Jukuri, open repository of the Natural Resources Institute Finland (Luke) 
   
 
   
All material supplied via Jukuri is protected by copyright and other intellectual property rights. Duplication 
or sale, in electronic or print form, of any part of the repository collections is prohibited. Making electronic 
or print copies of the material is permitted only for your own personal use or for educational purposes.  For 
other purposes, this article may be used in accordance with the publisher’s terms. There may be 
differences between this version and the publisher’s version. You are advised to cite the publisher’s 
version. 
 
This is an electronic reprint of the original article.  
This reprint may differ from the original in pagination and typographic detail. 
 
Author(s): Sara Fraixedas, Andreas Lindén, Markus Piha, Mar Cabeza, Richard Gregory & Aleksi 
Lehikoinen 
Title: A state-of-the-art review on birds as indicators of biodiversity: Advances, 
challenges, and future directions 
Year: 2020 
Version: Published version 
Copyright:   The Author(s) 2020 
Rights: CC BY-NC-ND 4.0 
Rights url: http://creativecommons.org/licenses/by-nc-nd/4.0/ 
 
Please cite the original version: 
Fraixedas S., Lindén A., Piha M., Cabeza M., Gregory R., Lehikoinen A. (2020). A state-of-the-art 
review on birds as indicators of biodiversity: Advances, challenges, and future directions. 
Ecological Indicators 118, 106728. https://doi.org/10.1016/j.ecolind.2020.106728. 
Contents lists available at ScienceDirect
Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind
Review
A state-of-the-art review on birds as indicators of biodiversity: Advances,
challenges, and future directions
Sara Fraixedasa,b,⁎, Andreas Lindénc,1, Markus Pihad, Mar Cabezaa,b, Richard Gregorye,f,
Aleksi Lehikoinend,g
aHelsinki Institute of Sustainability Science (HELSUS), Faculty of Biological and Environmental Sciences, P.O. Box 4, FI-00014 University of Helsinki, Finland
bGlobal Change and Conservation Lab, Faculty of Biological and Environmental Sciences, University of Helsinki, P.O. Box 65, FI-00014 Helsinki, Finland
cNovia University of Applied Sciences, FI-10600 Ekenäs, Finland
d Finnish Museum of Natural History, University of Helsinki, FI-00014 Helsinki, Finland
e RSPB Centre for Conservation Science, The Lodge, Potton Road, Sandy, Bedfordshire SG19 2DL, UK
f Centre for Biodiversity & Environment Research, Department of Genetics, Evolution and Environment, University College London, Darwin Building, Gower Street, London
WC1E 6BT, UK
g The Helsinki Lab of Ornithology (HelLO), Finnish Museum of Natural History, University of Helsinki, FI-00014 Helsinki, Finland
A R T I C L E I N F O
Keywords:
Biodiversity targets
Birds
Multi-species indicators
Site and species selection methodology
Uncertainty measures
Environmental monitoring
A B S T R A C T
The current loss of biodiversity has been broadly acknowledged as the main cause of ecosystem change. To halt
this trend, several international agreements have been made, and various biodiversity metrics have been de-
veloped to evaluate whether the targets of these agreements are being met. The process of developing good
indicators is not trivial. Indicators should be able to synthesize and communicate our current knowledge, but
they also need to meet both scientific and practical criteria. Since it would not be practical to monitor all species,
indicators are mainly built on the monitoring of some well-known taxa, such as birds. Here we systematically
review the wide spectrum of bird biodiversity indicators (hereafter indicators) available to: i) evaluate recent
methodological advances; ii) identify current knowledge gaps jeopardizing indicator interpretation and use in
guiding decision-making; and iii) examine challenges in their applicability across different spatial and temporal
contexts. We pay particular attention to indicator characteristics such as site and species selection, spatial,
seasonal and habitat coverage, and statistical issues in developing indicators and tools to tackle them, to provide
specific recommendations for the future construction of indicators. Several methodological advances have re-
cently been made to enhance the process of indicator development, including multiple ways to select sites and
species to increase their robustness. However, we found that there are strong spatial, seasonal and habitat biases
among the selected indicators. Most of them are from Europe, using mainly census data from the breeding season
and typically covering farmland and forest habitats. The major challenges that we detected in their applicability
were related to the modelling of the statistical uncertainty associated to the indicator. We recommend the use of
quantitative methods in site and species selection procedures whenever possible. Current indicators should be
expanded to areas outside Europe and to less studied habitats and should not neglect monitoring work outside
the breeding season. Time-series analyses studying temporal trends and using multi-species data should in
general account for temporal autocorrelation as well as for phylogenetic correlation. Multi-species hierarchical
models are a good alternative for analysing and constructing indicators, but they need to include annual random
effects allowing for unexplained annual variation in the average status of the community, i.e. the indicator
target. Despite methodological and context-specific differences in the indicators reviewed, most of them seem to
highlight the urgent need of devising strategic climate and conservation policies to improve the status and trends
of biodiversity.
https://doi.org/10.1016/j.ecolind.2020.106728
Received 13 April 2020; Received in revised form 8 July 2020; Accepted 13 July 2020
⁎ Corresponding author at: Helsinki Institute of Sustainability Science (HELSUS), Faculty of Biological and Environmental Sciences, P.O. Box 4, FI-00014 University
of Helsinki, Finland.
E-mail address: sara.fraixedas@helsinki.fi (S. Fraixedas).
1 Current address: Natural Resources Institute Finland (Luke), FI-00790 Helsinki, Finland.
Ecological Indicators 118 (2020) 106728
1470-160X/ © 2020 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license 
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
T
1. Introduction
The current loss of biodiversity has been broadly acknowledged as
the main cause of ecosystem change (Hooper et al., 2012; Ceballos
et al., 2015) and it has been suggested to undermine the ability of
mankind to adapt to global change (Cardinale et al., 2012), accelerating
ecosystem degradation (Hooper et al., 2012) and threatening human
well-being (Díaz et al., 2006; Hanski et al., 2012). In 2002, signatory
states agreed under the Convention on Biological Diversity (CBD) to
tackle both human-induced extinction of species and the loss of natural
habitats through a significant reduction of the current rate of biodi-
versity loss by 2010 (CBD, 2002; BirdLife International, 2013). A re-
lated target was set by European high-level delegations, who committed
to halt biodiversity decline by 2010 (Pereira and Cooper, 2006; Mace
and Baillie, 2007; van Strien et al., 2009). However, none of the targets
were met (Butchart et al., 2010; Rands et al., 2010), and similar, but
more explicit, goals were adopted within the CBD’s Strategic Plan for
Biodiversity 2011–2020 (CBD, 2010; Henle et al., 2013; Secretariat of
the Convention on Biological Diversity, 2014), widely referred to as the
Aichi Biodiversity Targets. So far, progress towards these goals has been
poor, mainly because of the lack of a clear definition and operational
indicators (Tittensor et al., 2014). To determine whether new targets
formulated within the post-2020 Global Biodiversity Framework are
accomplished, there is a need to develop summary statistics to describe
trends in components of biodiversity as accurately and robustly as
possible, while being at the same time useful and understandable for
decision-makers (Reid et al., 1993; Gregory et al., 2005; Mace et al.,
2018; Green et al., 2019).
Birds are a good candidate taxon for monitoring global environ-
mental change, because they have long been monitored worldwide
(Bibby, 1999; Pereira and Cooper, 2006; Schmeller et al., 2012). They
are relatively easy to detect and identify, census methods are well de-
veloped and cheap, and many skilled volunteers are willing to count
birds (Koskimies, 1989; Venier and Pearce, 2004; Gregory et al., 2005;
Pereira and Cooper, 2006). Our understanding of their population
biology, behaviour and life history (except in the tropics; Xiao et al.,
2017) is good (Venier and Pearce, 2004; Gregory et al., 2005) and they
are known to show predictable population responses to environmental
change (Järvinen and Väisänen, 1979). They are also widespread and
relatively diverse (~10,000 species globally) and tend to be at, or near,
the top of the food chain, making them sensitive to changes at lower
trophic levels (Koskimies, 1989; Gregory et al., 2005).
The need to measure biodiversity change has promoted the devel-
opment of what are known as composite or multi-species indicators
(Gregory et al., 2005; Maes and Van Dyck, 2005). The purpose of
composite indicators is to integrate data of species and ecosystem
change in such a way that information is reduced into simple mean-
ingful, visual summaries or “indices” (Gregory et al., 2003; Pereira and
Cooper, 2006; Fraixedas et al., 2015a,b). Some examples are the Bio-
diversity Intactness Index (BII; Scholes and Biggs, 2005) and the fol-
lowing associated Biodiversity Intactness Variance (BIV; Hui et al.,
2008), the Red List Index (RLI; Butchart et al., 2004, 2005, 2007), the
Living Planet Index (LPI; Loh et al., 2005), or the more recent Biodi-
versity Change Index (BCI; Normander et al., 2012).
