Content uploaded by Simone Stevenson
Author content
All content in this area was uploaded by Simone Stevenson on Apr 01, 2021
Content may be subject to copyright.
This article has been accepted for publication and undergone full peer review but has not been
through the copyediting, typesetting, pagination and proofreading process, which may lead to
differences between this version and the Version of Record. Please cite this article as doi:
10.1111/cobi.13575.
This article is protected by copyright. All rights reserved.
Matching biodiversity indicators to policy needs
Corresponding author
Simone L. Stevenson1
1 Deakin University, School of Life and Environmental Sciences, Centre for
Integrative Ecology, 221 Burwood Highway, Burwood, Victoria 3125, Australia.
ssteven@deakin.edu.au
Co-authors
Kate Watermeyer1, Giovanni Caggiano2, Elizabeth A. Fulton3,4, Simon Ferrier5 &
Emily Nicholson1
Affiliations
1 Deakin University, School of Life and Environmental Sciences, Centre for
Integrative Ecology, Victoria, Australia
2 Department of Economics, Monash University, Caulfield East, Victoria 3145
3 CSIRO Oceans and Atmosphere, GPO Box 1538, Hobart, Tasmania 7001
4 Centre for Marine Socioecology, University of Tasmania, Hobart, Tasmania 7001
5 CSIRO Land and Water, Canberra, ACT, Australia
Running head
Biodiversity Indicators
Keywords
Aichi Targets, Biodiversity Indicators, Sustainable Development Goals (SDGs),
Convention on Biological Diversity (CBD), Ecosystem Based Fisheries Management
(EBFM), IUCN Red List Index, Extinction, Population dynamics
Article Impact Statement
Some biodiversity indicators are best suited to prediction for preventative action and
others to retrospection and evaluating past actions.
This article is protected by copyright. All rights reserved.
2
Abstract
Policy makers rely on biodiversity indicators to assess when, where and how nature
is changing. Some indicators, however, respond more quickly to pressures than
others, measuring short-term and potentially reversible change, while others capture
permanent loss of biodiversity. These characteristics influence an indicator’s
suitability to perform predictive versus retrospective evaluation functions, but are
rarely considered when developing or interpreting indicators. We demonstrate how a
conceptual model from economics can be adapted for biodiversity to classify
indicators by three functions: Leading (indicators that change prior the subject of
interest, thus informing preventative measures); Coincident (indicators that measure
the subject of interest); or Lagging (indicators that signal change after the subject of
interest, used to evaluate past actions). We use this approach to classify existing
indicators for two case studies: global species extinction and marine ecosystem
collapse. We find that many existing indicators are theoretically capable of
performing different functions; data analysis will be required to confirm temporal
relationships between indicators. Classifying indicators according to function
enhances interpretability, supports preventative action and facilitates structured
decision-making.
This article is protected by copyright. All rights reserved.
3
Introduction
The world is experiencing widespread, rapid declines in biodiversity (Tittensor et al.
2014). The Convention on Biological Diversity’s Aichi Targets (CBD 1992, 2010) and
the United Nations Sustainable Development Goals (SDGs) (UNGA 2015) convey
global aspirations for the state of biodiversity. Decision-makers use indicators to
measure progress and inform action towards these targets (Tittensor et al. 2014).
Indicators are often implemented, however, without a clear function or consideration
of suitability for purpose (Collen & Nicholson 2014), inhibiting their usefulness in
decision-making (Mace & Baillie 2007). For example, there are increasing calls to
use indicators when predicting future conditions, thereby avoiding permanent
biodiversity loss (Collen & Nicholson 2014), but not all indicators are suited to this
function. Indicators measuring slow-to-emerge or permanent biodiversity change,
such as extinctions, are of limited use for preventing irreversible loss but remain
important for evaluating past actions (Pereira et al. 2012).
At the global scale, biodiversity indicators are primarily used to monitor general
trends (Tittensor et al. 2014), produce data and research, (Tittensor et al. 2014), and
raise awareness amongst the public and policy makers (Leadley et al. 2014).
Indicators remain under-used in active decision-making, such as evaluating past
management or policy actions (Hoffmann et al. 2015), informing policy decisions
(Jones et al. 2011), or setting biodiversity targets (Done & Reichelt 1998). A
nebulous approach to biodiversity monitoring is evident in the assignment of the Red
List Index (RLI) to monitor 9 very different Aichi Targets (CBD 2016). Further,
biodiversity indicators are usually grouped under a single function (e.g. measuring
‘state’) in existing policy frameworks (e.g. Driver-Pressure-State-Impact-Response
This article is protected by copyright. All rights reserved.
4
(DPSIR)), (Kristensen 2004; IPBES 2016). Such broad functions assume all
biodiversity indicators are equally capable of performing both predictive and
evaluative functions, which is rarely true (Pereira et al. 2012).
Other disciplines of complex systems, such as economics, offer valuable lessons for
enhancing the role of biodiversity indicators in active decision-making. The
economic theory used to measure and manage economic growth (Burns & Mitchell
1946) has considerable potential for defining conservation objectives, and matching
appropriate indicators to the required predictive or evaluative function. Economists
aim to maintain a state of stable economic growth (‘equilibrium expansion’) and
avoid economic recessions or excessive growth. Sets of multiple indicators are used
to confirm which state the economy is in, and predict transitions to a different state
(Stock & Watson 2003); this facilitates pre-emptive action if, for example, a recession
is predicted (Marcellino 2006). Conservation is similarly concerned with predicting
and confirming the state, trajectory and transitions of biodiversity. While suites of
biodiversity indicators have been proposed in the past (Noss 1990; Sparks et al.
2011), none have incorporated the differing speeds at which various components of
biodiversity react to change, nor the temporal relationships between them (Wu
1999).
Economists classify and interpret indicators by intended function, defined by their
temporal relationships to economic growth, and each other (Burns & Mitchell 1946).
