Citizen science reveals trends in bat populations: The National Bat
Monitoring Programme in Great Britain
, P.A. Briggs
, K.A. Haysom
, A.M. Hutson
, N.L. Lechiara
, P.A. Racey
, A.L. Walsh
Bat Conservation Trust, Quadrant House, 250 Kennington Lane, London SE11 5RD, UK
Steve Langton (Statistical Consultancy), Hallgarth, Leavening, Malton, North Yorkshire YO17 9SA, UK
Received 8 August 2014
Received in revised form 6 November 2014
Accepted 11 November 2014
Bats play an important role in ecosystems and are highly relevant as indicators of environmental change.
Long-term monitoring of bat populations is therefore fundamental to verifying environmental change
over time. Although in the past, signiﬁcant declines in bat populations have been reported across Europe,
only limited data are available from systematic monitoring schemes over long periods. In this study we
use data from the National Bat Monitoring Programme, a citizen science scheme drawing on over 3500
volunteers, to estimate changes in populations across Great Britain for 10 bat species or species groups
between 1997 and 2012. We demonstrate uniquely how data collected on relative abundance and activ-
ity of bats by volunteers at 3272 sites using standardised, multiple survey methods (counts at roosts and
bat detector surveys using tuneable, heterodyne detectors) can be successfully utilised to produce statis-
tically robust population indices for a large proportion of a country’s bat fauna. All trends calculated, with
the exception of one species (Rhinolophus ferrumequinum), had sufﬁcient power to detect Red Alert level
declines. Our results reveal a generally favourable picture for bats over the monitoring period; all species
showed a stable or increasing trend from at least one survey type, although for four species where there
were multiple trends from different survey types, the trend directions did not agree (Myotis nattereri, Pipi-
strellus pipistrellus, Pipistrellus pygmaeus and Eptesicus serotinus). This study demonstrates that use of vol-
unteer programmes can be successful in monitoring bat populations, provided that key features including
standardised survey methods and volunteer training are incorporated. Some species that are more difﬁ-
cult to detect and identify may however require specialist surveillance techniques.
Ó2014 Elsevier Ltd. All rights reserved.
Long term monitoring of plant and animal populations is of fun-
damental importance to the effective conservation of biodiversity
at all scales. Citizen science programmes are widely used for gath-
ering information on the presence and abundance of species and
for assessing changes in species distribution and population size
in a number of taxa including birds, butterﬂies, reptiles and mam-
mals (Dickinson et al., 2010; Gregory et al., 2005; Penone et al.,
2013; Sewell et al., 2012; Wright et al., 2013). Data generated by
monitoring schemes are used to inform conservation priorities,
assess responses to anthropogenic change, determine and evaluate
management actions and produce bioindicators to assess the state
of ecosystems (Balmford et al., 2003; Jones et al., 2011; Loh et al.,
2005; Magurran et al., 2010). For example, information generated
from breeding bird surveys carried out by volunteers in the UK
has been central to the demonstration of the impacts of agricul-
tural intensiﬁcation on farmland birds (Chamberlain et al., 2000).
Bats represent around a ﬁfth of the world’s mammal species
(Simmons, 2005) and therefore contribute signiﬁcantly to the
diversity of mammals at a global level. Bats also provide essential
ecosystem services (Kunz et al., 2011), for example as pollinators
(Fujita and Tuttle, 1991), or as predators by regulating populations
of their prey including insects (Kalka et al., 2008; Boyles et al.,
2011) and have traits that make them sensitive to a wide range
0006-3207/Ó2014 Elsevier Ltd. All rights reserved.
Corresponding author. Tel.: +44 207 8207173.
E-mail addresses: email@example.com (K.E. Barlow), firstname.lastname@example.org (P.A.
Briggs), email@example.com (K.A. Haysom), firstname.lastname@example.org (A.M.
Hutson), email@example.com (P.A. Racey), firstname.lastname@example.org (A.L. Walsh),
email@example.com (S.D. Langton).
Present address: Berks, Bucks & Oxon Wildlife Trust, Hasker House, Woolley Firs,
Cherry Garden Lane, Maidenhead, Berks, SL6 3LJ, UK.
Present address: Winkﬁeld, Station Road, Plumpton Green, East Sussex BN7 3BU,
Present address: Centre for Ecology and Conservation, University of Exeter
Cornwall Campus, Treliever Road, Penryn, Cornwall TR10 9EZ, UK.
Present address: San Diego Zoo Institute for Conservation Research, 15600 San
Pasqual Valley Road, Escondido, CA 92027, USA.
Biological Conservation 182 (2015) 14–26
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/biocon
of environmental impacts including habitat loss and fragmenta-
tion, climate and pollution (Ekman and de Jong, 1996; Gerell and
Lundberg, 1993; Kunz et al., 2007; Ransome, 1989; Rebelo et al.,
2010). For all these reasons they are highly relevant as indicators
of ecosystem change (Jones et al., 2009; Newson et al., 2009;
Haysom et al., 2014).
Many bat populations experienced large declines in the second
half of the twentieth century with inﬂuential drivers including
habitat loss, climate change, disease and development (Hutson
et al., 2001; Jones et al., 2009; Mickleburgh et al., 1992;
Puechmaille et al., 2011). Information on the status and trends of
bat populations is therefore critical, not only to ensure their effec-
tive conservation and to determine how bats are responding to the
unprecedented rates of current global change, but also to provide a
wider indication of the health of ecosystems and to contribute to
the global aim of reducing rates of biodiversity loss. The value of
long-term monitoring of bat populations has been recognised at
a global (Hutson et al., 2001; Jones et al., 2009) and European level
(Battersby, 2010), and indeed monitoring bats is a national obliga-
tion within Europe under the Habitats Directive, Council Directive
92/42/EEC on the conservation of natural habitats and of wild ﬂora
and fauna, (http://ec.europa.eu/environment/nature/legislation/
habitatsdirective/index_en.htm; accessed 1 August 2014) and the
Agreement on Conservation of Populations of Migratory Bats
(http://www.eurobats.org/; accessed 1 August 2014).
Bats are challenging to monitor as they are small, nocturnal and
can present difﬁculties in species identiﬁcation. Survey methods
used for bats include visual counts at winter and summer roosts,
capture surveys and more recently with the development of bat
detectors, acoustic surveys (Battersby, 2010; Kunz and Parsons,
2009; Walters et al., 2012). There have been a number of attempts
to determine the effectiveness of different survey methods
(MacSwiney et al., 2008; Meyer et al., 2010; O’Shea and Bogan,
2003) although there are no commonly used protocols at a global
scale. There is limited evidence of the scale of the declines in bat
populations in the UK and across western Europe however
(Haysom et al., 2010; Racey and Stebbings, 1972; Stebbings,
1988), and few large-scale programmes using standardised meth-
ods for assessing population trends at a national scale. A study
investigating habitat use by bats in the UK demonstrated that vol-
unteers could be coordinated successfully to carry out bat detector
surveys at a national level (Walsh and Harris, 1996a,b; Walsh et al.,
1995). The National Bat Monitoring Programme (NBMP) was sub-
sequently established in the UK in 1996, with surveys since
1997. It is a long-term, integrated, programme monitoring multiple
species, using data collected by a network of volunteers carrying
out different survey methods at different stages in the annual cycle
of bats, with the aim of identifying population declines requiring
conservation action (Walsh et al., 2001, 2003).
Here, we use data generated by the NBMP to evaluate whether
the use of data on the relative abundance and activity of bats col-
lected through a volunteer-based programme can provide statisti-
cally robust temporal trends and detect changes in bat populations.
Monitoring programmes must be able to detect population change
to be effective and the survey effort required to detect such change
over a set timescale is an important factor to consider in design and
implementation. Alerts are commonly used to identify changes in
species populations and there are a number of such approaches
to identify serious population declines, for example the IUCN Red
List (www.iucnredlist.org; accessed 1 August 2014). A similar
example is the system of Red Alerts (a > 50% decline over 25 years)
and Amber Alerts (a 25–49% decline over 25 years) developed for
use in the analysis of UK bird population trends (Marchant et al.,
1997). Whilst there are limitations to the alert approach
(Magurran et al., 2010), it is a useful method for indicating signif-
icant population change and is utilised within the NBMP. We focus
on data collected from Great Britain (England, Scotland and Wales),
as volunteer (and therefore survey) coverage is sparse for some
surveys in other areas covered by the programme. Speciﬁcally in
this study we aim to determine: (i) whether volunteer-based sur-
veys can provide robust population indices for bats using example
data from the NBMP in GB and whether there is sufﬁcient power to
detect Red and Amber Alert level changes in populations in all sur-
veys; (ii) whether bat populations in GB have changed over the
period of monitoring; (iii) whether there are differences in esti-
mated changes between species or survey methods; and (iv) what
lessons can be learnt from the programme that are applicable to
monitoring bats at a broader scale.