Composite trend indicators are one example where a group of spe-
cies population trends are taken together to describe environmental
changes. The main goal is to reflect the population response, but also to
mirror that of other species reacting to the same human disturbance,
therefore acting as a surrogate for ecosystem health (Caro and
D’Oherty, 1999; Gregory et al., 2005). Probably the best-known ex-
ample among birds is the Wild Bird Index (WBI; Gregory et al., 1999,
2003), which is one of the better summary statistics available at a re-
gional scale (Butchart et al., 2010; Tittensor et al., 2014). WBIs have
been adopted by the EU and incorporated in the ‘abundance and dis-
tribution of selected species’, one of the Streamlining European Biodi-
versity Indicators (SEBI) set to address the EU biodiversity targets (EEA,
2012). This indicator (SEBI 01) reflects the general state of common
birds in two main habitats (forest and farmland), which together re-
present the predominant land types in Europe (Gregory et al., 2005;
EEA, 2012; Vallecillo et al., 2016). Importantly, these indicators are
used to inform and improve current management practices, as is the
case of the European Farmland Bird Indicator (EFBI; Butler et al.,
2010). WBIs have also been used in global biodiversity syntheses
(Butchart et al., 2010; Secretariat of the Convention on Biological
Diversity, 2014; Tittensor et al., 2014).
Composite indicators have been further developed to describe the
impacts of climate change on biodiversity. Examples, also known as
pressure indicators, include the Community Temperature Index (CTI;
Devictor et al., 2008, 2012) and the Climate Impact Indicator (CII;
Gregory et al., 2009; Stephens et al., 2016). The latter, based on bird
data, has been also taken up by the EU (SEBI 011), under the name
‘impacts of climate change on bird populations’ (EEA, 2012), and used
in global biodiversity assessments (Butchart et al., 2010; Tittensor et al.,
2014).
Given the degree of integration of bird indicators into national and
international policies (Gregory and van Strien, 2010), the present study
represents a systematic review of recent scientific literature published
on multi-species indicators (i.e. multi-species indices or multi-species
annual summaries of a community reflecting changes in space and/or
time) including bird data at the global, regional and national levels. We
review four different types of indicators (state, benefit, response and
pressure; BirdLife International, 2013), specifically focusing on site and
species selection methodologies, spatial, seasonal and habitat coverage,
and statistical issues in developing indicators and tools to tackle them.
For this purpose, we summarise the general characteristics of the se-
lected indicators by evaluating recent methodological advances, iden-
tifying current knowledge gaps and recognizing the main challenges in
their applicability. We believe that a review such as this one is long
overdue, given the proliferation of different indicators, their incon-
sistent use and applications, and all the knowledge gaps that jeopardize
their interpretation across multiple spatial and temporal scales. Based
on this analysis, we formulate specific recommendations for the future
construction of indicators, including different ways to account for un-
certainty.
2. Materials and methods
We used the ISI Web of Science database as a literature search
method to access research articles. We carried out an advanced search
using two different key word combinations: i) (birds AND indicators*
AND biodiversity*); and ii) (birds AND multi-species* AND indices*).
We also discarded those studies written in a language other than
English. We looked for articles published in the first half of the im-
plementing period of the CBD’s Strategic Plan for Biodiversity
2011–2020 (i.e. 2011–2016), including studies up to December 2016.
Although we did not consider grey literature (i.e. reports, government
documents, etc.), we regard the obtained sample as representative of
the current scientific literature available on the subject of study.
Our combined search resulted in a total of 278 studies (269 from the
first key word combination, and 9 studies from the second one; searches
performed on 13 December 2019). We reviewed the title and abstract of
the articles, and chose the studies providing at least one multi-species
indicator for birds. Articles involving bird indicators together with in-
dicators for other taxa were likewise selected, but we excluded studies
where birds were combined together with other taxa in a single in-
dicator. All the bird-related indicators found in the Biodiversity
Indicators Partnership (BIP) website were covered in our literature
search, therefore including state, benefit, response and pressure in-
dicators (BirdLife International, 2013).
The final selection comprised a total of 47 articles that met our
criteria and were examined in more detail. The majority of the studies
reported more than one indicator, and we gathered information
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
2
separately for each of them. Indicators were collected both from the
main text and the supplementary material, obtaining a total of 254
indicators. The geographical extent of the final indicator set ranged
from a minimum spatial resolution of a region (e.g. southern Finland),
excluding indicators performed at the local scale (e.g. city forests/
woodlands), to a global level. We did not consider any indicators for
which there was no information available in relation to their outcome
(i.e. trend). For each of the indicators, we extracted: 1) the reference of
the article (authors and journal abbreviation); 2) type of indicator
(state, benefit, response or pressure, or a combination of these); 3) a
brief description of the indicator; 4) type of data used (e.g. abundance,
presence/absence data); 5) study period; 6) number of species involved;
7) season of the year; 8) study area; 9) type of habitat; 10) driver (type
of driver thought to affect the indicator); 11) whether site selection was
applied to develop the indicator; 12) whether species selection was
applied to develop the indicator; 13) whether expert opinion was used
in the species selection procedure; and 14) the general outcome of the
indicator (increase, decline or no trend; the latter including well de-
fined non-linear changes without a clear overall increase/decrease over
the full study period) (see Table A1).
In addition, we classified the uncertainty measures related to each
of the selected indicators, as well as other relevant data (see Table A2)
as follows:
1. Inferential approach for detecting temporal changes in indicators:
1a) Annual indicator values and their uncertainties are compared
against a reference year (e.g. the first year) or period (e.g. the average
of the first five years) with a fixed value, typically one or 100 for ease of
communication (Gregory and van Strien, 2010);
1b) Trends in indicator time-series are analysed with a separate
model (e.g. linear model, piecewise trend, generalized additive model);
this approach, which may include spatial replicates, also applies if in-
dices are first fitted with a reference year or the indicator is illustrated
using approach 1a, but after that temporal changes in the indicator are
analysed statistically in a separate model;
1c) Hierarchical approach, jointly modelling the raw data of sepa-
rate species and changes in the indicator (e.g. generalized linear mixed
model, dynamic factor analysis, Bayesian approaches). This is a similar
approach to 1b) in that a trend is modelled over the whole study period
(it also applies even if e.g. figures are made with a reference year), but
the analysis is done in only one step and the same model provides the
indicator values (the annual multi-species model fit). Even cases ig-
noring one or many levels of potentially relevant random variation are
listed here;
1d) Statistical uncertainty of the indicator or its changes is not as-
sessed at all. Indicator values are presented as naked point estimates
(e.g. RLI, where there is no statistical uncertainty given for the cate-
gories per se, although bootstrap is used to assess uncertainty due to
missing data).
2. Technical approach for estimating indicator uncertainty, or for
analysing temporal/spatial changes (variation) in the indicator:
2a) Simulation or bootstrap;
2b) Frequentist theory (e.g. null hypothesis tests or confidence in-
tervals – CIs, not obtained through simulation);
2c) Bayesian analysis.
3. Other relevant data:
3a) Accounts for species detectability when constructing the in-
dicator; indicators where species detectability is explicitly modelled;
3b) Accounts for (or tests for) phylogenetic correlation when con-
structing the indicator; especially with multi-species data, such as the
case of the indicators, it can be important to recognise that related
species’ indices may behave similarly;
3c) Accounts for (or tests for) temporal residual autocorrelation in
trend analyses of indicators; in this case autocorrelation only applies to
1b and 1c (not 1a). Even when present in the indicator, autocorrelation
is not an issue in pairwise comparisons of two years, and approach 1a
consists of several such comparisons (between the reference year and
all other values).
For all assessed traits or classification outcomes we reported the
proportions of indicators (%) that filled the criteria (binary outcome).
Using the function binom.test in R (package “stats”; R Core Team, 2018),
we obtained the 95% CIs as uncertainties associated with the reported
proportions. This was done under the assumption that the selected in-
dicators form a random sample, representing the approach preferred by
researchers in these types of studies, and that the proportion of in-
dicators is an estimate of the probability for applying a certain ap-
proach. Each of the indicators reviewed was considered as a single data
point in the proportion. We used the habitats classification scheme of
the International Union for Conservation of Nature and Natural Re-
sources to classify certain habitat types into broader categories (IUCN,
2020).
3. Results
3.1. Indicator typology and drivers
From a total of 254 indicators, we found that the minority were
benefit (n= 2) and response (n= 4) indicators, whereas almost half of
the cases (CI = 39%–52%; n = 116 of 254) were state indicators. In
addition, 30% of the selected indicators were classified into more than
one typology (e.g. pressure and state; CI = 24%–36%; n = 76 of 254).
The number of pressure indicators was almost half of state indicators
(CI = 39%–58%; n = 56 of 116). While the majority of pressure in-
dicators were either related to climate change or to land-use changes
(CI = 74%–94%; n= 48 of 56), combined pressure and state (n= 14),
and also benefit indicators were all linked to climate and land-use
changes. Pressure and response indicators showed a good balance
among the three different driver combinations (8 climate-change in-
dicators, 11 land-use change indicators, and 14 climate and land-use
change indicators). Combined state and response (n = 29), and re-
sponse indicators alone only referred to land-use changes. The majority
of state indicators were also linked to land-use changes
(CI = 83%–95%; n= 104 of 116). Therefore, up to 70% of the selected
indicators (CI = 64%–75%; n = 177 of 254) dealt with land-use
changes alone, while the percentage of indicators concerning both
drivers was less than 20% (CI = 14%–24%; n = 47 of 254) (see Table
A1 and Fig. 1).