Such a classification is possible because the target of management and prediction
(economic growth), its possible states, and the transitions between them have been
defined. Indicators that measure economic growth (often the indicator Gross
Domestic Product) are classified as coincident indicators (Marcellino 2006).
This article is protected by copyright. All rights reserved.
5
Indicators of other variables within the economy that typically change before the
state of economic growth itself, and thus perform predictive functions, are referred to
as leading indicators (e.g., the stock market) (Stock & Watson 2003). Lastly,
indicators that change after economic growth are lagging indicators (e.g., interest
rates), and play an important role in confirming recessions and evaluating past
actions (Zarnowitz & Boschan 1975).
In this paper, we aim to facilitate the theoretical classification of biodiversity
indicators by function, thereby improving their usefulness in conservation decisions.
When classifying indicators, economists begin with a conceptual model describing
how different components of the economy change over time. They use this model to
propose the expected order of change in indicators, then confirm the sequence of
change empirically (Marcellino 2006). We adapt this classification system by
defining the five key elements required to develop a comparable conceptual model of
biodiversity change over time. We explain how the resultant model can be used to
select and theoretically classify biodiversity indicators as leading, coincident and
lagging. Lastly, we demonstrate by developing two conceptual models of
biodiversity change – global species extinctions and collapse of fished marine
ecosystems – and use each model to classify a set of indicators. Classifying
indicators by their leading or lagging capacity is not intended to replace indicator
selection criteria (e.g. Mace and Baillie (2007)), but rather make biodiversity
indicators more informative.
This article is protected by copyright. All rights reserved.
6
Applying the concept of indicator classification to biodiversity
indicators
Classifying biodiversity indicators by function presents some challenges, because
there is no single, universally agreed target of conservation management akin to
economic growth, nor an existing conceptual model developed for this purpose. Like
the economy, biodiversity is inherently complex and multifaceted, and the
importance of different biodiversity variables can differ across space, time and value
systems (Duelli & Obrist 2003). Likewise, the conceptual model of biodiversity
change will be different depending on management objectives, and may bear little
resemblance to the cyclical process of economic expansion and recession. We
therefore describe the process of developing a model and classifying indicators
accordingly.
Five elements need to be defined in order to develop a conceptual classification
model, starting with the process of interest and component of biodiversity to be
predicted and managed. The first four elements we describe are essential for
classifying indicators, while the last is not essential for classification itself, but
required to convert the model and indicators into performance measures. The
definition of each element is likely to require iterative development rather than
forming a clear linear sequence.
1. Identify the objective of monitoring and the process of change to be
measured. In economics, the objective is to manage economic growth, and
the process of change is the co-movement of economic activities over time
(Koopmans 1947). The process of biodiversity change does not need to be
This article is protected by copyright. All rights reserved.
7
cyclical (switching between recurrent states) like the economy; it can be
directional or hierarchical, where the system transitions through states
progressively, rarely or never revisiting a previous state. Examples of such
processes include species extinction, or climate-driven distribution change.
2. Define important system dynamics, including states through which the
system passes during the process, and their sequence. While
management targets often focus on a specific component of biodiversity (e.g.
species richness), these components rarely change in isolation, but rather as
part of a broader process (Pereira et al. 2010). States are abstractions
encompassing a defined amount of variation in time (Westoby et al. 1989),
meaning that a process of change can be broken into user-defined states
describing any relevant component of biodiversity change. In the economy,
the defined states are mutually exclusive, being: stable growth, excessive
growth, or recession (Stock & Watson 2003). A biodiversity example might be
an ecosystem that transitions between alternate stable states characterised by
different dominant vegetation (Westoby et al. 1989). Alternatively, the process
of change may be hierarchical, where states are defined when change
progresses through different levels of biological organisation (e.g. loss of
individuals, then local populations, then species). Change to lower levels of
biological organisation occur on shorter time scales, accumulating over time
until it emerges as change to the higher levels as well (Simon 1962; Noss
1990; Wu 1999). The states and their sequence need to be grounded in
scientific theory and logic to facilitate selection of useful biodiversity indicators.
3. Identify the target variable(s). The target variable serves as an anchor for
classifying indicators because the designation of ‘leading’ and ‘lagging’ are
This article is protected by copyright. All rights reserved.
8
relative. The target variable should represent the subject of prediction and
evaluation, and can be any valued component of biodiversity that changes
within the defined process. For example, while the economy encompasses
many interacting components such as employment and housing, the primary
target of prediction and evaluation is the trend in economic growth. An
equivalent target for biodiversity could be species extinctions, for example.
More than one target variable can be chosen if there are multiple objectives or
components of interest. Indicator classification needs to be repeated for each
target variable, which can mean the same indicator acting in different
functions, depending on the management question. The target variable and its
relationship with the rest of the system forms the conceptual model.
4. Use the model to identify and classify indicators. This conceptual model
is equally applicable to modelled or empirically-derived indicators. Indicators
should be identified and assigned to one of the states if their input variables or
modelled outputs fall within the type of variation defining that state. Once
indicators are assigned to states, examining indicator trends or status can
convey information about what state the process is in. Indicators can be
identified from literature or past observation of the system. States without
appropriate indicators should still be included in the model to emphasise
monitoring gaps. Indicators that are constructed from, or model, the target
variable can be classified as coincident. All other indicators are then classified
according to their chronological relationship with the state measured by the
coincident indicator(s). Indicators matched to states occurring earlier than the
coincident indicator are classified as leading, and those matching later states
This article is protected by copyright. All rights reserved.
9
are lagging. The last element of the conceptual model cannot be defined until
coincident indicator(s) have been identified.