2.1. Survey methods
Four survey types (Roost Count, Hibernation Survey, Field Sur-
vey and Waterway Survey, Table 1) were carried out annually by
trained volunteer surveyors (Walsh et al., 2001), of which more
than 3500 took part in the programme between 1997 and 2012.
Where possible, more than one survey type was used to monitor
each species. To ensure volunteers had the relevant skills to take
part in surveys, training resources were made available to survey-
ors including bat detector workshops, online training tutorials,
ﬁeld notes on identiﬁcation and video demonstrations (www.bat-
s.org.uk; accessed 1 August 2014). Surveys were restricted to suit-
able weather conditions (avoiding heavy rain, high winds and
sunset temperatures below 7 °C) and basic environmental data
were collected by surveyors for each survey.
Roost Counts were carried out at summer roosts of seven bat
species (greater horseshoe bat Rhinolophus ferrumequinum, lesser
horseshoe bat Rhinolophus hipposideros, Natterer’s bat Myotis nat-
tereri, common pipistrelle Pipistrellus pipistrellus, soprano pipi-
strelle Pipistrellus pygmaeus, serotine Eptesicus serotinus and
brown long-eared bat Plecotus auritus) located in buildings and
other man-made structures. Emergence counts, starting 15 min
prior to sunset or at sunset depending on species, were completed
by surveyors at self-selected roost sites on two dates prior to par-
turition between May and July. The majority of sites have been
identiﬁed to species by experienced bat workers or by volunteers
using bat detectors and following guidance provided in the survey
instructions and training resources. At sites with mixed species
colonies, the focal species was distinguished by size, behaviour,
emergence time and echolocation calls (Walsh et al., 2001). Roost
Count data have been collected since 1997 for most species except
M. nattereri, for which there were surveys since 2000.
Winter counts (Hibernation Survey) were carried out at
hibernation sites, typically caves, mines and other underground
structures such as cellars or ice houses. Data were collected
on all species encountered. Two cryptic species, whiskered bat
Myotis mystacinus and Brandt’s bat Myotis brandtii were grouped
as they are difﬁcult to separate visually in hibernation. Two
daytime visits were made by groups of surveyors to self-
selected hibernation sites, one in January and one in February
under licence from the relevant statutory nature conservation
organisation. Volunteers searched for bats in open locations
and in crevices along a standard route around the site. Hiberna-
tion Surveys have been co-ordinated at a national scale since
1997. Data were also collated from additional counts in earlier
years (1990–1996) at some sites where monitoring was ongoing
prior to the start of the NBMP. Although the survey protocol
was for two surveys between January and February, the pro-
gramme also accepted data from surveys carried out from
December to March as at some sites variations to the protocol
were already in place at the start of the co-ordinated surveys.
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 15
The Field Survey was designed to collect activity data on four
bat species (P. pipistrellus, P. pygmaeus, noctule Nyctalus noctula
and E. serotinus) along transects by using bat detectors. Trained
volunteers were allocated a 1 km grid square selected using a ran-
dom stratiﬁed sampling approach based on proportional represen-
tation of 40 UK land classes from the Land Cover Map 2000
(LCM2000) as described in Walsh et al. (2001). A transect (of
approximately 3 km in length and as close as possible to an ideal
triangular design) was mapped out by the volunteer within each
1 km square and split into 12 approximately equal sections (see
August 2014, for an example). Commencing 20 min after sunset
on two dates in July separated by at least ﬁve days, surveyors
walked each section of the transect with a heterodyne bat detector
tuned to 25 kHz. Surveyors counted the number of bat passes of N.
noctula and E. serotinus, (a pass being a sequence of two or more
echolocation calls heard as a bat ﬂies past the detector and sepa-
rated from the previous pass by at least 1 second (Fenton,
1970)). At the end of each section surveyors stopped and com-
pleted a 2-min point count with the bat detector tuned to 50 kHz
in which they counted the number of bat passes of P. pipistrellus
and P. pygmaeus. Species identiﬁcation was carried out in the ﬁeld
by assessment of call characteristics: repetition rate, rhythm, tone
and peak frequency (P. pipistrellus 42–48 kHz, P. pygma-
eus > 52 kHz, N. noctula < 21 kHz and E. serotinus 25–29 kHz) based
on known characteristics of calls (Ahlén and Baagøe, 1999; Russ,
2012). Volunteers were provided with instructions on call identiﬁ-
cation. ‘Unsure’ categories for Pipistrellus and Nyctalus/Eptesicus
were included on the survey forms to minimise errors in species
identiﬁcation by allowing surveyors the option of recording bat
passes that they hear but are unable to identify to species with cer-
tainty. Field Survey data have been collected since 1998.
The Waterway Survey was designed to collect activity data for
Daubenton’s bat Myotis daubentonii along watercourses as this spe-
cies forages primarily over the surface of the water trawling for
prey (e.g. Harris and Yalden, 2008). Trained volunteers were allo-
cated a grid reference (accurate to 100 m) along a watercourse
greater than 2 m wide for which there was River Habitat Survey
information (Raven et al., 1998). A 1 km transect was mapped
out by the volunteer along the watercourse. This transect was cen-
tralised on the allocated grid reference and ten point count
locations were evenly spaced along it (see http://www.bats.or-
g.uk/nbmp_tutorials_waterway/tutorial6.htm, accessed 1 August
2014, for an example). The survey was carried out on two dates
in August, separated by at least ﬁve days, commencing 40 min after
sunset. Surveyors completed 4-min point counts at each location
along the transect with the bat detector tuned to 35 kHz counted
the number of M. daubentonii passes or, where activity levels were
very high, record activity as continuous. Species identiﬁcation was
conﬁrmed for each pass by using a torch to observe the bats ﬂying
across the surface of the water. An ‘unsure’ category was included
on the survey form to minimise errors in species identiﬁcation by
allowing surveyors the option of recording the presence of a possi-
ble M. daubentonii pass that was not conﬁrmed visually. Waterway
Survey data have been collected since 1997, however the ‘unsure’
option on the survey form was only introduced in 1998 and there-
fore data from 1997 were excluded from this analysis.
2.2. Power analysis
Power analysis was carried out in the ﬁrst ﬁve years of the mon-
itoring programme (Walsh et al., 2001, 2003), in order to assess the
ability of each survey to detect Red and Amber Alert level popula-
tion declines for each species monitored and to determine the min-
imum number of sites required for each species and survey over
speciﬁed periods. Since this initial power analysis, the width of
the conﬁdence intervals has been monitored to assess the precision
of each trend. This approach is used in preference to repeating
power analysis for considering the biological meaning of the
observed trend (Gerard et al., 1998; Hoenig and Heisey, 2001)
and also reﬂects aspects of the design, such as the pattern of miss-
ing values, which are difﬁcult to allow for within a power analysis.
Additional power analyses were carried out here for the three
widely distributed species monitored through the Field Survey (P.
pipistrellus, P. pygmaeus and N. noctula), in order to assess the
power of the binomial approach to trend analysis, using data from
surveys completed between 1998 and 2010. A simulation approach
was used to assess the ability of the analysis method to detect
trends, with variabilities based on Restricted Maximum Likelihood
(REML) estimates of transformed data from the survey, following
the methods of Roche et al. (2011). Simulations were carried out
over a 28 year period to determine the number of years of Field
Summary of NBMP survey data contributing to bat population trend analyses.