3.2. Seasonal, spatial and habitat coverage
Regarding the season of the year, more than 80% (CI = 79%–89%;
n = 214 of 254) of the indicators used only breeding season data,
whereas only four indicators relied on wintering data (2%;
CI = 0%–4%; n = 4 of 254). Less than 15% of the indicators were
combining both breeding and wintering data (CI = 9%–18%; n= 33 of
254) (Table A1).
At least 5% of the indicators were carried out at the global scale
(CI = 3%–9%; n = 13 of 254). The majority of the selected indicators
were exclusively focused on Europe (83%; CI = 78%–87%; n = 211 of
254), and many of them were performed at the national level (75%;
CI = 69%–81%; n = 159 of 211). Europe was the only continent
covering regional indicators (n= 19). All indicators from Oceania were
performed at the national level (n= 18), whereas in America almost all
indicators were classified as supranational (i.e. involving more than one
country; 91%; CI = 59%–100%; n = 10 of 11). Only one indicator
came from Africa (see Table A1 and Fig. 2) and none from Asia.
Around 20% of the selected indicators incorporated two or more
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
3
than two habitat types (CI = 16%–26%; n= 52 of 254). A considerable
percentage of the indicators did not include information about the
habitat type (21%; CI = 16%–27%; n = 54 of 254). Farmland was the
most recurrent habitat in our dataset among those indicators including
a single habitat type (n = 50), followed closely by forest indicators
(n = 41). Forest and farmland habitats made up almost 40% of the
indicators (CI = 30%–42%; n = 91 of 254). The worst represented
habitats were marine and shrubland (n = 4, and n = 3, respectively).
Indicators with no clear habitat classification were mostly pressure
indicators (65%; CI = 51%–77%; n = 35 of 54). This was also the case
for marine indicators. Half or more than half of farmland, forest and
wetland indicators were classified as state indicators (farmland: 50%;
CI = 36%–64%; n = 25 of 50; forest: 76%; CI = 60%–88%; n = 31 of
41; wetland: 82%; CI = 57%–96%; n = 14 of 17). All indicators be-
longing to grassland and shrubland habitats were also state indicators
(see Table A1 and Table 1).
Land-use change indicators were dominant among farmland (86%),
forest (73%), generalist (i.e. no marked habitat preferences; 67%),
grassland (75%), marine (75%) and wetland habitats (88%). They also
constituted more than half of the indicators for which no habitat data
were available (CI = 41%–69%; n = 30 of 54) (see Table A1 and
Table 2).
3.3. Site selection, species selection and expert advice
The majority of the indicators used species selection (85%;
CI = 80%–89%; n = 215 of 254), and nearly 50% of them used site
selection (CI = 43%–56%; n= 125 of 254). Almost all indicators using
site selection also used species selection (96%; CI = 91%–99%;
n= 120 of 125). A bit more than 30% of the indicators included expert
advice (CI = 26%–38%; n = 81 of 254), and all of them made use of
species selection. Indicators using both site selection and expert advice
were not very common in our dataset (22%; CI = 17%–27%; n= 55 of
254), but they all included a certain species selection procedure. Almost
half of the indicators including species selection methods were state
indicators (47%; CI = 41%–54%; n= 102 of 215). It was the same case
for those indicators incorporating site, species selection and expert
advice (49%; CI = 35%–63%; n = 27 of 55) (Table A1).
The application of site selection was not independent from species
selection, meaning that there was an overrepresentation of indicators
using both site selection and species selection, or neither one, as com-
pared to indicators using only one selection type (Pearson’s Chi-squared
test: χ2 = 22.72, df = 1, p < 0.0001). Species selection was not in-
dependent from expert advice (χ2 = 19.87, df = 1, p < 0.0001), with
more indicators relying on either species selection methods or both
species selection and expert advice.
3.4. Outcome of the indicators
In general terms, half of the selected indicators showed declining
trends (CI = 44%–56%; n = 127 of 254). From these, more than 60%
were pressure and state indicators (CI = 53%–71%; n = 79 of 127).
Only about 20% of the indicators increased (CI = 18%–29%; n= 59 of
254), and more than 30% of them were state indicators
(CI = 22%–47%; n = 20 of 59). Around 20% of the indicators had no
significant trends (CI = 18%–28%; n = 58 of 254), being the majority
state indicators (78%; CI = 65%–87%; n= 45 of 58). Almost 5% of the
chosen indicators could not be classified into any of the defined cate-
gories (i.e. decline, increase or no trend; CI = 2%–7%; n = 10 of 254),
and eight of out ten were state indicators (see Table A1 and Fig. 3).
Regarding indicator typology, all benefit indicators (n = 2) de-
clined. More than half of pressure (CI = 50%–77%; n = 36 of 56), and
pressure and state indicators showed declining trends (n = 8 of 14). At
least 75% of response (n= 3 of 4) and pressure and response indicators
declined (CI = 58%–89%; n = 25 of 33). State indicators had ap-
proximately the same number of indicators declining and showing no
clear trends (n = 43 of 116 and n = 45 of 116, respectively). Contrary
to other indicator types, a bit more than half of state and response in-
dicators showed increasing trends (n = 15 of 29) (see Table A1 and
Fig. 3).
More than half of farmland indicators and forest indicators showed
declining trends (farmland: n = 26 of 50; forest: n = 21 of 41).
Together, they summed up almost 40% of the all set of declining in-
dicators (CI = 29%–46%; n = 47 of 127). More than 50% of generalist
indicators had increasing trends (n = 5 of 9). However, farmland and
forest indicators made up almost 50% of the all set of increasing in-
dicators (n = 28 of 59). More than half of grassland, shrubland and
wetland indicators showed no clear trends (grassland: n = 6 of 8;
shrubland: n = 2 of 3; wetland = 9 of 17). Approximately 40% of the
all set of indicators showing no clear trends was found among farmland
and wetland indicators (CI = 24%–50%; n = 21 of 58) (see Table A1
and Table 3).
Almost 60% of the indicators showing declining trends were related
to land-use changes (CI = 48%–66%; n = 73 of 127). A similar case
was found for indicators with increasing and no clear trends, being the
majority land-use change indicators (increase: 68%; CI = 54%–79%;
n = 40 of 59; no trend: 93%; CI = 83%–98%; n = 54 of 58). While
Fig. 1. Indicator typology as defined by BirdLife International (2013) extracted
from the literature search and classified into different drivers (UNS = Un-
specified; CC = Climate Change; LUC = Land-Use Change).
Fig. 2. Spatial coverage of the indicators extracted from the literature search
and classified into different continents. The word “supranational” refers to
those indicators involving more than one country.
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
4
70% of climate change indicators and those including both climate and
land-use changes were declining (CC: CI = 50%–86%; n= 19 of 27; CC
&LUC: CI = 55%–83%; n = 33 of 47), only about 40% of land-use
change indicators showed declining trends (CI = 34%–49%; n = 73 of
177) (Table A1).
3.5. Indicator uncertainty
In more than 80% of the cases it was possible to clearly determine
the inferential approach used to detect temporal changes in indicators
(82%; CI = 77%–86%; n = 208 of 254). The most used inferential
approach was 1b, i.e. the indicator time-series were first constructed
but their trends were analysed with a separate model (29%;
CI = 23%–36%; n = 60 of 208). This included also cases where the
trend model was hierarchical, e.g. a mixed model analysing the
indicators and including a random effect of site. Only a few of these
indicators effectively applied a combination of approaches 1a (i.e.
particular years compared against a reference year or period) and 1b
(13%; CI = 9%–18%; n= 27 of 208). About 20% of the indicators used
hierarchical modelling for constructing and analysing the indicators
(approach 1c; CI = 14%–25%; n = 40 of 208). About 60% of cases
where mixed effects models were applied, covering cases from both
approach 1b and 1c, annual random variation in the indicator was ig-
nored (e.g. no random effect of factor variable year in addition to the
trend; CI = 48%–68%; n = 58 of 100). In almost 30% of the cases
statistical uncertainty was not assessed at all (approach 1d;
CI = 22%–35%; n = 59 of 208). In addition, there were special cases
(11%; CI = 8%–16%; n = 29 of 254) where indicators were classified
as using approach 1a. These referred to simulated future scenarios not
based on statistical models. There was missing information for more
than 5% of the indicators (CI = 4%–10%; n = 17 of 254) (Table A2).
Frequentist theory was the most popular technical approach for
estimating indicator uncertainty, or for analysing temporal/spatial
changes (variation) in the indicator (approach 2b; 43%;
CI = 37%–50%; n = 110 of 254), followed closely by simulation or
bootstrap (approach 2a; 34%; CI = 28%–40%; n = 87 of 254).
Surprisingly, none of the indicators used Bayesian analysis (i.e.
Table 1
Indicator typology as defined by BirdLife International (2013) extracted from the literature search and classified into different habitat types. In some cases, the same
indicator covered two or more habitat types. We used the habitats classification scheme of the International Union for Conservation of Nature and Natural Resources
to classify certain habitat types into broader categories (IUCN, 2020). “UNS” refers to unspecified habitat.