5. Identify a desirable state, undesirable state(s) and transition point of the
coincident indicator(s). The desirable state defines the level of variation or
value of the coincident indicator that managers aim to maintain. Undesirable
states are those which managers wish to avoid. In economics, stable economic
growth is the desired state, and recession or excessive growth are undesirable
states. The identification of desirable or undesirable biodiversity states will be a
political, economic and technical exercise highly dependent on management
objectives and context (Done & Reichelt 1998). Once the two states have been
defined, it is also possible to define the transition point(s) in the coincident
indicator that demarcate the desirable and undesirable states, thus converting
the indicator to a performance measure (Hall & Mainprize 2004). In economics,
these transition points are the turning points between stable growth, and
recession or excessive growth (Marcellino 2006).
Case study one – Global species extinctions
Identify the objective of monitoring and process of change to be measured
We selected global species extinctions as our first process of interest. Extinction is
the subject of Aichi Target 12 of the CBD, and Target 15.5 of the UN SDGs. The
extinction process has been well described by existing conceptual and quantitative
models of extinction, and is characterised by a hierarchy of changes to species
abundance, distribution and eventually extinction (Johnson 1998; Mace et al. 2018).
This article is protected by copyright. All rights reserved.
10
Define important system dynamics, including states through which the system
passes during the process, and their sequence
In this example, each state describes a change in population abundance at different
spatial scales (individual, local, global). The process of species extinction is
inherently hierarchical, and the states defined below represent the progression of
change through nested levels of biological organisation (being individuals,
populations and species), where change at lower levels also continue as the system
transitions into a new state (Simon 1962).
State 1. Changes to species abundance: Occurs when mortality of species’
individuals exceed births, or vice versa. Measures of abundance respond quickly
to pressures (Balmford et al. 2003). While abundance and distribution are highly
correlated, in the case of declines there is typically a lag between initial change in
abundance, and associated changes to distribution (Gaston et al. 2000). As
abundance declines, local populations (the higher order of organisation) become
more vulnerable to local extinctions (Johnson 1998).
State 2. Changes to species distribution: Occurs when sub-populations of a
species go locally extinct, where abundance in at least one formerly occupied
area declines to zero (Ceballos & Ehrlich 2002). Local extinctions directly affect
the likelihood of persistence, as species lose genetic and geographical diversity,
increasing their vulnerability to stochastic shocks and extinction (Johnson 1998).
State 3. Change to global extinction numbers: Occurs when the lower levels of
organisation that define a species - abundance, and distribution of local
populations - both decline to zero (Johnson 1998). Typically, there is a time lag
between declines in abundance and distribution, and extinction (Pereira et al.
This article is protected by copyright. All rights reserved.
11
2010). Unlike the previous two states, global extinction is an absorbing state that,
once entered, prevents any future changes to abundance or distribution in a
species.
Identify target variable(s)
The third element required is one or more target variables to serve as the anchor for
classification. We are interested in three variables – abundance, distribution and
extinction– setting each as the target variable in turn. This demonstrates how
multiple target variables can be identified for each process, and how indicator
classification changes relative to the designated target variable.
Use the model to identify and classify indicators
With the objective of measuring Aichi Target 12 and SDG Target 15.5, we selected
existing global indicators relevant to the process of extinction Eligible indicators
needed to be endorsed by the CBD and included on the Biodiversity Indicators
Partnership (BIP) website (CBD 2016), and could be drawn from any Aichi Target or
SDG, provided they represented change in population abundance at some scale.
We reviewed existing literature on each indicator to determine what type of variation
it measures or models (species abundance, distribution or extinction) (Table 1). Each
indicator was then assigned to the corresponding state in the process, based on
ecological theory and the indicator’s variables (Johnson 1998; Balmford et al. 2003;
Pereira et al. 2010) (indicators of abundance assigned to State 1, indicators of
distribution assigned to State 2, and measures of extinction to State 3). Most CBD
biodiversity indicators, e.g. the Living Planet Index (LPI), are aggregates of the state
of or trends in multiple species and therefore expected to approximate the process of
This article is protected by copyright. All rights reserved.
12
individual species extinction (Walpole et al. 2009). Finally, setting each state as the
target variable in turn, we assigned indicators the function of leading, coincident or
lagging, according to whether they measure change in an earlier, the same, or a
later state, than the target variable. See Figure 1 (b) for an example of classified
extinction indicators.
The Living Planet Index, for example, is driven by change in vertebrate population
abundance (Collen et al. 2009), and was therefore assigned to the state ‘Changes to
species abundance’. The Biodiversity Intactness Index can model both abundance
and richness (Newbold et al. 2016); we assigned the former to the abundance state;
the latter to the distribution state. The number of species extinctions was assigned
to the extinction state. Notably, the Red List Index (RLI) uses multiple criteria and
variables (including abundance and distribution, measured over past and future
timeframes) to determine a species’ Red List status and thus index values.. While
certainly a leading indicator of extinction, the combination of input variables makes it
difficult to assign the RLI to a single state. Therefore the RLI is included in both
abundance and distribution states (Table 1), assuming that the index could be
disaggregated by input, something not currently possible (Bubb et al. 2009).
Identify a desirable state, undesirable state and transition point of the coincident
indicator(s).
While the first four elements of this conceptual model are designed to be
generalised, the definition of a desirable and undesirable state, and specified
transition point(s) for performance management will be highly context dependent.
These elements should be developed by the experts and decision makers using the
indicators (e.g. Done and Reichelt (1998)). For example, if number of extinctions
This article is protected by copyright. All rights reserved.
13
were the coincident indicator for Aichi Target 12, decision makers may aim for no
further extinctions (desired state), with any increase in indicator value considered an
undesirable state.
Case study two - The collapse of a fished marine ecosystem
Identify the objective of monitoring and process of change to be measured
Fisheries management is evolving from a single-species to ecosystem-based
approaches that incorporate ecological interactions and processes (Larkin 1996).