Survey Species Total no. sites contributing
to trend analysis
Average annual no. sites
contributing to trend
Roost Count R. ferrumequinum 28 16.7 ± 1.28 1997–2012
R. hipposideros 242 126.6 ± 11.0 1997–2012
M. nattereri 72 34.8 ± 2.76 2000–2012
P. pipistrellus 442 187.2 ± 11.0 1997–2012
P. pygmaeus 333 122.3 ± 6.29 1997–2012
E. serotinus 89 35.7 ± 1.28 1997–2012
P. auritus 139 47.7 ± 5.88 1997–2012
Hibernation Survey R. ferrumequinum 204 59.5 ± 5.67 1990–2012
R. hipposideros 256 85.3 ± 10.2 1990–2012
M. daubentonii 324 149.3 ± 16.3 1990–2012
M. nattereri 448 183.4 ± 20.7 1990–2012
M. mystacinus/brandtii 189 89.3 ± 8.84 1990–2012
P. auritus 375 160.9 ± 19.1 1990–2012
Field Survey P. pipistrellus 532 171.3 ± 12.5 1998–2012
P. pygmaeus 532 181.5 ± 13.3 1998–2012
E. serotinus 394 133.4 ± 10.4 1998–2012
N. noctula 529 168.6 ± 12.4 1998–2012
Waterway Survey M. daubentonii 785 244.3 ± 22.1 1998–2012
16 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26
Survey required for 80% power to detect Red and Amber Alert level
declines for varying numbers of sites.
2.3. Population trend analysis
Population indices were calculated for each of the species for
which sufﬁcient data were available using a General Additive
Model (GAM) approach which was developed to describe popula-
tion trends in breeding birds (Fewster et al., 2000). This approach
has been used for analysis of bat populations elsewhere (Roche
et al., 2011). Only sites within GB with at least two years of survey
data were included. Analysis was completed using a log-linear gen-
eralised linear model (GLM) with a Poisson error distribution ﬁtted
to the relative abundance or bat activity data from each survey
type and species. A site term was included in the model to allow
for differences in relative abundance between sites and the time
trend was modelled using the GAM framework to ﬁt a smoothed
curve to show long-term trends over and above annual ﬂuctua-
tions. The degree of smoothing in the GAM was controlled by spec-
ifying the degrees of freedom for the smoothing process. This may
vary between one (equivalent to a simple linear trend) and one less
than the number of years (a ‘saturated model’ equivalent to ﬁtting
individual annual means) and was usually set to the default sug-
gested by Fewster et al. (2000) of 0.3 times the number of years
of survey data. For each species and survey, curves were ﬁtted with
a range of degrees of freedom and goodness-of-ﬁt statistics calcu-
lated in order to ensure that the default provided a good ﬁt to the
annual means without being unduly inﬂuenced by individual out-
lying years. To allow for the large number of sites in some models,
in some cases models were ﬁtted as a two-step process, ﬁrst ﬁtting
the saturated model and then ﬁtting the GAM with the appropriate
degrees of freedom to the annual means, appropriately weighted to
reﬂect the different numbers of sites contributing to each value.
Simulations indicated that this modiﬁcation had negligible impact
on precision but allowed the model to be ﬁtted when the number
of sites was very large. To avoid the problems of temporal autocor-
relation within abundance data (Fewster et al., 2000) and over-dis-
persion (Zuur et al., 2009), a bootstrap approach was used to
produce conﬁdence limits around the smoothed trends. At least
400 bootstrap samples were used for each model in which new
datasets were created by resampling with replacement sites from
the original dataset to ensure robust 95% conﬁdence intervals.
Smoothed GAMs are quite robust against random variation
between years, except for the ﬁrst and last year of the time series
where annual ﬂuctuations and extreme outliers can have a large
impact on trend direction and therefore may bias estimates of pop-
ulation change. The baseline year was therefore set as 1999
(index = 100) where sample sizes were sufﬁcient, and later for sur-
veys where initial sample sizes were small. For each species and
survey, where conﬁdence intervals of the trend did not overlap
the index of 100 in the last year of survey, there was a signiﬁcant
difference in the population from the base to the last year
The geographical distribution of volunteers leads to unequal
representation of countries within GB (Fig. 1). To account for this,
counts were weighted in proportion to the ratio of non-upland area
to the number of sites surveyed, thus ensuring that each country
contributes equally to the trends based on land area. Upland areas
were excluded from the weighting calculations due to the rela-
tively low bat abundance in this land class (Walsh and Harris,
1996a). For the species with ranges that are restricted to the more
southern parts of England and Wales (R. ferrumequinum, R. hippo-
sideros and E. serotinus), un-weighted results were used.
The maximum count for each year was used as the response
variable in the Poisson models for Roost Counts, or a single count
where only one survey was carried out; this approach produced
more precise results than including each individual count. Dummy
zeros were included in the model for Roost Count at sites where
the survey forms returned by the surveyors implied no bats were
present during the survey period, or where a count was made
but the date was unknown. For Hibernation Surveys, the count of
bats observed was the response variable. Where additional counts
were available for Hibernation Surveys from years prior to the start
of the NBMP in 1997, these were included in the GAMs to improve
trend estimation, although the model results reported here are
from the baseline year for each species.
Over-dispersion is a particular issue for bat activity data from
bat detector surveys, because a single bat may repeatedly ﬂy past
the surveyor giving rise to a large count of bat passes or continuous
activity. Whilst the bootstrapping process ensures that the conﬁ-
dence limits from the Poisson GAM models have appropriate cov-
erage, the precision of the ﬁtted trend curve is substantially
reduced by these high counts. Therefore, a different approach
was taken for trend analysis of Field and Waterway Survey data,
using binomial models of the proportion of point counts or transect
sections in each survey where the species was observed (excluding
GAM models can include covariates for factors that could inﬂu-
ence the means (e.g. bat detector make, temperature, survey date).
Such factors can inﬂate random variability and can cause bias
Fig. 1. Distribution of 10 km squares in GB containing survey sites from which data
have contributed to species trends.
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 17
where their average value changes over time, unless they are
included in the models. Generalised Linear Mixed Models (GLMMs)
were used to investigate factors that might inﬂuence results and
these were included in subsequent GAMs where they were statis-
tically signiﬁcant with a biologically plausible relationship. When
the Field Survey started in 1998, >80% of surveys were carried
out using the Batbox III (Batbox Ltd, West Sussex) heterodyne
bat detector. Since then, a much wider range of bat detector
models has become available. By 2012, only 10% of surveys were
completed using the Batbox III, with the most commonly used
detector (25% of surveys) being the Duet (Batbox Ltd, West Sussex).
The various models of bat detectors used in surveys have a range of
different microphone types, each with a unique frequency
response leading to differences between detectors in detectability
of different bat species (Adams et al., 2012). An analysis without
covariates may produce a biased estimate of trend due to changes
in technology being confounded with population change. To
account for this issue, Field Survey GAM models were ﬁtted with
covariates for microphone type and sensitivity range.
3.1. Survey effort
In total, from all survey types, data from 3272 sites collected
between 1997 and 2012 contributed to species trends for ten spe-
cies or species group (Fig. 1,Table 1). A total of 610 Hibernation
Survey sites had valid counts from more than one year since
1990 (with 86.5% of surveys being carried out in the core survey
period of January and February). Not all species were encountered
at all sites however, therefore the number of sites that contributed
to the Hibernation Survey trend for each species was lower than
the total. Sites with no bats in any year were excluded from the
model for that species (Table 1). Not all species recorded during
Hibernation Surveys were encountered sufﬁciently often or in
large enough numbers to produce robust trends and analysis was
carried out on Hibernation Survey data for the following
species or species group only: R. ferrumequinum, R. hipposideros,
M. nattereri, M. daubentonii,M. mystacinus/brandtii, and P. auritus.
The number of Field Survey sites that contributed to the trend
for E. serotinus between 1998 and 2012 was lower than for the
other Field Survey species due to the limited distribution of this
species in the UK (Table 1). Scotland, north-west England and
north-east England were excluded from the E. serotinus model as
records for this species were either not present or very low (fewer
than ten instances over the survey period).
3.2. Power analysis
As expected, the power analysis of bat activity data from the
Field Survey showed that Red Alert level declines will be detected
much more quickly than Amber Alert declines (Table 2). The
existing Field Survey coverage, of almost 200 sites per year for
the period monitored to date (Table 1), was sufﬁcient to detect
Red Alert level declines of 2.73% per year in three of the species
(P. pipistrellus, P. pygmaeus and N. noctula) monitored through the
Field Survey with 80% power in 9 years (Table 2). Similarly, with
the current level of survey effort, an Amber Alert level decline of
1.14% per year could be detected in 10 years for P. pipistrellus
(Table 2), and 17–18 years for P. pygmaeus and N. noctula.