Typology
Habitat Benefit Pressure Pressure & response Pressure & state Response State State & response Total %
>2 habitats 0 5 4 0 2 11 0 22 9
2 habitats 0 2 14 0 0 2 12 30 12
Farmland 0 3 7 3 0 25 12 50 20
Forest 0 3 4 3 0 31 0 41 16
Generalist 0 1 0 2 0 1 5 9 4
Grassland 0 0 0 0 0 8 0 8 3
Marine (*) 0 3 0 0 0 1 0 4 2
Shrubland 0 0 0 0 0 3 0 3 1
Urban 0 1 0 3 0 4 0 8 3
Wetland (**) 0 3 0 0 0 14 0 17 7
Other (***) 0 0 2 1 0 5 0 8 3
UNS 2 35 2 2 2 11 0 54 21
Total 2 56 33 14 4 116 29 254 100
% 1 22 13 6 2 46 11 100
(*) Including dune/shore, estuaries and sea
(**) Including bog/marsh, freshwater, inland water, lakes and mires
(***) Including artificial, natural, near human and open-water environments
Table 2
Indicator drivers extracted from the literature search and classified into dif-
ferent habitat types. In some cases, the same indicator covered two or more
habitat types. We used the habitats classification scheme of the International
Union for Conservation of Nature and Natural Resources to classify certain
habitat types into broader categories (IUCN, 2020). The meaning of the acro-
nyms is the following: UNS= Unspecified, CC = Climate Change, LUC= Land-
Use Change.
Driver
Habitat CC LUC CC & LUC UNS Total %
>2 habitats 7 13 2 0 22 9
2 habitats 2 19 9 0 30 12
Farmland 1 43 6 0 50 20
Forest 6 30 5 0 41 16
Generalist 1 6 2 0 9 4
Grassland 0 6 2 0 8 3
Marine (*) 1 3 0 0 4 2
Shrubland 0 1 2 0 3 1
Urban 1 4 3 0 8 3
Wetland (**) 0 15 2 0 17 7
Other (***) 0 7 1 0 8 3
UNS 8 30 13 3 54 21
Total 27 177 47 3 254 100
% 11 70 19 1 100
(*) Including dune/shore, estuaries and sea
(**) Including bog/marsh, freshwater, inland water, lakes and mires
(***) Including artificial, natural, near human and open-water environments
Fig. 3. Indicator typology as defined by BirdLife International (2013) extracted
from the literature search and classified into different outcomes. “UNS” refers to
unspecified outcome.
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
5
approach 2c), and in more than 20% of the cases it was not possible to
evaluate the technical approach given that uncertainty was not assessed
(i.e. indicators previously classified as 1d; CI = 18%–29%; n = 59 of
254) (Table A2).
In addition, only 15% of the indicators accounted for species de-
tectability (approach 3a; CI = 11%–20%; n= 38 of 254), about 1% for
phylogenetic correlation (approach 3b; CI = 0%–3%; n = 2 of 254),
and less than 20% for temporal autocorrelation, only considering in-
dicators classified as either 1b or 1c (approach 3c; CI = 10%–26%;
n = 17 of 100) (Table A2). The modelling of temporal autocorrelation
was not relevant when the inferential approach was 1a (see Methods
section). Therefore, in the majority of cases indicators did not in-
corporate any of these features.
4. Discussion
Three main research findings emerge from this study. First, im-
portant methodological advances have been made to improve the ro-
bustness of the reviewed indicators, as illustrated by the extensive use
of site and/or species selection methods in their construction. Second,
strong spatial, seasonal and habitat biases have been identified, with
most indicators being habitat-specific (mainly covering farmland and
forest habitats), mostly focused on Europe and based on data from the
breeding season. Third, statistical uncertainty was ignored in a sub-
stantial number of the indicators reviewed. We elaborate on these
findings and their implications in the subsequent sections.
4.1. Recent methodological advances
Our review shows a notable number of indicators for which site
selection, species selection or a combination of both methods are ap-
plied, particularly in the case of state indicators. Species selection
procedures are key elements when designing indicators. Several studies
have demonstrated that indices are more representative of the wider
community when species selection criteria are used to identify the
potential indicator species pool (Butler et al., 2012; Wade et al., 2014).
A good example are studies including methods to quantify species
habitat preferences based on habitat data derived from transects
(Larsen et al., 2011; Jiguet et al., 2012a; Renwick et al., 2012; Fraixedas
et al., 2015b). Other recent methodological advances in species selec-
tion include the construction of indicators using a ratio of species po-
sitively and negatively affected by a driver, which enables a finer in-
terpretation of the patterns affecting the shape of the final indicators
(Herrando et al., 2014, 2016; Palmer et al., 2015; Stephens et al.,
2016). However, even in this case, it is necessary to select species po-
sitively and negatively affected by the specific driver. Given that species
can behave differently within these two groups, it is essential to rely on
either evidence-based selection or, if that is not possible, use expert
opinion to complement species selection procedures (Roth et al., 2014),
or in other cases to simply assign species to a particular habitat
(Galewski and Devictor, 2016).
4.2. Current knowledge gaps
We discovered several biases in the indicators reviewed. Most of the
indicators analysing drivers of population change only considered land-
use changes. The combination of climate and land-use changes was not
very common in our dataset, though their interaction is critical in un-
derstanding environmental change, as it is to conservation policies
(Burns et al., 2016; Oliver et al., 2017). There was also a bias in terms of
indicator typology, with the number of state indicators clearly sur-
passing other indicator types such as benefit or response. Moreover,
since many bird species are migratory, population drivers may influ-
ence species during different phases of their life cycle. The selected
indicators are mostly based on data from the breeding season because
territorial bird populations can often be censused more easily, and in-
dicators using data from the non-breeding season are scarce (but see
e.g. Godet et al., 2011; Fraixedas et al., 2015a). Another important gap
found at the global scale is the spatial coverage of the indicators, since
they are mostly focused on Europe, where most of national indicators
come from. We specifically found a lack of indicators from Africa and
Asia. Out of all published state indicators found in the peer-reviewed
literature, the most common are habitat-specific ones, which are biased
towards farmland and forests. Although these are the most widespread
habitats in Europe, others such as marine or shrubland habitats should
not be neglected. Despite the impacts caused by human activities on
marine ecosystems and species, the current lack of information on the
status of species and habitats limits our understanding of the pressures
that may affect biodiversity and ecosystems (EEA, 2017). However, it is
known that seabirds are among the most threatened groups of birds
(Hoffmann et al., 2018). As for shrubland habitats, most bird popula-
tions have been severely declining since the 1950 s (e.g. in eastern
North America; Hunter et al., 2001), which highlights the need to
measure the effectiveness of conservation programs for these species
(Schlossberg and King, 2015).
4.3. Statistical issues in indicator development
The results showed that statistical uncertainty was ignored in sev-
eral of the indicators reviewed. Realistic measures of uncertainty ac-
companying the indicator and its trends are essential for reliable in-
ference and sensible interpretation of changes in any indicator (Soldaat
et al., 2017). Similarly as for error estimation, the number of indicators
accounting for species detectability, phylogenetic correlation and
temporal autocorrelation was very low. Variation in detectability be-
tween species may not represent a large problem when species included
in the indicator analyses are given an equal weight (Lehikoinen, 2013).
However, changes in detectability over time due to, for instance phe-
nology, may generate a bias in the estimated abundance (Lehikoinen,
2013; Chambert et al., 2015). Moreover, few studies have considered
the importance of spatial variation in species detectability when esti-
mating temporal trends in bird populations (e.g. MacLeod et al., 2012;
Massimino et al., 2015). Temporal autocorrelation is another key
Table 3
Indicator outcome (increase, decline or no trend; the latter referring to well
defined non-linear changes without a clear overall increase/decrease over the
full study period) extracted from the literature search and classified into dif-
ferent habitat types. In some cases, the same indicator covered two or more
habitat types. We used the habitats classification scheme of the International
Union for Conservation of Nature and Natural Resources to classify certain
habitat types into broader categories (IUCN, 2020). “UNS” refers to unspecified
outcome/habitat.
Outcome
Habitat Decline Increase No trend UNS Total %
>2 habitats 9 4 5 4 22 9
2 habitats 16 9 5 0 30 12
Farmland 26 12 12 0 50 20
Forest 21 16 4 0 41 16
Generalist 3 5 1 0 9 4
Grassland 2 0 6 0 8 3
Marine (*) 0 1 1 2 4 2
Shrubland 1 0 2 0 3 1
Urban 2 3 3 0 8 3
Wetland (**) 7 1 9 0 17 7
Other (***) 2 2 0 4 8 3
UNS 38 6 10 0 54 21
Total 127 59 58 10 254 100
% 50 23 23 4 100
(*) Including dune/shore, estuaries and sea.
(**) Including bog/marsh, freshwater, inland water, lakes and mires.
(***) Including artificial, natural, near human and open-water environments.