Ecosystem responses to fishing, if not properly managed, may culminate in
ecosystem collapse (e.g. 1970s collapse of the Northern Benguela system (Bland et
al. 2018))
Define important system dynamics, including states through which the system
passes during the process, and their sequence
This process describes changes to a fished marine ecosystem. The first state
consists of changes in fishing pressure, and the subset of ecosystem components it
affects directly. The second state describes changes to structural and functional
properties of the whole ecosystem (fished and unfished). The third state is defined
by the same variables as the previous state, entered when these variables change
beyond a specified collapse threshold. This process is only partially hierarchical, as
changes to ecosystem structure and function could continue to emerge even if the
immediate effects of fishing cease (Babcock et al. 2010), but ecosystem collapse (or
recovery) cannot, by definition, occur without simultaneous changes to structure and
function (Bland et al. 2018).
This article is protected by copyright. All rights reserved.
14
State 1. Direct effects on fished target and non-target species: Fishing pressure
is a direct threat through removal of individuals from the ecosystem, and in this
context can be considered the beginning of the process. This state is defined by
changes driven directly by fisheries landings, methods, and strategies, which are
quick to respond to fishing pressure (Shin et al. 2010). This state may
encompass changes to biomass landed, biomass remaining in the ocean,
discards and exploitation levels (Shin et al. 2010).
State 2. Changes to ecosystem structure and function: Removal of individuals
(particularly under prolonged or intense fishing pressure) may precede indirect
effects on the wider system’s biodiversity (Babcock et al. 2010). These effects
can be driven by changed community composition, trophic or habitat interactions,
or population level changes (e.g. via fisheries-induced evolution (Jørgensen et al.
2007)). This state may include changes to species trait variables such as body
size (Babcock et al. 2010), mean trophic level, ecosystem stability (Fulton et al.,
2005) and ultimately ecosystem structure and function.
State 3. Ecosystem collapse: The endpoint of ecosystem decline. While the
specifics of collapse will be different for each ecosystem, it is characterised by
loss of the system’s defining biological and environmental features (Bland et al.
2018) and severe degradation of ecosystem structure and function. This can be
measured via ecosystem assessment and listing as collapsed under the IUCN
Red List of Ecosystems (Keith et al. 2013).
Identify target variable(s)
If monitoring a single species, biomass or abundance would (for example) make
ideal target variables. As the entire fished ecosystem is the property of interest, we
This article is protected by copyright. All rights reserved.
15
set each state of the process (direct fishing effects, changes to ecosystem structure
and function and ecosystem collapse) as the target variable in turn.
Use the model to identify and classify indicators
For this case study, we selected the list of indicators developed by the INDISEAS
initiative (established by the European Union to identify marine ecosystem indicators
that are relevant and comparable across different ecosystems) (Shin et al. 2010; Coll
et al. 2016). The indicators capture ecological effects of fishing and changes to
biodiversity and conservation status (Shin et al. 2010; Coll et al. 2016). We also
incorporated the IUCN Red List of Ecosystems as an indicator of ecosystem collapse
(Keith et al. 2013).
As with the previous case study, we reviewed the literature to identify each
indicator’s inputs or modelled outputs, then assigned them to one of the states
described in Table 2. Indicators of direct fishing effects were assigned to State 1.
Indicators of ecosystem structure and function were assigned to State 2, and
indicators measuring ecosystem collapse were assigned to State 3. Then setting
each state as the target variable in turn, assigned their function of leading, coincident
or lagging according to whether they measure change in an earlier state, the same
state, or a later state, than the target variable (Table 2). See Figure 1 (c) for an
example of classified marine indicators.
Identify a desirable state, undesirable state and transition point of the coincident
indicator(s).
As with the previous case study, the definition of a desirable and undesirable state,
and specified transition point(s) for the purposes of performance management
should be developed by the users. For example, if the coincident indicator was the
This article is protected by copyright. All rights reserved.
16
mean trophic level of the community, managers may select a specific value or range
of mean trophic level (e.g. the mean trophic level from a healthy system) as their
desired state, with any departure from this level considered an undesirable state.
Discussion
Our analysis shows that while many biodiversity indicators are coincident or lagging
relative to the target variable, there are also indicators capable of triggering action
before permanent change occurs. Multiple indicators of species extinctions and
marine ecosystem collapse can function as leading indicators. Our classification
demonstrates how indicators can be explicitly assigned to functions, clarifying their
intended purpose of cautionary signal or retrospective evaluation. This extends the
current use of biodiversity indicators, and could support fundamental management
improvements; creating a transparent link between monitoring, decision-making and
performance evaluation (Nicholson et al. 2012; Punt et al. 2016).
We found that most extinction-related indicators we assessed can act as leading
indicators of global extinction. Five were also leading indicators of species
distribution change (Table 1) but changing abundance, the first state in the process,
has no leading indicators. Predicting abundance change may require demographic
indicators (e.g. mortality or growth rates) not currently in wide use (Tittensor et al.
2014), or pressure-based indicators (Kristensen 2004). With so many leading
indicators, capacity already exists to improve proactive decision-making within the
CBD (CBD 1992), and could be facilitated by the Biodiversity Indicators Partnership
(BIP). Our approach complements existing policy frameworks like DPSIR by making
This article is protected by copyright. All rights reserved.
17
the relationships between indicators, and their intended use within the different
stages in the policy cycle, explicit (Kristensen 2004; IPBES 2016). While classifying
indicators can increase clarity of decision-making, adopting specific performance
measures (limits of acceptable change) for each indicator would further enhance
their utility.
Our approach shows existing biodiversity indicators can be re-framed for use in
active decision-making. International non-government organisations (INGOs) and
funding bodies can use international targets to form proactive strategic objectives
(e.g. preventing species richness decline), using leading indicators to identify spatial
priorities, develop strategic programmes and allocate funding (Collen & Nicholson
2014). Coincident and lagging indicators of richness and extinction (e.g. the BII)
could be used to evaluate return on investment and alignment of outcomes with
objectives (Hoffmann et al. 2015). National or sub-national governments or
managers could use indicators of local abundance and extinction risk as leading
indicators to prioritise species or ecosystems for intervention, with outcomes
reviewed against lagging indicators of population size, reintroductions and range
(Brazill-Boast 2018).