For the Roost Count and Hibernation Survey data, the ability to
detect alert level trends in relative abundance was estimated by
examining the width of the 95% conﬁdence intervals from the
smoothed GAM curves for each species. These ﬁgures (Table 3)
are most relevant when no signiﬁcant trend is apparent from the
analysis. A Red Alert decline of 2.73% per year would have reduced
the population by 30% between 1999 and 2012, whilst an Amber
Alert decline of 1.14% per year would have reduced the 2012 pop-
ulation by 14% from a base year of 1999. Thus, for example, the
width of the conﬁdence intervals for E. serotinus from Roost Counts
is 21.3% in 2012 (Table 3), suggesting that the trend would have
sufﬁcient power to detect a Red Alert but not an Amber Alert
decline. For the remaining species and surveys where no signiﬁ-
cant change was shown from the trend analysis (R. ferrumequinum
from the Hibernation Survey, M. nattereri from the Roost Count and
P. auritus from the Hibernation Survey and the Roost Count), the
conﬁdence intervals suggested that there was sufﬁcient power to
detect Red Alert changes only, except for winter counts of R.
3.3. Population trends
Changes in population trends across bat species in GB moni-
tored as part of the NBMP between 1997 or 1998 (depending on
the survey type) and 2012 were estimated for ten species or spe-
cies group, with all but two being monitored using two survey
types (Table 3,Figs. 2–4). Six species or species groups, R. fer-
rumequinum, R. hipposideros, M. daubentonii, M. nattereri, M. mys-
tacinus/brandtii and P. pipistrellus showed signiﬁcantly increasing
trends from at least one survey type, although where there were
trends from two survey types, the trend directions did not always
agree. The remaining species showed no overall change, except for
P. pygmaeus which showed no overall change from the Field Sur-
vey, but a signiﬁcant decrease from the Roost Count (Table 3).
Both Rhinolophus species showed an almost linear, signiﬁcant
increasing trend in population from the base year of 1999 to
2012 from summer roost counts (65.5% increase in R. hipposideros,
156.8% increase in R. ferrumequinum;Fig. 2,Table 3). A signiﬁcant
increase was also seen in the R. hipposideros trend from winter
hibernation counts, whilst winter counts of R. ferrumequinum
showed an increasing but not signiﬁcant trend due to high levels
of variation in counts and therefore wide conﬁdence limits
(Fig. 2,Table 3).
Field Survey power analysis. Number of years of Field Survey required for 80% power to detect Amber and Red Alert declines (i.e. 25% or
50% decline over 25 years respectively) for different numbers of sites surveyed per year for P. pipistrellus, P. pygmaeus and N. noctula.
Standard errors of estimates are less than 1 year, except for estimates involving substantial extrapolation beyond the 28 year simulations
which are shown as >30.
P. pipistrellus P. pygmaeus N. noctula
No. sites Red Alert Amber Alert Red Alert Amber Alert Red Alert Amber Alert
30 10 20 18 >30 20 >30
50 9 19 14 >30 15 >30
100814 1123 1223
200 6 10 9 17 9 18
18 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26
All species of Myotis bats showed signiﬁcant increases in popu-
lation from the base year of 1999 to 2012 from hibernation counts
(Fig. 3,Table 3). The population trend in M. daubentonii showed a
small but steady increase of 26.8% overall, with a similar, but not
signiﬁcant increase also seen in the summer activity data from
the Waterway Survey (Fig. 3,Table 3). The increasing trend in win-
ter counts of M. nattereri since 1999 was not supported by summer
roost counts which showed no overall trend in the period from
2000 to 2012 (Fig. 3,Table 3). The population trend in the small
Myotis species group (M. mystacinus/brandtii) from hibernation
counts showed a signiﬁcant increase of 41.7% overall since 1999
(Fig. 3,Table 3). No overall change was seen in the population of
P. auritus from either winter hibernation counts or summer roost
counts (Fig. 3,Table 3).
Trends in summer activity from the four species monitored
through the Field Survey differed considerably. P. pipistrellus
showed a signiﬁcant, increasing trend (52.2% increase) from sum-
mer activity data with a peak in 2010 but a contrasting almost lin-
ear, signiﬁcant decline of a similar magnitude (54.3% decrease)
from summer roost counts since the base year of 1999 (Fig. 4,
Table 3). P. pygmaeus showed no overall trend from summer activ-
ity data over the entire monitoring period, although there was a
signiﬁcant increase between 2005 and 2011. However, this species
also showed a signiﬁcant negative decline from summer roost
counts (Fig. 4,Table 3). Neither the population trend from summer
activity data nor from summer roost counts showed an overall sig-
niﬁcant change for E. serotinus from the 1999 base year to 2012,
although the trend directions were opposing, with a positive trend
Estimates of changes in bat populations from the base year to 2012 by species and survey type. Overall change indicates direction and signiﬁcance of
change (P< 0.05). The width of conﬁdence interval for the value of the index in 2012 is also given (the difference between the smoothed curve (solid
lines on Figs. 2–4) and its lower 95% limit (dashed lines on Figs. 2–4)).
Species Survey Base year % change from base
year to 2012
Overall change Conﬁdence interval
R. ferrumequinum Hibernation Survey 1999 61.2 Not signiﬁcant 74.9
Roost Count 1999 156.8 Increase 74.6
R. hipposideros Hibernation Survey 1999 75.2 Increase 28.6
Roost Count 1999 65.5 Increase 23.4
M. daubentonii Hibernation Survey 1999 26.8 Increase 20.5
Waterway Survey 1999 6.96 Not signiﬁcant 9.0
M. nattereri Hibernation Survey 1999 61.4 Increase 22.8
Roost Count 2002 22.2 Not signiﬁcant 28.4
M. mystacinus/brandtii Hibernation Survey 1999 41.7 Increase 36.1
P. auritus Hibernation Survey 1999 6.04 Not signiﬁcant 27.6
Roost Count 2001 18.5 No change 29.0
P. pipistrellus Field Survey 1999 52.2 Increase 17.9
Roost Count 1999 54.3 Decrease 11.1
P. pygmaeus Field Survey 1999 24.4 Not signiﬁcant 24.4
Roost Count 1999 45.1 Decrease 12.3
E. serotinus Field Survey 1999 30.8 Not signiﬁcant 33.3
Roost Count 1999 30.8 Not signiﬁcant 21.3
N. noctula Field Survey 1999 20.1 Not signiﬁcant 29.1
Fig. 2. Population trends for R. hipposideros (a and b) and R. ferrumequinum (c and d) from winter hibernation (a and c) and summer roost (b and d) counts. Graphs show
estimated trend from GAMs (solid line) with 95% conﬁdence limits (dashed lines) based on an index of 100 (grey line) at base year of 1999.
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 19
from summer activity data (30.8% increase) and a negative trend
from summer roost counts of the same magnitude (30.8% decrease;
Fig. 4,Table 3). There was no overall signiﬁcant trend in N. noctula,
although there was a signiﬁcant increase in activity from 2005 to a
peak in 2008 followed by a decline to 2012 (Fig. 4,Table 3).
3.4. Covariates included in trends
In most cases the differences between the GAMs with and with-
out covariates were minimal and related to the width of the conﬁ-
dence intervals, with the covariates reducing these rather than
altering the magnitude of the trend. For some of the Field Survey
results, however, bat detector model had a marked impact on trends.
The microphone sensitivity variable had a signiﬁcant impact on
trend analysis of both P. pipistrellus and P. pygmaeus as both species
were recorded more frequently when the most common bat detec-
tor model switched from a Batbox III to a Batbox Duet during the
monitoring period (F= 4.06, df = 5, 1776, P= 0.001 for P. pipistrellus
and F= 5.45, df = 5, 1678,P< 0.001 for P. pygmaeus testing for differ-
ences between all sensitivity groups). A number of other covariates
were also included in the Field Survey models (Table 4), but these
were variables such as wind, temperature and survey times, which
showed no consistent trend over time and so did not carry the same
risk of bias to the trend as the detector variables.