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
6
component of estimating population change (Thomas, 1996). In-
corporating a temporally correlated component in the trend analysis
can smooth out fluctuations caused by e.g. an anomalous weather year
(Harrison et al., 2014), and ignoring it may leave lead to under-
estimated uncertainly (e.g. too narrow CIs), which is difficult to spot
and may actually look like a desired result. The same can happen if
phylogeny is ignored, since it is assumed that species are independent
from each other even if they may share a common ancestry (Johnston
et al., 2014; Sólymos et al., 2018). To account for phylogenetic corre-
lation one could, for instance, weight species according to uniqueness
(i.e. a weight that is different for each species), or add several random
effects (e.g. family, genus and species) on the intercept and slope when
modelling trends (Ives and Helmus, 2011; Tucker et al., 2017).
A separate, and often missing, component of many indicator ana-
lyses with multiple species or sites is the modelling of common annual
random variation, i.e. between-species synchrony affecting the in-
dicator or spatial synchrony when analysing variation in the indicator.
This can be simply modelled by including year as a factor variable with
random effects (on the intercept) in the model, and it is particularly
important both in log-linear trend models and in models allowing for
flexible and more complex trends, such as generalized additive models
(Knape, 2016). Although researchers may be reluctant about including
year as both a fixed effect (as a continuous variable – describing the
trend) and as a random effect in the same model, these variables have
very different function and the model is in most cases fully identifiable.
4.4. Major challenges in indicator applicability
There are inevitable challenges concerning the use and development
of biodiversity indicators in Europe and beyond (Walpole et al., 2009).
One potential explanation for studies not including statistical un-
certainty may be that current methods for error estimation are poorly
developed in multi-species biodiversity indices (Gregory et al., 2019).
Lack of data also constitutes a limiting factor when trying to estimate
error in trends, leaving some options out of range (e.g. bootstrapping
species*sites data to fully account for sampling error; Buckland et al.,
2005). Even when data are available, there are certain technical chal-
lenges and limitations when reporting statistical uncertainty (e.g. con-
structing CIs around indicators and their trends; Soldaat et al., 2017).
As an example, there is currently no statistical uncertainty given for the
red list categories, which makes difficult to communicate the un-
certainty associated to the RLI values and trends (Butchart et al., 2007;
Rueda-Cediel et al., 2018). In addition, the ability of statistical models
to react to recent changes in indicator trends depends on the length of
the time series, among other aspects (Tittensor et al., 2014). This is a
major limiting factor to all existing indicators looking at trends, since
large-scale land-use changes, and therefore their impacts on vertebrate
diversity, have already taken place before the commonly used baselines
(e.g. the 1970s; Hoffmann et al., 2018; Fraixedas et al., 2019). More-
over, even in the case of WBIs, where bird observations are collected
through systematic annual breeding surveys to provide scientifically
robust and representative trend information (Gregory and van Strien,
2010; Hoffmann et al., 2018), data coverage is currently limited to
Europe and North America (but see for instance Wotton et al., 2017).
Another problem associated with current monitoring schemes is that
they may not be able to monitor populations of rarer species, poten-
tially excluding the more sensitive species (Battisti and Fanelli, 2016).
The omission of less common species associated with pristine habitats
can produce an over-optimistic assessment of the health of the eco-
system (Renwick et al., 2012). At the same time, accounting for de-
tectability will be only possible if we choose monitoring methods that
provide data, where requirements for analysing observation process are
fulfilled (e.g. Kéry and Schmidt, 2004, 2008). Finally, problems of
species selection persist (e.g. differences in the allocation of species into
groups based on their habitat preference depending on the region stu-
died), which may strongly affect the trends in indices (Fraser et al.,
2017).
4.5. Recommendations and future directions
Several methodological advances have been made in the develop-
ment of biodiversity indicators. Quantitative methods for species and
site selection will likely improve the robustness of existing indicators
(Gregory and van Strien, 2010; Wade et al., 2013, 2014), and we
strongly advise paying attention to species selection when establishing
a new indicator. If expert opinion is required, methods for engaging
with experts that enhance the accuracy and calibration of their judge-
ments should be used (Sutherland and Burgman, 2015; Fraser et al.,
2017). Currently, there are strong spatial, habitat and seasonal biases in
peer-reviewed indicators. Therefore, we call for further studies to fill
the gaps, particularly in ecosystems outside Europe (e.g. Szabo et al.,
2012; Ingram et al., 2015), and without neglecting the non-breeding
season (e.g. Godet et al., 2011; Fraixedas et al., 2015a). There is also a
need to integrate the impacts of climate change and land use on bio-
diversity into the same modelling framework (e.g. Clavero et al., 2011;
Eglington and Pearce-Higgins, 2012; Ay et al., 2014), which may prove
especially valuable for the assessment of conservation policies. Several
of the revised indicators relied on an initial baseline year to measure
trends. Poor estimation in the baseline year due to e.g. low sample size
or large annual fluctuations in the index usually leads to inaccurate
estimates of the population trend, with biodiversity declines being
overlooked (Buckland and Johnston, 2017). However, there are
methods to reduce the sensitivity to the choice of baseline year, such as
smoothing the index using generalized additive models (Fewster et al.,
2000; Buckland et al., 2005). It is also possible to fit models with an-
nually separate intercepts, so that all years are associated with their
own uncertainty, rather that integrating the uncertainty of the baseline
year(s) to be compared with all other years.
Multi-species hierarchical models are a good alternative for con-
structing and analysing indicators (Amano et al., 2012), but they need
to include annual random effects allowing for annual variation in the
indicators (Knape, 2016). Another attractive but seldom seen modelling
approach that could be used for simultaneously constructing and ana-
lysing indicators is dynamical factor analysis (Zuur et al., 2003). Time-
series analyses studying temporal trends and using multi-species data
should in general account for temporal autocorrelation (e.g. Lindström
et al., 2013; Galewski and Devictor, 2016; Schipper et al., 2016). In
addition, phylogenetic correlation is an issue worth to consider in
modelling decisions (e.g. Jiguet et al., 2012b). At the outset, further
efforts should be made to improve monitoring design to obtain high
quality data on biodiversity at acceptable spatial and temporal resolu-
tions (de Heer et al., 2005; Normander et al., 2012). By increasing the
robustness of monitoring schemes, we hope to see a greater diversifi-
cation of the indicator typology, including the development of more
benefit and response indicators. Moreover, common methods to apply
and harmonise data from different monitoring schemes would be highly
desirable to optimize sampling effort and be more efficient in the sta-
tistical use of data (Normander et al., 2012; Schmeller et al., 2012).
Applying monitoring methods that enable accounting for detectability
will be most important in studies that cover a wide range of habitat
types (Studeny et al., 2013; Harrison et al., 2014). Finally, cross-vali-
dation with other biodiversity indicators is a means of improving the
interpretation of results (Bailey et al., 2007; European Communities,
2009), because birds may respond differently to environmental factors
when compared to other taxa (Gregory and van Strien, 2010).
5. Concluding remarks
To effectively estimate changes in biodiversity and evaluate pro-
gress towards global biodiversity targets in the context of the post-2020
Global Biodiversity Framework, greater commitment and increased
resources are needed to expand current monitoring schemes in order to
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
7
improve the spatial representativeness and the habitat and seasonal
coverage of the indicators. This will be key to improve our under-
standing of how ecosystems and the living resources react to environ-
mental changes. Furthermore, to obtain a clear picture of biodiversity
trends and being able to communicate the results to policy makers,
researchers must find appropriate statistical methods to calculate in-
formative and robust indicators with a realistic level of fluctuations and
uncertainty around them. This will ensure the formulation of well-in-
formed conservation programmes and a resource for practitioners to
mitigate or even reverse current biodiversity losses. Our results re-
garding the outcome of the revised indicators give cause for concern,
since most of them were more often displaying declining trends in
biodiversity than increasing ones. Parties to the CBD prepare to adopt
the post-2020 Global Biodiversity Framework and do so facing an
overall failure in achieving most of the Aichi Biodiversity Targets. We
believe that this review can help to improve indicator development and
application, and thereby contribute to ensure a roadmap to monitor
success of the next generation of biodiversity goals.
Author contributions
SF conceived the study, carried out the analyses and writing, ALe
and ALi supervised the project, ALi and MP helped with data collection;
all authors contributed to the study design, write and approved the final
version of the paper.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgments
Sara Fraixedas received financial support for conducting this re-
search from the Maj and Tor Nessling Foundation, the Finnish Cultural
Foundation, and the Helsinki Institute of Sustainability Science
(HELSUS), and Aleksi Lehikoinen from the Academy of Finland (grant
275606). The authors would like to thank Andrea Santangeli and
Álvaro Fernández-Llamazares for their insightful comments and sug-
gestions.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.ecolind.2020.106728.
References
Amano, T., Okamura, H., Carrizo, S.F., Sutherland, W.J., 2012. Hierarchical models for
smoothed population indices: the importance of considering variations in trends of
count data among sites. Ecol. Indic. 13, 243–252.
Ay, J.S., Chakir, R., Doyen, L., Jiguet, F., Leadley, P., 2014. Integrated models, scenarios
and dynamics of climate, land use and common birds. Clim. Change 126, 13–30.
Bailey, D., Billeter, R., Aviron, S., Schweiger, O., Herzog, F., 2007. The influence of
thematic resolution on metric selection for biodiversity monitoring in agricultural
landscapes. Landsc. Ecol. 22, 461–473.