In the marine realm, reconciling new ecosystem-based fisheries management
approaches with appropriate performance measures is an active area of research
(Moffitt et al. 2016). Our approach provides a way to structure multiple indicators of
ecosystem health along a gradient of change (Table 2), in turn facilitating the
development of performance measures. Considering direct fishing effects as leading
indicators supports development of early warning reference points tied to
preventative action, for example informing the ‘harvest control rules’ under
This article is protected by copyright. All rights reserved.
18
management strategy evaluation (Punt et al. 2016). Coincident and lagging
indicators support outcome measurement, and are potential subjects for the
establishment of ecosystem-based target and limit reference points (Hall & Mainprize
2004).
The conceptual models we describe here are designed to be broadly applicable,
allowing flexibility while enhancing comparability between systems. Our approach is
deliberately general, using elements shared by most managed complex systems. In
principal, this generality provides a method for users (e.g. governments, managers) to
develop their own conceptual models and classification, specific to their own systems,
processes, and scales. Alternately, the same conceptual model (including our case
studies) could be applied across different scales (i.e. global and national), or systems
(i.e. different countries or ecosystems). Shared conceptual models facilitate
comparability between systems by identifying their state, even if the indicators used in
each system or scale are different (Shin et al. 2010). Thus different indicators (e.g.
leading indicators in one ecosystem might be landings, discards in another) can act in
broadly comparable units (both warning of later changes to ecosystem structure and
function), supporting aggregation for global targets (Tittensor et al. 2014). Similarly,
using the same model also allows for substitution of scale-relevant indicators when
upscaling or downscaling, and the use of either model-based or empirically derived
indicators (Guerra et al. 2019).
A key assumption of our approach is that the theoretical sequence in the conceptual
model accurately reflects progression of indicator change in the real world. Lags in
data publication, indicator calculation, and model-based, pressure-as-proxy
This article is protected by copyright. All rights reserved.
19
biodiversity indicators have the potential to alter the speed and order in which
indicators flag change. An understanding of an indicator’s ability to detect trends in
variables of interest, as well as the impacts of common issues such as data lag and
bias is vital, however many remain untested (Fulton et al. 2005; Nicholson et al.
2012). Economists rely on statistical methods with long time-series to confirm the
sequence of change (Marcellino 2006). Methods typically include multi-state models
that identify correlations between different indicators, probabilities and timeframes of
transition between states (Hougaard 1999). Testing the order of signals from
biodiversity indicators with empirical or simulated data would enhance the application
of this model and improve certainty in decision-making. .
Developing or applying conceptual models as described here can address identified
gaps in global target setting, indicator development, data collection and testing
(Collen & Nicholson 2014). The power of the economic model lies in the uniting
objective to maintain desirable economic growth and clarity regarding what
constitutes an acceptable outcome. This key foundation is missing from the
surveillance approach taken to monitoring biodiversity. By linking explicit monitoring
objectives, target development and indicator selection, our approach actively
supports the development of biodiversity objectives that are measurable and time-
bound (‘SMART’) (Green et al. 2019). Further, assigning a purpose to the indicators
establishes an expectation against which indicator performance can be tested,
adding further justification for data collection. Without appropriate data and
validation, the accuracy of using indicators for this, or indeed other purposes, cannot
be guaranteed (Collen & Nicholson 2014).
This article is protected by copyright. All rights reserved.
20
Conclusion
The CBD Aichi Targets will be revised in 2020, along with many Sustainable
Development Goals (SDGs). This is the ideal time to ensure the indicators
associated with these targets are fit-for-purpose (Mace et al. 2018; Nicholson et al.
2018; Green et al. 2019), and have specific roles in supporting decisions as leading,
coincident and lagging indicators. Our approach can be applied to different systems
and scales, as demonstrated through the two case studies. It can also support
comparability between systems and scales that require different indicators,
particularly when performance measures are implemented. Defining the process of
biodiversity change over time, with specific targets of management and indicator
functions, provides a useful structure for facilitating a much-needed proactive,
purposeful approach to biodiversity monitoring and policy (Collen & Nicholson 2014).
Glossary
Coincident indicator
Indicator considered to best measure change in the target variable in real time.
Species distribution
The observed, inferred or projected geographic range of a species (Johnson 1998;
Bubb et al. 2009).
Indicator
Metrics that quantify changes in different types of variation present in complex
systems (Bland et al. 2018).
This article is protected by copyright. All rights reserved.
21
Lagging indicator
Indicators that measure aspects of the system that only change after the coincident
indicator.
Leading indicator
Indicators that measure aspects of the system that change before the coincident
indicator.
Limit reference points
Fisheries term - the indicator value at which stock stress may occur, indicates the
system has entered an undesirable state (fisheries analogue of transition point)
(Caddy & Mahon 1995).
Performance measure
Difference between the actual indicator value, and the reference point, intended to
show the performance of management and the system against the desired state
(Hall & Mainprize 2004).
Reference point
Indicator value benchmarks that, when reached, should activate predetermined
management attention and ideally, actions (Caddy & Mahon 1995).
Target reference points
Fisheries term - the desired value of an indicator where stock levels are balanced
with optimum fishery yield (Caddy & Mahon 1995).
This article is protected by copyright. All rights reserved.
22
Target variable
An aspect of a complex system that is the subject of management objectives,
monitoring and prediction (Marcellino 2006).
Threshold reference points
Fisheries term - value of an indicator that when reached, is considered to be an
‘early warning’ that the system may be in danger of exceeding target or limit
reference points (Hall & Mainprize 2004).