Fig. 3. Population trends for M. mystacinus/brandtii (a), M. daubentonii (b and c), M. nattereri (d and e), P. auritus (f and g) from winter hibernation counts (a, b, d and f),
summer bat detector surveys (c) and summer roost counts (e and g). Graphs show estimated trends from GAMs (solid line) with 95% conﬁdence limits (dashed lines) based on
an index of 100 (grey line) at base year of 1999 in all cases except M. nattereri Roost Count (base year 2002) and P. auritus Roost Count (base year 2001).
20 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26
Similarly, the covariates for the other survey types were mainly
used to help reduce the level of unexplained random variability
rather than to prevent bias. For each species trend, the analysis
with covariates was used when this achieved an increase in preci-
sion compared to the unadjusted trend (Table 4). For Hibernation
Survey GAM models, temperature parameters as measured by
surveyors (internal and external temperature) and the month of
survey had an impact on winter counts, with some differences
between species, except for those of R. ferrumequinum and
M. nattereri. For all Roost Counts, GAM models were based on the
maximum count at each roost in each year and GLMMs showed
that where only a single count was made at a site in a given year,
counts were lower than the peak of two counts. For the Waterway
Survey a number of covariates were ﬁtted that affected bat activity
recorded (Table 4).
In this study we used data from the National Bat Monitoring
Programme to estimate changes in populations across Great Britain
for ten bat species or species group using multiple survey types
assessing relative bat abundance in different seasons and summer
bat activity from 1997 to 2012. This dataset is unique in demon-
strating how data collected on relative abundance and activity of
bats by volunteers using standardised, multiple survey methods
can be utilised to produce statistically robust population indices
Fig. 4. Population trends for P. pipistrellus (a and b), P. pygmaeus (c and d), E. serotinus (e and f) and N. noctula (g) from summer bat detector surveys (a, c, e and g) and summer
roost counts (b, d and f). Graphs show estimated trend from GAMs (solid line) with 95% conﬁdence limits (dashed lines) based on an index of 100 (grey line) at base year of
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 21
for such a large proportion of a country’s bat fauna at a national
scale. A number of other studies have described changes in bat
populations over time from monitoring programmes across Europe
(reviewed in Battersby, 2010; Haysom et al., 2014) including some
that cover longer time series (Kervyn et al., 2009; Meschede and
Rudolph, 2010; Uhrin et al., 2010; Lesin
´ski et al., 2011). However,
many of these focussed on a single or small number of species, pro-
vided data from a single survey method, were limited to one phase
of the lifecycle, or were limited in geographical scale (Ingersoll
et al., 2013; Warren and Witter, 2002; Lesin
´ski et al., 2005,
2011). By equipping volunteers with the training and experience
required to take part in surveys using different methods, we have
developed a cost-efﬁcient way of determining changes in national
bat populations. On average, over 2000 surveys are carried out as
part of the NBMP each year. This involves volunteers investing
around 19 000 h in the programme annually, with an estimated
value of £276975, assuming a rate of £75 per 8-h day for volun-
teers taking part in the Roost Count and £150 per day for volun-
teers taking part in the other survey types which require a
higher skill level. By involving volunteers in bat monitoring, the
NBMP also connects people with wildlife and contributes to raising
awareness of bats and conservation among the public, which has
been identiﬁed as an important role of biodiversity monitoring at
a global scale (Bell et al., 2008; Jones et al., 2011).
Data collected through citizen science programmes are at risk of
a number of potential biases, for example due to observer quality
and sampling bias (Dickinson et al., 2010). Training of volunteers
and use of standardised methods have been identiﬁed as crucial
to the success of volunteer surveys (Newman et al., 2003;
Magurran et al., 2010). As part of the NBMP, training opportunities
are provided to volunteers to ensure that they can gain the skill
levels required to complete the different surveys. In this study
we have shown that the population indices estimated from the
NBMP data have sufﬁcient power to detect trends of at least a
50% decline over a 25-year period (Red Alert trend) for all species
with the exception of R. ferrumequinum, suggesting that our
approach to volunteer-based surveys results in high quality data
for population trend analysis at this scale, as has been demon-
strated elsewhere (Schmeller et al., 2009). The inclusion of covari-
ates in the trend models improved precision by taking into account
factors that result in increased variability or potential systematic
bias in the data. A number of factors can affect bat activity, as mon-
itored through the bat detector surveys, such as variations between
sites, between nights and the effects of weather (Verboom and
Spoelstra, 1999; Fischer et al., 2009). Differences in the detection
rates of bat detector types have been highlighted elsewhere
(Waters and Walsh, 1994; Adams et al., 2012) and by taking into
account changes over time in the availability of models of bat
detector, we have reduced the potential bias that could be intro-
duced by systematic changes in equipment used in the bat detector
surveys. The level of experience of volunteers may also affect their
ability to detect and identify some species that are monitored
using bat detectors, as has been found in other citizen science pro-
grammes for other taxa (Dickinson et al., 2010).
The NBMP was originally designed to detect serious declines in
bat species in order to ensure timely conservation action could be
taken (Walsh et al., 2001). This study, however, has revealed a gen-
erally favourable picture for bats over the monitoring period with
all species monitored showing a stable or increasing trend in GB
from the base year of surveys to 2012 from at least one survey
type. Although there is evidence of signiﬁcant decreases in bat
populations in the UK and Europe in the second half of the last cen-
Description of the covariates included in the population trend analyses by survey type and species.
Survey Species Covariates Effect of covariate
Roost Count All species Single count Analyses based on maximum count in
each roost in each year. Maxima from
multiple counts tend to be higher than
those from single counts
Hibernation Survey R. hipposideros, M. daubentonii,
M. mystacinus/brandtii, P. auritus
Month Temporal trends vary between species
Temperature Fitted as Principal Components Analysis of
external temperature and internal
temperatures at three locations
Field Survey P. pipistrellus Detector sensitivity Grouped variable for microphone
Wind Fewer bats in windy conditions
Temperature Fewer bats in cool conditions
Identiﬁcation skill level Bats more likely to be detected by
surveyors with higher skill levels
Survey start time Fewer bats when survey started closer to
P. pygmaeus Detector sensitivity Grouped variable for microphone
Date More bats at end of survey period
E. serotinus Wind Fewer bats in windy conditions
Temperature Fewer bats in cool conditions
Survey time More bats with long survey length
N. noctula Date More bats at end of survey period
Temperature Fewer bats in cool conditions
Survey time More bats with long survey length
Waterway Survey M. daubentonii Wind Fewer bats in windy conditions
Rain Fewer bats in wet conditions
Temperature Fewer bats at low temperatures
Identiﬁcation skill level ‘Sure’ presence conﬁrmed at more point
locations with increasing surveyor skill
Number of clear spots More bats when higher number of clear
spots along transect
22 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26
tury, thought to be caused by loss of habitat and disturbance or
destruction of roosts (Racey and Stebbings, 1972; Stebbings,
1988; Harris et al., 1995; Kervyn et al., 2009), the population indi-
ces from this study suggest that in GB, bats have started to show
signs of recovery in the last two decades. A recent study investigat-
ing the possibilities of using counts from hibernation surveys to
indicate bat population trends in Europe between 1993 and
2011, including data from this programme, presented a similar
pattern over nine countries, with eight of the species monitored
in the NBMP also showing stable or increasing trends at the
European level (Haysom et al., 2014; Van der Meij et al., in press).