Battisti, C., Fanelli, G., 2016. Applying indicators of disturbance from plant ecology to
vertebrates: the hemeroby of bird species. Ecol. Indic. 61, 799–805.
Bibby, C.J., 1999. Making the most of birds as environmental indicators. Ostrich 70 (1),
81–88.
BirdLife International, 2013. Spotlight on birds as indicators. Presented as part of the
BirdLife State of the world’s birds website. http://datazone.birdlife.org/home (ac-
cessed 9 November 2016).
Buckland, S.T., Johnston, A., 2017. Monitoring the biodiversity of regions: Key principles
and possible pitfalls. Biol. Conserv. 214, 23–34.
Buckland, S.T., Magurran, A.E., Green, R.E., Fewster, R.M., 2005. Monitoring change in
biodiversity through composite indices. Phil. Trans. R. Soc. B: Biol. Sci. 360,
243–254.
Burns, F., Eaton, M.A., Barlow, K.E., Beckmann, B.C., Brereton, T., Brooks, D.R., Brown,
P.M.J., Al Fulaij, N., Gent, T., Henderson, I., Noble, D.G., Parsons, M., Powney, G.D.,
Roy, H.E., Stroh, P., Walker, K., Wilkinson, J.W., Wotton, S.R., Gregory, R.D., 2016.
Agricultural management and climatic change are the major drivers of biodiversity
change in the UK. PLOS ONE 11 (3), e0151595.
Butchart, S.H.M., Stattersfield, A.J., Baillie, J., Bennun, L.A., Stuart, S.N., Akçakaya, H.R.,
Hilton-Taylor, C., Mace, G.M., 2005. Using Red List Indices to measure progress to-
wards the 2010 target and beyond. Phil. Trans. R. Soc. B 360, 255–268.
Butchart, S.H.M., Stattersfield, A.J., Bennun, L.A., Shutes, S.M., Akçakaya, H.R., Baillie,
J.E.M., Stuart, S.N., Hilton-Taylor, C., Mace, G.M., 2004. Measuring global trends in
the status of biodiversity: red list indices for birds. PLOS Biol. 2 (12), e383.
Butchart, S.H.M., Akçakaya, H.R., Chanson, J., Baillie, J.E.M., Collen, B., Quader, S.,
Turner, W.R., Amin, R., Stuart, S.N., Hilton-Taylor, C., 2007. Improvements to the
Rest List Index. PLOS ONE 2 (1), e140.
Butchart, S.H.M., Walpole, M., Collen, B., van Strien, A., Scharlemann, J.P., Almond, R.E.,
Baillie, J.E., Bomhard, B., Brown, C., Bruno, J., Carpenter, K.E., Carr, G.M., Chanson,
J., Chenery, A.M., Csirke, J., Davidson, N.C., Dentener, F., Foster, M., Galli, A.,
Galloway, J.N., Genovesi, P., Gregory, R.D., Hockings, M., Kapos, V., Lamarque, J.F.,
Leverington, F., Loh, J., McGeoch, M.A., McRae, L., Minasyan, A., Hernández
Morcillo, M., Oldfield, T.E., Pauly, D., Quader, S., Revenga, C., Sauer, J.R., Skolnik,
B., Spear, D., Stanwell-Smith, D., Stuart, S.N., Symes, A., Tierney, M., Tyrrell, T.D.,
Vié, J.C., Watson, R., 2010. Global biodiversity: indicators of recent declines. Science
328 (5982), 1164–11168.
Butler, S.J., Freckleton, R.P., Renwick, A.R., Norris, K., 2012. An objective, niche-based
approach to indicator species selection. Methods Ecol. Evol. 3 (2), 317–326.
Butler, S.J., Boccaccio, L., Gregory, R.D., Voříšek, P., Norris, K., 2010. Quantifying the
impact of land-use change to European farmland bird populations. Agri. Ecosystems
Environ. 137, 348–357.
Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P., Narwani,
A., Mace, G.M., Tilman, D., Wardle, D.A., Kinzig, A.P., Daily, G.C., Loreau, M., Grace,
J.B., Larigauderie, A., Srivastava, D.S., Naeem, S., 2012. Biodiversity loss and its
impact on humanity. Nature 486, 59–67.
Caro, T.M., D’Oherty, G., 1999. On the use of surrogate species in conservation biology.
Conserv. Biol. 13 (4), 805–814.
CBD, 2002. The Convention of Biological Diversity and the 6th Meeting of the Conference
of Parties (CoP 6): Decision VI/26 on Strategic Plan for Biodiversity 2002–2010.
accessed 27 October 2016. https://www.cbd.int/decision/cop/?id=7200.
CBD, 2010. The Convention of Biological Diversity and the 10th Meeting of the
Conference of Parties (CoP 10): Decision X/2 on Strategic Plan for Biodiversity
2011–2020. accessed 16 January 2017. https://www.cbd.int/decision/cop/?id=
12268.
Ceballos, G., Ehrlich, P.R., Barnosky, A.D., García, A., Pringle, R.M., Palmer, T.M., 2015.
Accelerated modern human-induced species losses: entering the sixth mass extinc-
tion. Sci. Adv. 1, e1400253.
Chambert, T., Kendall, W.L., Hines, J.E., Nichols, J.D., Pedrini, P., Waddle, J.H.,
Tavecchia, G., Walls, S.C., Tenan, S., 2015. Testing hypotheses on distribution shifts
and changes in phenology of imperfectly detectable species. Methods Ecol. Evol. 6,
638–647.
Clavero, M., Villero, D., Brotons, L.l., 2011. Climate change or land use dynamics: do we
know what climate change indicators indicate? PLOS ONE 6 (4), e18581.
de Heer, M., Kapos, V., ten Brick, B.J.E., 2005. Biodiversity trends in Europe: develop-
ment and testing of a species trend indicator for evaluating progress towards the 2010
target. Philos. Trans. R. Soc. B 360, 297–308.
Devictor, V., Julliard, R., Couvet, D., Jiguet, F., 2008. Birds are tracking climate warming,
but not fast enough. Proc. Roy. Soc. B. 275, 2743–2748.
Devictor, V., van Swaay, C., Brereton, T., Brotons, L., Chamberlain, D., Heliölä, J.,
Herrando, S., Julliard, R., Kuussaari, M., Lindström, Å., Reif, J., Roy, D.B., Schweiger,
O., Settele, J., Stefanescu, C., van Strien, A., van Turnhout, C., Vermouzek, Z.,
WallisDeVries, M., Wynhoff, I., Jiguet, F., 2012. Differences in the climatic debts of
birds and butterflies at a continental scale. Nat. Clim. Change 2, 121–124.
Díaz, S., Fargione, J., Chapin III, F.S., Tilman, D., 2006. Biodiversity loss threatens human
well-being. PLOS Biol. 4 (8), e277.
EEA, 2012. Streamlining European biodiversity indicators 2020: Building a future on
lessons learnt from the SEBI 2010 process. EEA Technical report No 11/2012.
Copenhagen, European Environment Agency. https://www.eea.europa.eu/ (accessed
1 December 2016).
EEA, 2017. State of marine ecosystems and associated descriptors. Report 2017.
Copenhagen, European Environment Agency. https://www.eea.europa.eu/ (accessed
20 February 2020).
Eglington, S., Pearce-Higgins, J.W., 2012. Disentangling the relative importance of
changes in climate and land-use intensity in driving recent bird population trends.
PLOS ONE 7 (3), e30407.
Communities, E., 2009. Guidance Document: The Application of the High Nature Value
Impact Indicator 2007–2013. accessed 24 January 2017. https://enrd.ec.europa.eu/
enrd-static/fms/pdf/6A6B5D2F-ADF1-0210-3AC3-AD86DFF73554.pdf.
Fewster, R.M., Buckland, S.T., Siriwardena, G.M., Baillie, S.R., Wilson, J.D., 2000.
Analysis of population trends for farmland birds using generalized additive models.
Ecology 81, 1970–1984.
Fraixedas, S., Galewski, T., Ribeiro-Lopes, S., Loh, J., Blondel, J., Fontès, H., Grillas, P.,
Lambret, P., Nicolas, D., Olivier, A., Geijzendorffer, I.R., 2019. Estimating biodi-
versity changes in the Camargue wetlands: an expert knowledge approach. PLOS ONE
14 (10), e0224235.
Fraixedas, S., Lehikoinen, A., Lindén, A., 2015a. Impacts of climate and land-use change
on wintering bird populations in Finland. J. Avian Biol. 46, 63–72.
Fraixedas, S., Lindén, A., Lehikoinen, A., 2015b. Population trends of common breeding
forest birds in southern Finland are consistent with trends in forest management and
climate change. Ornis Fenn. 92, 187–203.
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
8
Fraser, H., Pichancourt, J.-B., Butet, A., 2017. Tiny terminological disagreements with far
reaching consequences for global bird trends. Ecol. Ind. 73, 79–87.
Galewski, T., Devictor, V., 2016. When common birds became rare: historical records
shed light on long-term responses of bird communities to global change in the largest
wetland of France. PLOS ONE 11 (11), e0165542.
Godet, L., Jaffré, M., Devictor, V., 2011. Waders in winter: long-term changes of mi-
gratory bird assemblages facing climate change. Biol. Lett. 7 (5), 714–717.