Transition point
Indicator value or range representing a benchmark that delineates between a
desirable and undesirable state of a system.
This article is protected by copyright. All rights reserved.
23
References cited
Allnutt T, et al. 2008. A method for quantifying biodiversity loss and its application to a 50-
year record of deforestation across Madagascar. Conservation Letters 1:173 - 181.
Babcock RC, Shears NT, Alcala AC, Barrett NS, Edgar GJ, Lafferty KD, McClanahan TR,
Russ GR. 2010. Decadal trends in marine reserves reveal differential rates of change
in direct and indirect effects. Proceedings of the National Academy of Sciences of the
United States 107:18256 - 18261.
Balmford A, Green RE, Jenkins M. 2003. Measuring the changing state of nature. Trends in
Ecology & Evolution 18:326-330.
Bland LM, Rowland JA, Linn M, Nicholson E, Regan TJ, Keith DA, Murray NJ, Lester RE,
Rodríguez JP. 2018. Developing a standardized definition of ecosystem collapse for
risk assessment. Frontiers in Ecology and the Environment 16:29-36.
Brazill-Boast J. 2018. Saving our Species: a cost-effective, large-scale monitoring and
evaluation program for threatened species. Monitoring Threatened Species and
Ecological Communities:225.
Bubb P, Butchart S, Collen B, Dublin H, Kapos V, Pollock C, Stuart S, Vié J-C. 2009. IUCN
Red List Index: Guidance for national and regional use.
Burns AF, Mitchell WC 1946. Measuring business cycles. NBER Studies in Business Cycles
no.2 New York.
Caddy JF, Mahon R 1995. Reference points for fisheries management. Food and Agriculture
Organization of the United Nations Rome.
CBD. 1992. Convention on Biological Diversity in CBD, editor. CBD.
This article is protected by copyright. All rights reserved.
24
CBD. 2010. Strategic Plan for Biodiversity 2011 - 2020 and the Aichi Targets. Montreal,
Canada.
CBD. 2016. Decision XIII/28 Indicators for the Strategic Plan for Biodiversity 2011-2020 and
the Aichi Biodiversity Targets in Convention on Biological Diversity, editor.
Convention on Biological Diversity, Cancun, Mexico.
Ceballos G, Ehrlich PR. 2002. Mammal population losses and the extinction crisis. Science
296:904 - 907.
Coll M, et al. 2016. Ecological indicators to capture the effects of fishing on biodiversity and
conservation status of marine ecosystems. Ecological Indicators 60:947-962.
Collen B, Loh J, Whitmee S, McRae L, Amin R, Baillie JEM. 2009. Monitoring Change in
Vertebrate Abundance: the Living Planet Index. Conservation Biology 23:317-327.
Collen B, Nicholson E. 2014. Taking the measure of change. Science 346:166-167.
Done T, Reichelt R. 1998. Integrated coastal zone and fisheries ecosystem management:
generic goals and performance indices. Ecological Applications 8:S110 - S118.
Duelli P, Obrist MK. 2003. Biodiversity indicators: the choice of values and measures.
Agriculture, ecosystems & environment 98:87-98.
Fulton EA, Smith ADM, Punt AE. 2005. Which ecological indicators can robustly detect
effects of fishing? ICES Journal of Marine Science 62:540 - 551.
Gaston KJ, Blackburn TM, Greenwood JJD, Gregory RD, Quinn RM, Lawton JH. 2000.
Abundance-occupancy relationships. Journal of Applied Ecology 37:39-59.
Green EJ, Buchanan GM, Butchart SHM, Chandler GM, Burgess ND, Hill SLL, Gregory RD.
2019. Relating characteristics of global biodiversity targets to reported progress.
Conservation Biology:1360.
This article is protected by copyright. All rights reserved.
25
Gregory RD, Willis SG, Jiguet F, Voříšek P, Klvaňová A, van Strien A, Huntley B, Collingham
YC, Couvet D, Green RE. 2009. An Indicator of the Impact of Climatic Change on
European Bird Populations. PLoS ONE 4:1-6.
Guerra CA, et al. 2019. Finding the essential: Improving conservation monitoring across
scales. Global Ecology and Conservation 18:e00601.
Hall SJ, Mainprize B. 2004. Towards ecosystem‐based fisheries management. Fish and
Fisheries 5:1-20.
Hoffmann M, Duckworth JW, Holmes K, Mallon DP, Rodrigues ASL, Stuart SN. 2015. The
difference conservation makes to extinction risk of the world's ungulates.
Conservation Biology 29:1303-1313.
Hougaard P. 1999. Multi-state models: a review. Lifetime Data Analysis 5:239-264.
IPBES. 2016. Summary for policymakers of the assessment report of the methodological
assessment of scenarios and models of biodiversity and ecosystem services. Bonn,
Germany.
Johnson CN. 1998. Species extinction and the relationship between distribution and
abundance. Nature 394:272 - 274.
Jones JPG, et al. 2011. The why, what, and how of global biodiversity indicators beyond the
2010 target. Conservation Biology: The Journal Of The Society For Conservation
Biology 25:450-457.
Jørgensen C, et al. 2007. Managing Evolving Fish Stocks. Science 318:1247-1248.
Joseph LN, Field SA, Wilcox C, Possingham HP. 2006. Presence–absence versus
abundance data for monitoring threatened species. Conservation biology 20:1679-
1687.
This article is protected by copyright. All rights reserved.
26
Keith DA, et al. 2013. Scientific foundations for an IUCN Red List of ecosystems. PLoS ONE
8:e62111-e62111.
Koopmans TC. 1947. Measurement without theory. The Review of Economics and Statistics
29:161-172.
Kristensen P. 2004. The DPSIR framework. National Environmental Research Institute,
Denmark 10.
Larkin PA. 1996. Concepts and issues in marine ecosystem management. Reviews in Fish
Biology and Fisheries 6:139-164.