Three of the species monitored in the programme, M. mystaci-
nus/brandtii which were considered together and N. noctula, were
monitored using a single survey method only. Of the remaining
eight species, three showed agreement in the direction and
signiﬁcance level of population change between survey methods
(R. hipposideros,P. auritus and E. serotinus), two showed agreement
in trend direction but not signiﬁcance level (R. ferrumequinum and
M. daubentonii) and three (M. nattereri,P. pipistrellus and P. pygma-
eus) showed differences in both trend direction and signiﬁcance of
trend between survey types. Further analysis is needed to under-
stand differences found in trends from relative abundance in dif-
ferent seasons (e.g. M. nattereri) and differences between trends
based on relative abundance of bats and trends based on bat activ-
ity data (P. pipistrellus and P. pygmaeus). It is possible that the dif-
ferences in distribution of sites between the random stratiﬁed
sampling approach of the bat detector surveys compared to the
self-selection of sites for the counts at summer and winter roosts
could lead to differences in the overall trends. Furthermore, many
bat species regularly move summer roost locations both within
and between years (Lewis, 1995). A bat colony will often use a
number of roost sites during the summer breeding period and
move between sites (Zeale et al., 2014), although the frequency
of roost switching varies between species. For example, P. pygma-
eus appears to switch roost sites less often than P. pipistrellus
(Barlow and Jones, 1999; Davidson-Watts and Jones, 2005). The
movement of bats between roost sites is likely to results in biases
in the trends deduced from summer Roost Counts. Because the
roost locations for monitoring are selected by volunteers, entire
colonies of bats are not usually monitored over time, only the num-
ber of bats in particular roost locations that are part of the pro-
gramme. Population trends based on bat activity data may be
more representative of changes in abundance than summer roost
counts therefore, particularly for Pipistrellus species, as it has been
shown that bat activity measured from acoustic surveys correlates
with relative abundance measured from trapping studies of these
bats (Lintott et al., 2014). The differences in trends between survey
types do however highlight the need to ensure that surveys using
different methods and covering more than one season are consid-
ered in the further development of monitoring programmes, as has
been shown in tropical habitats (MacSwiney et al., 2008; Meyer
et al., 2011): it is important for comprehensive bat monitoring pro-
grammes to assess bat populations throughout the year where
there is seasonality in habitat use.
4.1. Limitations of the NBMP
Whilst these results present a generally positive picture for bat
populations in GB since 1997, the species coverage of the pro-
gramme is not yet fully comprehensive, and there are some impor-
tant omissions. Rarer bat species that are habitat specialists, for
example Bechstein’s bat Myotis bechsteinii and barbastelle Barba-
stella barbastellus, both woodland specialist species (Harris and
Yalden, 2008), and also grey long-eared bat Plecotus austriacus,a
grassland specialist whose population is probably less than 1000
bats in GB (Razgour et al., 2011, 2013), are not included as part
of the programme as they are difﬁcult to monitor or rarely encoun-
tered (Walsh et al., 2001). The NBMP therefore currently describes
trends in the more common and widespread species found in GB
and may not wholly reﬂect pressures and impacts on all resident
bat species. For example, due to roosts and hibernacula being more
easily detected in some locations than others, summer roost count
surveys show bias towards species that are closely associated with
buildings and other structures and winter surveys towards species
that hibernate in large structures that are easily accessible to vol-
unteers. Population changes in species that are more commonly
found roosting in trees both in summer and winter, for example
B. barbastellus (Russo et al., 2004; Harris and Yalden, 2008), cannot
be effectively assessed using these monitoring techniques. The
NBMP aspires to correct such omissions and biases by increasing
the number of species that are surveyed, including those that are
most challenging, pending the development of new surveillance
approaches. It is particularly important that rarer species with spe-
cialist habitat requirements are equally represented in monitoring
schemes as common and widespread habitat generalists, since in
some other taxa it has been shown that whilst common or gener-
alist species have increased, specialists have declined (Clavel et al.,
2011; Davey et al., 2012; Le Viol et al., 2012).
The bat detector surveys in the NBMP use tuneable, heterodyne
bat detectors that can be used to determine differences between
echolocation calls produced by different bat species (Ahlén and
Baagøe, 1999). Identiﬁcation is carried out in the ﬁeld and it is
therefore important that surveyors taking part can accurately dis-
tinguish between different species. Although it has been argued
that accurate species identiﬁcation is possible for a wide range of
European bat species using this method (Ahlén and Baagøe,
1999), some error in identiﬁcation is likely. Our programme utilis-
es popular bat detectors that are accessible and affordable to a
wide range of volunteers and, in order to encourage mass partici-
pation, focuses on a smaller number of species that are relatively
easy to identify. More objective methods of species identiﬁcation
using quantitative analysis of parameters measured from recorded
calls are now available however for a wide range of species
(Walters et al., 2012).
The application of smoothed curves to the population trend
data carried out in this study has highlighted the overall pattern
of change over the monitoring period whilst taking into account
annual ﬂuctuations in abundance (Fewster et al., 2000). The
trend analysis does not provide information on the causes of
change during this period. The NBMP does however provide
the information required to carry out further work to link the
trends to changes in land use and climate for example, as has
been investigated for birds (Pearce-Higgins and Gill, 2010;
Eglington and Pearce-Higgins, 2012). Factors that may have
inﬂuenced changes in bat populations since 1997 include the
inﬂuence of legal protection and raised awareness of bat conser-
vation, changes in climate and changes in agricultural practices.
The legal protection of bats has led to an increasing interest in
their conservation since the 1980s (Mitchell-Jones et al., 1993).
The resulting improvements in public awareness and develop-
ment of the infrastructure to support bat conservation
(Haysom et al., 2010) have contributed to the favourable trends
in bat populations shown in this study, demonstrating that spe-
cies protection can assist in improving understanding and con-
servation of species. Future work to determine the relative
importance of key drivers of change in bat populations is the
Monitoring population change through assessment of relative
abundance and activity provides a measure of relative change over
time but does not give estimates of population size. Estimating
population size accurately may be important to quantify the extent
of population change and hence future viability, in response to spe-
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 23
ciﬁc threats (Jones et al., 2009). This has been attempted in a small
number of studies, for example in the case of the impact of the
novel fungal disease, white nose syndrome, currently affecting
bat populations in the USA (Thogmartin et al., 2013). For the bat
species monitored through the NBMP, populations estimates are
mainly based on expert opinion (Harris et al., 1995) and there is
currently only very limited information on the additional
parameters required to enable more quantitative assessments of
population size to be estimated, such as demographic rates and
population dynamics (Chauvenet et al., 2014).
4.2. Future directions in bat monitoring
This study has shown that volunteer programmes can be used
as a successful approach to monitoring bat populations and that
data collected by volunteers can be used to provide statistically
robust species population trends. Data from the programme have
also contributed to an improved understanding of bat distribution,
to the production of bat indicators which contribute to the ofﬁcial
government biodiversity indicators in the UK (Anon, 2013) and in
recent research to improve our understanding of landscape inﬂu-
ences, particularly the importance of agricultural woodlands and
hedgerows for bats (Boughey et al., 2011a,b).
The integration of bat detector surveys into the programme in
addition to the more traditional methods of counting bats at sum-
mer and winter roosts, that have been the core element of most
long-term bat monitoring schemes in Europe (Kervyn et al.,
2009; Battersby, 2010; Horác
ˇek, 2010; Meschede and Rudolph,
´ski et al., 2011), provides trends based on bat activity
in foraging habitats. Since the NBMP was designed, bat detector
technology has moved forward and broadband bat detectors that
allow simultaneous sampling of all echolocation call frequencies
are more widely available at lower cost. These have been used in
more recent approaches to designing bat monitoring programmes
(Rodhouse et al., 2012). A number of quantitative methods for spe-
cies identiﬁcation have also been developed (Parsons and Jones,
2000; Obrist et al., 2004; Walters et al., 2012). These techniques
have been successfully used in surveys utilising volunteers, partic-
ularly through the use of driven transects, for example in France
(Kerbiriou et al., 2010), Ireland (Roche et al., 2011) and across East-
ern Europe through the iBats programme (Jones et al., 2013). We
suggest that, where a source of potential volunteer surveyors are
available, the model of the NBMP using volunteers to take part in
surveys of different levels through provisioning of appropriate
training and central volunteer support, in combination with the
integration of broadband bat detector surveys that allow increased
species coverage and objectivity of species identiﬁcation, is the
way forward for developing comprehensive bat monitoring pro-
grammes across a broad geographical scale.
This study has also shown that the structured approach to the
citizen science scheme utilised in the UK to monitor bats provides
a suitable framework for a long-term species monitoring system
(Beeker et al., 2013). The approach comprises the development
and testing of simple, standardised survey methods which focus
on species that are relatively easy to identify using the selected
methods (Walsh et al., 2001), central co-ordination of the pro-
gramme, investment in recruitment, training and retention of vol-
unteers, the use of selected volunteers to provide training and local
support to other volunteers and statistical analyses designed to
minimise inherent biases in the data. The implementation of the
programme through a partnership between government and a
not-for-proﬁt organisation has inﬂuenced its tenacity (Racey,
2013). We recommend that these key features should be consid-
ered in the development of other citizen science monitoring pro-
grammes, being applicable to a wide range of schemes, survey
methods and taxa.