Green, E.J., Buchanan, G.M., Butchart, S.H.M., Chandler, G.M., Burgess, N.D., Hill, S.L.L.,
Gregory, R.D., 2019. Relating characteristics of global biodiversity targets to reported
progress. Conserv. Biol. 33 (6), 1360–1369.
Gregory, R.D., Gibbons, D.W., Impey, A., Marchant, J.H., 1999. Generation of the
headline indicator of wild bird populations. BTO Research Report 221. British Trust
for Ornithology, Thetford, and Royal Society for the Protection of Birds, Sandy.
Gregory, R.D., Noble, D., Field, R., Marchant, J., Raven, M., Gibbons, D.W., 2003. Using
birds as indicators of biodiversity. Ornis Hungarica 12–13, 11–24.
Gregory, R.D., Skorpilova, J., Voříšek, P., Butler, S., 2019. An analysis of trends, un-
certainty and species selection shows contrasting trends of widespread forest and
farmland birds in Europe. Ecol. Ind. 103, 676–687.
Gregory, R.D., van Strien, A., 2010. Wild bird indicators: using composite population
trends of birds as measures of environmental health. Ornithol. Sci. 9 (1), 3–22.
Gregory, R.D., van Strien, A., Voříšek, P., Gmelig-Meyling, A.W., Noble, D.G., Foppen,
R.P.B., Gibbons, D.W., 2005. Developing indicators for European birds. Philos. T.
Roy. Soc. B 360, 269–288.
Gregory, R.D., Willis, S.G., Jiguet, F., Voříšek, P., Klvanová, A., van Strien, A., Huntley, B.,
Collingham, Y.C., Couvet, D., Green, R.E., 2009. An indicator of the impact of cli-
matic change on european bird populations. PLOS ONE 4 (3), e4678.
Hanski, I., von Hertzen, L., Fyhrquist, N., Koskinen, K., Torppa, K., Laatikainen, T.,
Karisola, P., Auvinen, P., Paulin, L., Mäkelä, M.J., Vartiainen, E., Kosunen, T.U.,
Alenius, H., Haahtela, T., 2012. Environmental biodiversity, human microbiota and
allergy are interrelated. Proc. Natl. Acad. Sci. USA 109 (21), 8334–8339.
Harrison, P.J., Buckland, S.T., Yuan, Y., Elston, D.A., Brewer, M.J., Johnston, A., Pearce-
Higgins, W., 2014. Assessing trends in biodiversity over space and time using the
example of British breeding birds. J. Appl. Ecol. 51, 1650–1660.
Henle, K., Bauch, B., Auliya, M., Külvik, M., Pe’er, G., Schmeller, D.S., Framstad, E., 2013.
Priorities for biodiversity monitoring in Europe. A review of supranational policies
and a novel scheme for integrative prioritization. Ecol. Indic. 33, 5–18.
Herrando, S., Anton, M., Sardà-Palomera, F., Bota, G., Gregory, R.D., Brotons, L., 2014.
Indicators of the impact of land use changes using large-scale bird surveys: land
abandonment in a Mediterranean region. Ecol. Indic. 45, 235–244.
Herrando, S., Brotons, L., Anton, M., Páramo, F., Villero, D., Titeux, N., Quesada, J.,
Stefanescu, C., 2016. Assessing impacts of land abandonment on Mediterranean
biodiversity using indicators based on bird and butterfly monitoring data. Environ.
Conserv. 43 (1), 69–78.
Hoffmann, M., Brooks, T.M., Butchart, S.H.M., Gregory, R.D., McRae, L., 2018. Trends in
biodiversity: vertebrates, in: DellaSala, D.A., Goldstein, M.I. (Eds.), The Encyclopedia
of the Anthropocene. Elsevier, Oxford, vol. 3, p. 175–184.
Hooper, D.U., Adair, E.C., Cardinale, B.J., Byrnes, J.E.K., Hungate, B.A., Matulich, K.L.,
Gonzalez, A., Duffy, J.E., Gamfeldt, L., O’Connor, M.I., 2012. A global synthesis re-
veals biodiversity loss as a major driver of ecosystem change. Nature 486, 105–108.
Hui, D., Biggs, R., Scholes, R.J., Jackson, R.B., 2008. Biol. Conserv. 141, 1091–1094.
Hunter, W.C., Buehler, D.A., Canterbury, R.A., Confer, J.L., Hamel, P.B., 2001.
Conservation of disturbance-dependent birds in eastern North America. Wildl. Soc.
Bull. 29, 440–455.
Ingram, D.J., Coad, L., Collen, B., Kümpel, N.F., Breuer, T., Fa, J.E., Gill, D.J.C., Maisels,
F., Schleicher, J., Stokes, E.J., Taylor, G., Scharlemann, J.P.W., 2015. Indicators for
wild animal offtake: methods and case study for African mammals and birds. Ecol.
Soc. 20 (3), 40.
IUCN, 2020. Habitats Classification Scheme (Version 3.1). https://www.iucnredlist.org/
(accessed 29 February 2020).
Ives, A.R., Helmus, M.R., 2011. Generalized linear mixed models for phylogenetic ana-
lyses of community structure. Ecol. Monogr. 81, 511–525.
Järvinen, O., Väisänen, R.A., 1979. Changes in bird populations as criteria of environ-
mental changes. Holarctic Ecol. 2, 75–80.
Jiguet, F., Devictor, V., Julliard, R., Couvet, D., 2012a. French citizens monitoring or-
dinary birds provide tools for conservation and ecological sciences. Acta Oecol. 44,
58–66.
Jiguet, F., Godet, L., Devictor, V., 2012b. Hunting and the fate of French breeding wa-
terbirds. Bird Stud. 59 (4), 474–482.
Johnston, A., Newson, S.E., Risely, K., Musgrove, A.J., Massimino, D., Baillie, S.R.,
Pearce-Higgins, J.W., 2014. Species traits explain variation in detectability of UK
birds. Bird Stud. 61 (3), 340–350.
Kéry, M., Schmidt, B.R., 2004. Monitoring programs need to take into account imperfect
species detectability. Basic Appl. Ecol. 5, 65–73.
Kéry, M., Schmidt, B.R., 2008. Imperfect detection and its consequences for monitoring
for conservation. Community Ecol. 9 (2), 207–2016.
Knape, J., 2016. Decomposing trends in Swedish bird populations using generalized ad-
ditive mixed models. J. Appl. Ecol. 53, 1852–1861.
Koskimies, P., 1989. Birds as a tool in environmental monitoring. Ann. Zool. Fennici 26,
153–166.
Larsen, J.L., Heldbjerg, H., Eskildsen, A., 2011. Improving national habitat specific bio-
diversity indicators using relative habitat use for common birds. Ecol. Indic. 11,
1459–1466.
Lehikoinen, A., 2013. Climate change, phenology and species detectability in monitoring
scheme. Popul. Ecol. 55, 315–323.
Lindström, Å., Green, M., Paulson, G., Smith, H.G., Devictor, V., 2013. Rapid changes in
bird community composition at multiple temporal and spatial scales in response to
recent climate change. Ecography 36, 313–322.
Loh, J., Green, R.E., Ricketts, T., Lamoreux, J., Jenkins, M., Kapos, V., Randers, J., 2005.
The Living Planet Index: using species population time series to track trends in bio-
diversity. Phil. Trans. R. Soc. B 360, 289–295.
Mace, G.M., Baillie, J.E.M., 2007. The 2010 biodiversity indicators: challenges for science
and policy. Conserv. Biol. 21 (6), 1406–1413.
Mace, G.M., Barrett, M., Burgess, N.D., Cornell, S.E., Freeman, R., Grooten, M., Purvis, A.,
2018. Aiming higher to bend the curve of biodiversity loss. Nat. Sustain. 1, 448–451.
MacLeod, C.J., Greene, T.C., MacKenzie, D.I., Allen, R.B., 2012. Monitoring widespread
and common bird species on New Zealand’s conservation lands: a pilot study. New
Zeal. J. Ecol. 36 (3), 300–311.
Maes, D., Van Dyck, H., 2005. Habitat quality and biodiversity indicator performances of
a threatened butterfly versus a multispecies group for wet heathlands in Belgium.
Biol. Conserv. 123, 177–187.
Massimino, D., Johnston, A., Noble, D.G., Pearce-Higgins, J.W., 2015. Multi-species
spatially-explicit indicators reveal spatially structured trends in bird communities.
Ecol. Indic. 58, 277–285.
Normander, B., Levin, G., Auvinen, A.P., Bratli, H., Stabbetorp, O., Hedblom, M.,
Glimskär, A., Gudmundsson, G.A., 2012. Indicator framework for measuring quantity
and quality of biodiversity – exemplified in the Nordic countries. Ecol. Indic. 13,
104–116.
Oliver, T.H., Gillings, S., Pearce-Higgins, J.W., Brereton, T., Crick, H.Q.P., Duffield, S.J.,
Morecroft, M.D., Roy, D.B., 2017. Large extents of intensive land use limit community
reorganization during climate warming. Glob. Change. Biol. 23, 2272–2283.
