF, Swetnam TW, Schulke T, Stacey and Monitoring Conservation and Development Projects. Washington
PB, Morgan P, Hoffman M, Klingel JT. 2002. Ecological restoration of (DC): Island Press.
southwestern ponderosa pine ecosystems: A broad perspective. Ecolog- ———. 2001. Is Our Project Succeeding? A Guide to Threat Reduction
ical Applications 12: 1418–1433. Assessment for Conservation. Washington (DC): Biodiversity Support
Angliss M, et al. 2001. State of the Rivers Report: Letaba and Luvuvhu River Program.
Systems 2001. Pretoria (South Africa): Water Research Commission. Noss RF. 1990. Indicators for monitoring biodiversity: A hierarchical approach.
Report TT 165/01. (18 August 2003; www.csir.co.za/rhp/) Conservation Biology 4: 355–364.
Brandon K, Redford KH, Sanderson SE. 1998. Parks in Peril: People, Poli- ———. 1996. Ecosystems as conservation targets. Trends in Ecology and
tics, and Protected Areas. Washington (DC): Island Press. Evolution 11: 351.
Bryant D, Burke L, McManus J, Spalding M. 1998. Reefs at Risk: A Map-Based Noss RF, Cooperrider AY. 1994. Saving Nature’s Legacy: Protecting and
Indicator of Threats to the World’s Coral Reefs. Washington (DC): Restoring Biodiversity. Washington (DC): Island Press.
World Resources Institute. Noss RF, LaRoe ET, Scott JM. 1995. Endangered Ecosystems of the United
Carignan V, Villard M-A. 2002. Selecting indicator species to monitor eco- States: A Preliminary Assessment of Loss and Degradation. Washington
logical integrity: A review. Environmental Monitoring and Assessment (DC): US Department of the Interior, National Biological Service.
78: 45–61. Biological Report no. 28.

September 2003 / Vol. 53 No. 9 • BioScience 859

Articles
O’Connell TJ, Jackson LE, Brooks RP. 1998. The Bird Community Index: Sayre R, Roca E, Sedagatkish G, Young B, Keel S, Roca R, Sheppard S. 2000.
A Tool for Assessing Biological Integrity in the Mid-Atlantic Highlands. Nature in Focus: Rapid Ecological Assessment. Washington (DC): Island
University Park (PA): Penn State Cooperative Wetlands Center, Penn- Press.
sylvania State University. Report no. 98-4. Scheffer M, Carpenter S, Foley JA, Folke C, Walker B. 2001. Catastrophic shifts
Pace ML, Cole JJ, Carpenter SR, Kitchell JF. 1999. Trophic cascades revealed in ecosystems. Nature 413: 591–596.
in diverse ecosystems. Trends in Ecology and Evolution 14: 483–488. Schwartz MW. 1999. Choosing the appropriate scale of reserves for conser-
Poff NL, Allan JD, Bain MB, Karr JR, Prestegaard KL, Richter BD, Sparks RE, vation. Annual Review of Ecology and Systematics 30: 83–108.
Senge PM. 1994. The Fifth Discipline: The Art and Practice of the Learning
Stromberg JC. 1997. The natural flow regime. BioScience 47: 769–784.
Organization. New York: Currency Doubleday.
Poiani KA, Baumgartner JV, Buttrick SC, Green SL, Hopkins E, Ivey GD,
Stein BA, Kutner LS, Adams JS. 2000. Precious Heritage: The Status of
Seaton KP, Sutter RD. 1998. A scale-independent site conservation plan-
Biodiversity in the United States. New York: Oxford University Press.
ning framework in The Nature Conservancy. Landscape and Urban Swetnam TW, Allen CD, Betancourt JL. 1999. Applied historical ecology:
Planning 43: 143–156. Using the past to manage for the future. Ecological Applications 9:
Poiani KA, Richter BD, Anderson MG, Richter HE. 2000. Biodiversity con- 1189–1206.
servation at multiple scales: Functional sites, landscapes, and networks. [TNC] The Nature Conservancy. 1998. The Parks in Peril Site Consolidation
BioScience 50: 133–146. Scorecard Manual. (18 August 2003; http://parksinperil.org/05tools. shtml)
Redford KH, Taber A. 2000. Writing the wrongs: Developing a safe-fail ———. 2000a. The Five-S Framework for Site Conservation: A Practi-
culture in conservation. Conservation Biology 14: 1567–1568. tioner’s Handbook for Site Conservation Planning and Measuring
Salafsky N, Margoluis R. 1999a. Greater than the Sum of Their Parts: Designing Success. 2nd ed. Arlington (VA): TNC.
Conservation and Development Programs to Maximize Results and ———. 2000b. Conservation by Design: A Framework for Mission Success.
Learning. Washington (DC): Biodiversity Support Program. Arlington (VA): TNC.
———. 1999b. Threat reduction assessment: A practical and cost-effective ———. 2003. Measures and Audit Program, Project Measures Workshop,
approach to evaluating conservation and development projects. Dec. 2002: Assessing the Impact of Project Strategies and Actions.
Arlington (VA): TNC. (29 July 2003; www.conserveonline.org/
Conservation Biology 13: 830–841.
experimental/m&a_workshop)
Salafsky N, Margoluis R, Redford KH. 2001. Adaptive Management: A Tool
[WRI] World Resources Institute. 2000. People and Ecosystems: The
for Conservation Practitioners. Washington (DC): Biodiversity Support
Fraying Web of Life. Washington (DC): WRI.
Program. Yoshiyama RM, Fisher FW, Moyle PB. 1998. Historical abundance and de-
Salafsky N, Margoluis R, Redford KH, Robinson JG. 2002. Improving the prac- cline of chinook salmon in the Central Valley region of California. North
tice of conservation: A conceptual framework and research agenda for American Journal of Fisheries Management 18: 487–521.
conservation science. Conservation Biology 16: 1469–1479. Young TF, Sanzone S, eds. 2002. A Framework for Assessing and Reporting
Sanderson EW, Redford KH, Vedder A, Coppolillo PB, Ward SE. 2002. A on Ecological Condition: An SAB Report. Washington (DC): EPA
conceptual model for conservation planning based on landscape species Science Advisory Board. Report no. EPA-SAB-EPEC-02-009. (29 July 2003;
requirements. Landscape and Urban Planning 58: 41–56. www.epa.gov/sab)

860 BioScience • September 2003 / Vol. 53 No. 9