Role of the funding source
The National Bat Monitoring Programme was funded between
1996 and 2000 by the Department of the Environment, Transport
and the Regions (DETR Contract Reference No. CR018). The pro-
gramme is now run as a partnership between the Bat Conservation
Trust, Joint Nature Conservation Committee, National Resources
Wales and Defra with additional funding provided by Natural Eng-
land. All partners have provided a ﬁnancial contribution to the run-
ning of the programme and agree to the submission of this article
based on the outputs of the programme for publication. An annual
report of the National Bat Monitoring Programme is published
online at http://www.bats.org.uk/
We are extremely grateful to all the dedicated volunteers who
have taken part in the NBMP since its inception in 1996, without
which the programme could not continue. We are also grateful
to all those who have worked at BCT and who have contributed
to the NBMP during this period, in particular Colin Catto and Jules
Agate. We thank Kirsty Park, Kate Jones, Brock Fenton, an anony-
mous reviewer and Robin Pakeman for useful comments on the
manuscript. The NBMP is a partnership between the Bat Conserva-
tion Trust, Joint Nature Conservation Committee, National
Resources Wales and Defra with additional funding provided by
Natural England. Data are available from the authors.
Adams, A.M., Jantzen, M.K., Hamilton, R.M., Fenton, M.B., 2012. Do you hear what I
hear? Implications of detector selection for acoustic monitoring of bats.
Methods Ecol. Evol. 3, 992–998.
Ahlén, I., Baagøe, H.J., 1999. Use of ultrasound detectors for bat studies in Europe:
experiences from ﬁeld identiﬁcation, surveys, and monitoring. Acta Chiropterol.
Anon, 2013. UK Biodiversity Indicators in Your Pocket 2013. Department for
Environment, Food and Rural Affairs, London. <http://jncc.defra.gov.uk/pdf/
Balmford, A., Green, R.E., Jenkins, M., 2003. Measuring the changing state of nature.
Trends Ecol. Evol. 18, 326–330.
Barlow, K.E., Jones, G., 1999. Roosts, echolocation calls and wing morphology of two
phonic types of Pipistrellus pipistrellus. Z. Säugertierkd. 64, 257–268.
Battersby, J., 2010. Guidelines for Surveillance and Monitoring of European Bats.
EUROBATS Publication Series No. 5. UNEP/EUROBATS Secretariat, Bonn,
Beeker, T.A., Millenbah, K.F., Gore, M.L., Lundrigan, B.L., 2013. Guidelines for
creating a bat-speciﬁc citizen science acoustic monitoring programme. Hum.
Dimens. Wildl. 18, 58–67.
Bell, S., Marzano, M., Cent, J., Kobierska, H., Podjed, D., Vandzinskaite, D., Reinart, H.,
Armaitiene, A., Grodzin
´ska-Jurczak, M., Muršic
ˇ, R., 2008. What counts?
Volunteers and their organisations in the recording and monitoring of
biodiversity. Biodivers. Conserv. 17, 3443–3454.
Boughey, K.L., Lake, I.R., Haysom, K.A., Dolman, P.A., 2011a. Improving the
biodiversity beneﬁts of hedgerows: How physical characteristics and the
proximity of foraging habitat affect the use of linear features by bats. Biol.
Conserv. 144, 1790–1798.
Boughey, K.L., Lake, I.R., Haysom, K.A., Dolman, P.A., 2011b. Effects of landscape-
scale broadleaved woodland conﬁguration and extent on roost location for six
bat species across the UK. Biol. Conserv. 144, 2300–2310.
Boyles, J.G., Cryan, P.M., McCracken, G.F., Kunz, T.H., 2011. Economic importance of
bats in agriculture. Science 332, 41–42.
Chamberlain, D.E., Fuller, R.J., Bunce, R.G.H., Duckworth, J.C., Shrubb, M., 2000.
Changes in the abundance of farmland birds in relation to the timing of
agricultural intensiﬁcation in England and Wales. J. Appl. Ecol. 37, 771–788.
Chauvenet, A.L.M., Hutson, A.M., Smith, G.C., Aegerter, J.N., 2014. Demographic
variation in the U.K. serotine bat: ﬁlling gaps in knowledge for management.
Ecol. Evol. 4, 3820–3829. http://dx.doi.org/10.1002/ece3.1174.
Clavel, J., Julliard, R., Devictor, V., 2011. Worldwide decline of specialist species:
toward a global functional homogenization? Front. Ecol. Environ. 9, 222–228.
Davey, C.M., Chamberlain, D.E., Newson, S.E., Noble, D.G., Johnston, A., 2012. Rise of
the generalists: evidence for climate driven homogenization in avian
communities. Global Ecol. Biogeogr. 21, 568–578.
Davidson-Watts, I., Jones, G., 2005. Differences in foraging behaviour between
Pipistrellus pipistrellus (Schreber, 1774) and Pipistrellus pygmaeus. J. Zool. 268,
24 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26
Dickinson, J.L., Zuckerberg, B., Bonter, D.N., 2010. Citizen science as an ecological
research tool: challenges and beneﬁts. Annu. Rev. Ecol. Evol. Syst. 41, 149–172.
Eglington, S.M., 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, e30407.
Ekman, M., de Jong, J., 1996. Local patterns of distribution and resource utilisation of
four bat species (Myotis brandtii,Eptesicus nilssoni, Plecotus auritus and
Pipistrellus pipistrellus) in patchy and continuous environments. J. Zool. 238,
Fenton, M.B., 1970. A technique for monitoring bat activity with results obtained
from different environments in southern Ontario. Can. J. Zool. 48, 847–851.
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.
Fischer, J., Stott, J., Law, B.S., Adams, M.A., Forrester, R.I., 2009. Designing effective
habitat studies: quantifying multiple sources of variability in bat activity. Acta
Chiropterol. 11, 127–137.
Fujita, M.S., Tuttle, M.D., 1991. Flying foxes (Chiroptera: Pteropodidae): threatened
animals of key ecological and economic importance. Conserv. Biol. 5, 455–463.
Gerard, P.D., Smith, D.R., Weerakkody, G., 1998. Limits of retrospective power
analysis. J. Wildlife Manage. 62, 801–807.
Gerell, R., Lundberg, K.G., 1993. Decline of a bat Pipistrellus pipistrellus population in
an industrialized area in south Sweden. Biol. Conserv. 65, 153–157.
Gregory, R.D., van Strien, A., Vorisek, P., Gmelig Mayling, A.W., Noble, D.G., Foppen,
R.P.B., Gibbons, D.W., 2005. Developing indicators for European birds. Philos.
Trans. Roy. Soc. London B 360, 269–288.
Harris, S., Yalden, D.W., 2008. Mammals of the British Isles: Handbook, fourth ed.
Mammal Society, 799 pp.
Harris, S., Morris, P., Wray, S., Yalden, D., 1995. A review of British Mammals:
Population Estimates and Conservation Status of British Mammals Other Than
Cetaceans. Joint Nature Conservation Committee, Peterborough, 183 pp.
Haysom, K.A., Jones, G., Merrett, D., Racey, P.A., 2010. Bats. In: Maclean, N. (Ed.),
Silent Summer. The State of Wildlife in Britain and Ireland. Cambridge
University Press, Cambridge, pp. 259–280.
Haysom, K.A., Dekker, J., Russ, J., Van der Meij, T., Van Strien, A. 2014. European bat
population trends. A prototype biodiversity indicator. European Environment
Agency Technical Report No 19/2013.
Hoenig, J.M., Heisey, D.M., 2001. The abuse of power: the pervasive fallacy of power
calculations for data analysis. Am. Stat. 55, 1–6.
ˇek, I., 2010. Monitoring bats in underground hibernacula. In: Horác
Uhrin, M. (Eds.), A tribute to Bats. Lesnické Práce, Kostelec nad C
Hutson, A.M., Mickleburgh, S.P., Racey, P.A., 2001. IUCN/SSC Chiroptera Specialist
Group. Microchiropteran Bats: Global Status Survey and Conservation Action
Ingersoll, T.E., Sewall, B.J., Amelon, S.K., 2013. Improved analysis of long-term
monitoring data demonstrates marked regional declines of bat populations in
the Eastern United States. PLoS ONE 8, e65907.
Jones, G., Jacobs, D.S., Kunz, T.H., Willig, M.R., Racey, P.A., 2009. Carpe noctem: the
importance of bats as bioindicators. Endanger. Species Res. 8, 93–115.