Leadley PW, et al. 2014. Progress towards the Aichi Biodiversity Targets: An Assessment of
Biodiversity Trends, Policy Scenarios and Key Actions. Technical Series 78.
Montreal, Canada.
Mace GM, Baillie JEM. 2007. The 2010 Biodiversity Indicators: Challenges for Science and
Policy. Conservation Biology 21:1406 - 1413.
Mace GM, Barrett M, Burgess ND, Cornell SE, Freeman R, Grooten M, Purvis A. 2018.
Aiming higher to bend the curve of biodiversity loss. Nature Sustainability 1:448 -
451.
Marcellino M. 2006. Leading indicators. Handbook of Economic Forecasting 1:879-960.
Moffitt EA, Punt AE, Holsman K, Aydin KY, Ianelli JN, Ortiz I. 2016. Moving towards
ecosystem-based fisheries management: Options for parameterizing multi-species
biological reference points. Deep Sea Research (Part II, Topical Studies in
Oceanography) 134:350 - 359.
Newbold T, et al. 2016. Has land use pushed terrestrial biodiversity beyond the planetary
boundary? A global assessment. Science 353:288 - 291.
This article is protected by copyright. All rights reserved.
27
Nicholson E, et al. 2012. Making robust policy decisions using global biodiversity indicators.
PLoS ONE 7:e41128.
Nicholson E, Fulton EA, Brooks TM, Blanchard R, Leadley P, Metzger JP, Mokany K,
Stevenson S, Wintle BA, Woolley SN. 2018. Scenarios and Models to Support Global
Conservation Targets. Trends in Ecology & Evolution 34:57 - 68.
Noss RF. 1990. Indicators for monitoring biodiversity - a heirarchical approach Conservation
Biology 4:355-364.
Pedersen EJ, Thompson PL, Ball RA, Fortin M-J, Gouhier TC, Link H, Moritz C, Nenzen H,
Stanley RR, Taranu ZE. 2017. Signatures of the collapse and incipient recovery of an
overexploited marine ecosystem. Royal Society open science 4:170215.
Pereira HM, et al. 2010. Scenarios for global biodiversity in the 21st century. Science
330:1496 - 1501.
Pereira HM, Navarro LM, Martins IS. 2012. Global Biodiversity Change: The Bad, the Good,
and the Unknown. Annual Review of Environment and Resources 37:25.
Punt AE, Butterworth DS, de Moor CL, De Oliveira JA, Haddon M. 2016. Management
strategy evaluation: best practices. Fish and Fisheries 17:303-334.
Sheehan DK, Gregory RD, Eaton MA, Bubb PJ, Chenery AM. 2010. The Wild Bird Index –
Guidance for National and Regional Use.
Shin YJ, et al. 2010. Using indicators for evaluating, comparing, and communicating the
ecological status of exploited marine ecosystems. 2. Setting the scene. ICES Journal
of Marine Science 67:692-716.
Simon HA. 1962. The architecture of complexity. Proceedings of the American Philosophical
Society 106:457-476.
This article is protected by copyright. All rights reserved.
28
Soroka SN, Stecula DA, Wlezien C. 2015. It's (Change in) the (Future) Economy, Stupid:
Economic Indicators, the Media, and Public Opinion. American Journal of Political
Science:457.
Sparks TH, et al. 2011. Linked indicator sets for addressing biodiversity loss. ORYX 45:411-
419.
Stock JH, Watson MW. 2003. How did leading indicator forecasts perform during the 2001
recession? Economic Quarterly - Federal Reserve Bank of Richmond 89:71-90.
Tittensor DP, et al. 2014. A mid-term analysis of progress toward international biodiversity
targets. Science 346:241-244.
UNGA. 2015. Resolution 70/1 - Transforming Our World: the 2030 Agenda for Sustainable
Development in Assembly UNG, editor. United Nations General Assembly, New
York.
Walpole M, et al. 2009. Tracking Progress Toward the 2010 Biodiversity Target and Beyond.
Science 325:1503 - 1504.
Westoby M, Walker B, Noy-Meir I. 1989. Opportunistic management for rangelands not at
equilibrium. Journal of Range Management:266-274.
Wu J. 1999. Hierarchy and scaling: extrapolating information along a scaling ladder.
Canadian journal of remote sensing 25:367-380.
Zarnowitz V, Boschan C. 1975. New composite indexes of coincident and lagging indicators.
Business Conditions Digest 20.
This article is protected by copyright. All rights reserved.
29
Tables
Table 1
Table 1: Global indicators of biodiversity and their classification for each of the three changing states in the
process for global species extinction. Leading classifications have no shading, coincident indicators have light
grey shading and lagging indicators are shaded dark grey.
Indicator
Type of variation measured
Indicator relationship to:
Abundance
Distribution
Extinction
Indicators of State 1: Change in species abundance
Climatic Impact
Index for birds
Change in abundance of birds
affected by climate change
(Gregory et al. 2009).
coincident
leading
leading
Living Planet
Index
Change in relative abundance
of vertebrate populations
(Collen et al. 2009)
coincident
leading
leading
Biodiversity
Intactness Index
(abundance)
Change in proportion of
original species abundance
remaining in a location
(Newbold et al. 2016).
coincident
leading
leading
Red List Index
(past
abundance)
Change in species’ Red List
status under abundance
criterion* (Bubb et al. 2009)
coincident
leading
leading
Wild Bird Index
Change in relative abundance
of a group of bird species
(Sheehan et al. 2010)
coincident
leading
leading
Indicators of State 2: Change in species distribution
Biodiversity
Habitat Index
Change in gamma diversity,
driven by changes to species
distributions. (Allnutt et al.