Palmer, G., Stephens, P.A., Ward, A.I., Willis, S.G., 2015. Nationwide trophic cascades:
changes in avian community structure driven by ungulates. Sci. Rep. 5, 15601.
Pereira, H.M., Cooper, D., 2006. Towards the global monitoring of biodiversity change.
Trends Ecol. Evol. 21 (3), 123–129.
R Core Team, 2018. R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna. Available online at: https://www.R-
project.org/ (accessed 17 January 2020).
Rands, M.R.W., Adams, W.M., Bennun, L., Butchart, S.H.M., Clements, A., Coomes, D.,
Entwistle, A., Hodge, I., Kapos, V., Scharlemann, J.P.W., Sutherland, W.J., Vira, B.,
2010. Biodiversity conservation: challenges beyond 2010. Science 329, 1298–1303.
Reid, W.V., McNeely, J.A., Tunstall, D.B., Bryant, D.A., Winograd, M., 1993. Biodiversity
Indicators for Policy-makers. WRI and IUCN, Washington, DC and Gland.
Renwick, A.R., Johnston, A., Joys, A., Newson, S.E., Noble, D.G., Pearce-Higgins, J.W.,
2012. Composite bird indicators robust to variation in species selection and habitat
specificity. Ecol. Indic. 18, 200–207.
Roth, T., Plattner, M., Amrhein, V., 2014. Plants, birds and butterflies: short-term re-
sponses of species communities to climate warming vary by taxon and with altitude.
PLOS ONE 9 (1), e82490.
Rueda-Cediel, P., Anderson, K.E., Regan, T.J., Regan, H.M., 2018. Effects of uncertainty
and variability on population declines and IUCN Red List classifications. Conserv.
Biol. 32 (4), 916–925.
Schipper, A.M., Belmaker, J., De Miranda, M.D., Navarro, L.M., Böhning-Gaese, K.,
Costello, M.J., Dornelas, M., Foppen, R., Hortal, J., Huijbregts, M.A.J., Martín-López,
B., Pettorelli, N., Queiroz, C., Rossberg, A.G., Santini, L., Schiffers, K., Steinmann,
Z.J.N., Visconti, P., Rondinini, C., Pereira, H.M., 2016. Contrasting changes in the
abundance and diversity of North American bird assemblages from 1971 to 2010.
Glob. Chang. Biol. 22, 3948–3959.
Schlossberg, S., King, D.I., 2015. Measuring the effectiveness of conservation programs
for shrubland birds. Glob. Ecol. Conserv. 4, 658–665.
Schmeller, D.S., Henle, K., Loyau, A., Besnard, A., Henry, P.Y., 2012. Bird-monitoring in
Europe – a first overview of practices, motivations and aims. Nature Conserv. 2,
41–57.
Scholes, R.J., Biggs, R., 2005. A biodiversity intactness index. Nature 434, 45–49.
Secretariat of the Convention on Biological Diversity, 2014. Global Biodiversity Outlook
4. Montréal, 155 pages.
Soldaat, L.L., Pannekoek, J., Verweij, R.J.T., van Turnhout, C.A.M., van Strien, A.J., 2017.
A Monte Carlo method to account for sampling error in multi-species indicators. Ecol.
Indic. 81, 340–347.
Sólymos, P., Matsuoka, S.M., Stralberg, D., Barker, N.K.S., Bayne, E.M., 2018. Phylogeny
and species traits predict detectability. Ecography 41, 1595–1603.
Stephens, P.A., Mason, L.R., Green, R.E., Gregory, R.D., Sauer, J.R., Alison, J., Aunins, A.,
Brotons, L., Butchart, S.H.M., Campedelli, T., Chodkiewicz, T., Chylarecki, P., Crowe,
O., Elts, J., Escandell, V., Foppen, R.P.B., Heldbjerg, H., Herrando, S., Husby, M.,
Jiguet, F., Lehikoinen, A., Lindström, Å., Noble, D.G., Paquet, J.Y., Reif, J., Sattler, T.,
Szép, T., Teufelbauer, N., Trautmann, S., van Strien, A.J., van Turnhout, C.A.M.,
Voříšek, P., Willis, S.G., 2016. Consistent response of bird populations to climate
change on two continents. Science 352 (6281), 84–87.
Studeny, A.C., Buckland, S.T., Harrison, P.J., Illian, J.B., Magurran, A.E., Newson, S.E.,
2013. Fine-tuning the assessment of large-scale temporal trends in biodiversity using
the example of British breeding birds. J. Appl. Ecol. 50, 190–198.
Sutherland, W.J., Burgman, M.A., 2015. Use experts widely. Nature 526, 317–318.
Szabo, J.K., Butchart, S.H.M., Possingham, H.P., Garnett, S.T., 2012. Adapting global
biodiversity indicators to the national scale: a Red List Index for Australian birds.
Biol. Conserv. 123, 177–187.
Thomas, L., 1996. Monitoring long-term population change: why are there so many
analysis methods? Ecology 77, 49–58.
Tittensor, D.P., Walpole, M., Hill, S.L.L., Boyce, D.G., Britten, G.L., Burgess, N.D.,
Butchart, S.H.M., Leadley, P.W., Regan, E.C., Alkemade, R., Baumung, R., Bellard, C.,
Bouwman, L., Bowles-Newark, N.J., Chenery, A.M., Cheung, W.W.L., Christensen, V.,
Cooper, H.D., Crowther, A.R., Dixon, M.J.R., Galli, A., Gaveau, V., Gregory, R.D.,
Gutierrez, N.L., Hirsch, T.L., Höft, R., Januchowski-Hartley, S.R., Karmann, M., Krug,
C.B., Leverington, F.J., Loh, J., Lojenga, R.K., Malsch, K., Marques, A., Morgan,
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
9
D.H.W., Mumby, P.J., Newbold, T., Noonan-Mooney, K., Pagad, S.N., Parks, B.C.,
Pereira, H.M., Robertson, T., Rondinini, C., Santini, L., Scharlemann, J.P.W.,
Schindler, S., Sumaila, U.R., Teh, L.S.L., van Kolck, J., Visconti, P., Ye, Y., 2014. A
mid-term analysis of progress toward international biodiversity targets. Science 346,
241–244.
Tucker, C.M., Cadotte, M.W., Carvalho, S.B., Davies, T.J., Ferrier, S., Fritz, S.A., Grenyer,
R., Helmus, M.R., Jin, L.S., Mooers, A.O., Pavoine, S., Purschke, O., Redding, D.W.,
Rosauer, D.F., Winter, M., Mazel, F., 2017. A guide to phylogenetic metrics for
conservation, community ecology and macroecology. Biol. Rev. Camb. Philos. Soc. 92
(2), 698–715.
Vallecillo, S., Maes, J., Polce, C., Lavalle, C., 2016. A habitat quality indicator for common
birds in Europe based on species distribution models. Ecol. Indic. 69, 488–499.
van Strien, A.J., van Duuren, L., Foppen, R.P.B., Soldaat, L.L., 2009. A typology of in-
dicators of biodiversity change as a tool to make better indicators. Ecol. Indic. 9,
1041–1048.
Venier, L.A., Pearce, J.L., 2004. Birds as indicators of sustainable forest management. For.
Chron. 80 (1), 61–66.
Wade, A.S.I., Barov, B., Burfield, I.J., Gregory, R.D., Norris, K., Butler, S.J., 2013.
Quantifying the detrimental impacts of land-use and management change on eur-
opean forest bird populations. PLOS ONE 8 (5), e64552.
Wade, A.S.I., Barov, B., Burfield, I.J., Gregory, R.D., Norris, K., Voříšek, P., Wu, T., Butler,
S.J., 2014. A niche-based framework to assess current monitoring of European forest
birds and guide indicator species’ selection. PLOS ONE 9 (5), e97217.
Walpole, M., Almond., R.E.A., Besançon, C., Butchart, S.H.M., Campbell-Lendrum, D.,
Carr, G.M., Collen, B., Collette, L., Davidson, N.C., Dulloo, E., Fazel, A.M., Galloway,
J.N., Gill, M., Goverse, T., Hockings, M., Leaman, D., Morgan, D.H.W., Revenga, C.,
Rickwood, C.J., Schutyser, F., Simons, S., Stattersfield, A.J., Tyrrell, T.D., Vié, J-C.,
Zimnsky, M., 2009. Tracking Progress Towards the 2010 Biodiversity Target and
Beyond. Science 325, 1503–1504.
Wotton, S.R., Eaton, M.E., Sheehan, D., Munyekenye, F., Burfield, I.J., Butchart, S.H.M.,
Gregory, R.D., 2017. Developing biodiversity indicators for African birds. Oryx 1–12.
Xiao, H., Hu, Y., Lang, Z., Fang, B., Guo, W., Zhang, Q., Pan, X., Lu, X., 2017. How much
do we know about the breeding biology of bird species in the world? J. Avian Biol. 48
(4), 513–518.
Zuur, A.F., Tuck, I.D., Bailey, N., 2003. Dynamic factor analysis to estimate common
trends in fisheries time series. Can. J. Fish. Aquat. Sci. 60, 542–552.
S. Fraixedas, et al. Ecological Indicators 118 (2020) 106728
10