2008)
lagging
coincident
leading
Biodiversity
Intactness Index
(richness)
Change in proportion of
original species richness
remaining in a location
(Newbold et al. 2016)
lagging
coincident
leading
Red List Index
(past distribution)
Change in species’ Red List
status under distribution
criterion* (Bubb et al. 2009)
lagging
coincident
leading
Wildlife Picture
Index
Change in species
presence/absence in a
location (O'Brien et al. 2010)
lagging
coincident
leading
Indicators of State 3: Global extinction
Number of
species
extinctions (birds
& mammals)
Change in number of species
that have been listed as
extinct
lagging
lagging
coincident
* Hypothetical disaggregation. In reality, the RLI mixes abundance and distribution
criteria, and thus, states of change (Bubb et al. 2009)
This article is protected by copyright. All rights reserved.
30
Table 2
Table 2: Ecosystem-scale indicators of marine ecosystems, and their classification for each of the
four stages of marine ecosystem collapse. Leading classifications have no shading, coincident
indicators have light grey shading and lagging indicators are shaded dark grey.
Indicator
Type of variation
measured
Indicator relationship to:
Direct fishing effects
Ecosystem
structure
& function
Ecosystem
collapse
Indicators of State 1: Direct effects of fishing
Proportion of
discards in fishery
Change in the
proportion of total
catch that is discarded
(Coll et al. 2016).
coincident
leading
leading
1/
(landings/biomass)
Changes in the overall
volume of fish caught
(Shin et al. 2010).
coincident
leading
leading
Total biomass of
surveyed species
Change in total weight
of ecosystem
biodiversity measured
via surveys (Shin et al.
2010).
coincident
leading
leading
Catch-based
marine trophic
index
Change in trophic
composition of species
caught (Coll et al.
2016)
coincident
leading
leading
Proportion non-
declining exploited
species
Changes in number of
fished species with
declining biomass
trends (Coll et al.
2016)
coincident
leading
leading
Proportion of
under and
moderately
exploited stocks
Changes in assessed
status of stocks (Shin
et al. 2010)
coincident
leading
leading
Mean trophic level
of the landed
catch
Change in the trophic
composition of species
caught (Shin et al.
2010).
coincident
leading
leading
Mean intrinsic
vulnerability index
of the fish landed
catch
Change in the
vulnerability (based on
life history parameters)
of species caught (Coll
et al. 2016).
coincident
leading
leading
Indicators of State 2: Changes to ecosystem structure and function
Mean fish length in
the surveyed
community
Change in mean length
of fish remaining in the
ecosystem measured
via surveys (Shin et al.
2010).
lagging
coincident
leading
This article is protected by copyright. All rights reserved.
31
Indicator
Type of variation
measured
Indicator relationship to:
Direct fishing effects
Ecosystem
structure
& function
Ecosystem
collapse
Proportion of
predatory fish in
the surveyed
community
Change in proportion
of predatory fish in
ecosystem (relative to
other fish) (Shin et al.
2010).
lagging
coincident
leading
Mean trophic level
of the surveyed
community
Changes in trophic
composition of the
ecosystem measured
via surveys (Coll et al.
2016).
lagging
coincident
leading
Mean trophic level
of the modelled
community
Changes in the
modelled trophic
composition of the
ecosystem (Coll et al.
2016).
lagging
coincident
leading
1/coefficient of
variation of total
biomass of
surveyed species
Change in variability of
ecosystem biomass
over time, driven by
fishing strategy (Shin
et al. 2010).
lagging
coincident
leading
Mean maximum
life span of
surveyed fish
species
Change in mean
lifespan of ecosystem
biodiversity, driven by
fishing strategy (Shin
et al. 2010).
lagging
coincident
leading
Indicators of State 3: Ecosystem collapse
Ecosystem
collapsed
Ecosystem status
changes to ‘collapsed’
following an RLE
assessment (Keith et
al. 2013).
lagging
lagging
coincident
This article is protected by copyright. All rights reserved.
32
Figures
Figure 1
Figure 1: Examples of how indicators of different functions might change over time for three examples: Economics (a),
species extinctions (b), and marine ecosystems (c). In each panel, dotted lines represent the leading indicator, solid lines
represent the coincident indicator and dashed lines represent the lagging indicator. Shaded areas in the economics
example (a) represent recessions. Species extinction time-series (c) were simulated to provide an illustrative example
(following a similar approach to (Joseph et al. 2006)). Time-series for the economics (a) and marine leading and
coincident (c) examples were drawn from published data (Soroka et al. 2015; Pedersen et al. 2017). We added indicative
ecosystem collapse to (c), where we assume ecosystem collapse to be the loss of characteristic features (species or
community structure).
This article is protected by copyright. All rights reserved.
33
Table legends
Table 1: Global indicators of biodiversity and their classification for each of the three
changing states in the process for global species extinction. Leading classifications
have no shading, coincident indicators have light grey shading and lagging indicators
are shaded dark grey.
Table 2: Ecosystem-scale indicators of marine ecosystems, and their classification
for each of the four stages of marine ecosystem collapse. Leading classifications
have no shading, coincident indicators have light grey shading and lagging indicators
are shaded dark grey.
Figure legends
Figure 1: Examples of how indicators of different functions might change over time for three
examples: Economics (a), species extinctions (b), and marine ecosystems (c). In each panel, dotted
lines represent the leading indicator, solid lines represent the coincident indicator and dashed lines
represent the lagging indicator. Shaded areas in the economics example (a) represent recessions.
Species extinction time-series (c) were simulated to provide an illustrative example (following a
similar approach to (Joseph et al. 2006)). Time-series for the economics (a) and marine leading and
coincident (c) examples were drawn from published data (Soroka et al. 2015; Pedersen et al. 2017).
We added indicative ecosystem collapse to (c), where we assume ecosystem collapse to be the loss of
characteristic features (species or community structure).
A preview of this full-text is provided by Wiley.
Content available from Conservation Biology
This content is subject to copyright. Terms and conditions apply.