Jones, J.P.G., Collen, B., Atkinson, G., Baxter, P.W.J., Bubb, P., Illian, J.B., Katzner, T.E.,
Keane, A., Loh, J., McDonald-Madden, E., Nicholson, E., Pereira, H.M.,
Possingham, H.P., Pullin, A.S., Rodrigues, A.S.L., Ruiz-Gutierrez, V.,
Sommerville, M., Milner-Gulland, E.J., 2011. The why, what, and how of global
biodiversity indicators beyond the 2010 target. Conserv. Biol. 25, 450–457.
Jones, K.E., Russ, J.A., Bashta, A., Bilhari, Z., Catto, C., Cs}
osz, I., Gorbachev, A., Gy}
P., Hughes, A., Ivashkiv, I., Koryagina, N., Kurali, A., Langton, S., Collen, A.,
Margiean, G., Pandourski, I., Parsons, S., Prokofev, I., Szodoray-Paradi, A.,
Szodoray-Paradi, F., Tilova, E., Walters, C.L., Weatherill, A., Zavarzin, O., 2013.
Indicator bats program: a system for the global acoustic monitoring of bats. In:
Collen, B., Pettorelli, N., Baillie, J.E.M., Durant, S.M. (Eds.), Biodiversity
Monitoring and Conservation: Bridging the Gap Between Global Commitment
and Local Action. John Wiley & Sons, Ltd., pp. 211–247.
Kalka, M.B., Smith, A.R., Kalko, E.K.V., 2008. Bats limit arthropods and herbivory in a
tropical forest. Science 320, 71.
Kerbiriou, C., Julien, J.F., Deguines, N., Gasc, A., Lorillière, R., Lois, G., Jiguet, F.,
Juilliard, R., Couvet, D., 2010. Suivi temporel des chauves-souris communes,
premiers résultats et perspectives. Symbioses 25, 73–79.
Kervyn, T., Lamotte, S., Nyssen, P., Verschuren, J., 2009. Major decline of bat
abundance and diversity during the past 50 years in southern Belgium. Belg. J.
Zool. 139, 124–132.
Kunz, T.H., Arnett, E.B., Erickson, W.P., Hoar, A.R., Johnson, G.D., Larkin, R.P.,
Strickland, M.D., Thresher, R.W., Tuttle, M.D., 2007. Ecological impacts of wind
energy development on bats: questions, research needs, and hypotheses. Front.
Ecol. Environ. 5, 31–324.
Kunz, T.H., Braun de Torrez, E., Bauer, D., Lobova, T., Fleming, T.H., 2011. Ecosystem
services provided by bats. Ann. N. Y. Acad. Sci. 1223, 1–38.
Kunz, T.H., Parsons, S., 2009. Ecological and Behavioural Methods for the Study of
Bats. Johns Hopkins University Press, pp. 901.
Le Viol, I., Jiguet, F., Brotons, L., Herrando, S., Lindström, A., Pearce-Higgins, J.W., Reif,
J., Van Turnhout, C., Devictor, V., 2012. More and more generalists: two decades
of changes in the European avifauna. Biol. Lett. 8, 780–782.
´ski, G., Fuszara, E., Fuszara, M., Jurczyszyn, M., Urban
´czyk, Z., 2005. Long-term
changes in the numbers of the barbastelle Barbastella barbastellus in Poland.
Folia Zool. 54, 351–358.
´ski, G., Ignaczak, M., Kowalski, M., 2011. Increasing bat abundance in a major
winter roost in central Poland over 30 years. Mammalia 75, 163–167.
Lewis, S.E., 1995. Roost ﬁdelity of bats: a review. J. Mammal. 76, 481–496.
Lintott, P.R., Fuentes-Montemayor, E., Goulson, D., Park, K.J., 2014. Testing the
effectiveness of surveying techniques in determining bat community
composition within woodland. Wildlife Res. 40, 675–684.
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
biodiversity. Philos. Trans. Roy. Soc. London B 360, 289–295.
MacSwiney, G.M.C., Clarke, F.M., Racey, P.A., 2008. What you see is not what you
get: the role of ultrasonic detectors in increasing inventory completeness in
Neotropical bat assemblages. J. Appl. Ecol. 45, 1364–1371.
Magurran, A.E., Baillie, S.R., Buckland, S.T., Dick, J.M.P., Elston, D.A., Scott, E.M.,
Smith, R.I., Somerﬁeld, P.J., Watt, A.D., 2010. Long-term datasets in biodiversity
research and monitoring: assessing change in ecological communities through
time. Trends Ecol. Evol. 25, 574–582.
Marchant, J.H., Forrest, C., Greenwood, J.J.D., 1997. A review of large-scale
generic population monitoring schemes in Europe. Bird Census News
Meschede, A., Rudolph, B.U., 2010. 1985–2009: 25 Jahre Fledermausmonitoring in
Bayern, Bayerisches Landesamt für Umwelt, Augsburg. <http://
Meyer, C.F.J., Aguiar, L.M.S., Aguirre, L.F., Baumgarten, J., Clarke, F.M., Cosson, J.-F.,
Estrada Villegas, S., Fahr, J., Faria, D., Furey, N., Henry, M., Hodgkison, R.,
Jenkins, R.K.B., Jung, K.G., Kingston, T., Kunz, T.H., MacSwiney Gonzalez, M.C.,
Moya, I., Pons, J.-M., Racey, P.A., Rex, K., Sampaio, E.M., Stoner, K.E., Voigt, C.C.,
von Staden, D., Weise, C.D., Kalko, E.K.V., 2010. Long-term monitoring
of tropical bats for anthropogenic impact assessment: Gauging the
statistical power to detect population change. Biol. Conserv. 143,
Meyer, C.F.J., Aguiar, L.M.S., Aguirre, L.F., Baumgarten, J., Clarke, F.M., Cosson, J.-F.,
Villegas, S.E., Fahr, J., Faria, D., Furey, N., Henry, M., Hodgkison, R., Jenkins, R.K.B.,
Jung, K.G., Kingston, T., Kunz, T.H., Cristina MacSwiney Gonzalez, M., Moya, I.,
Patterson, B.D., Pons, J.-M., Racey, P.A., Rex, K., Sampaio, E.M., Solari, S., Stoner,
K.E., Voigt, C.C., von Staden, D., Weise, C.D., Kalko, E.K.V., 2011. Accounting for
detectability improves estimates of species richness in tropical bat surveys. J.
Appl. Ecol. 48, 777–787.
Mickleburgh, S.P., Hutson, A.M., Racey, P.A., 1992. IUCN/SSC Chiroptera Specialist
Group. Old World Fruit Bats: An Action Plan for their Conservation, IUCN.
Mitchell-Jones, A.J., Hutson, A.M., Racey, P.A., 1993. The growth and development of
bat conservation in Britain. Mammal Rev. 23, 139–148.
Newman, C., Buesching, C.D., Macdonald, D.W., 2003. Validating mammal
monitoring methods and assessing the performance of volunteers in wildlife
conservation—‘‘Sed quis custodiet ipsos custodies?’’. Biol. Conserv. 113, 189–
Newson, S.E., Mendes, S., Crick, H.Q.P., Dulvy, N.K., Houghton, J.D.R., Hays, G.C.,
Hutson, A.M., MacLeod, C.D., Pierce, G.J., Robinson, R.A., 2009. Indicators of the
impact of climate change on migratory species. Endanger. Species Res. 7, 101–
O’Shea, T.J., Bogan, M.A., 2003. Monitoring trends in bat populations of the United
States and territories: problems and prospects. U.S. Geological Survey,
Biological Resources Discipline, Information Technology Report USGS/BRD/
Obrist, M.K., Boesch, R., Flückiger, P.F., 2004. Variability in echolocation call design
of 26 Swiss bat species: consequences, limits and options for automated ﬁeld
identiﬁcation with a synergetic pattern recognition approach. Mammalia 68,
Parsons, S., Jones, G., 2000. Acoustic identiﬁcation of twelve species of echolocating
bat by discriminant function analysis and artiﬁcial neural networks. J. Exp. Biol.
Pearce-Higgins, J.W., Gill, J.A., 2010. Unravelling the mechanisms linking climate
change, agriculture and avian population declines. Ibis 152, 439–442.
Penone, C., Le Voil, I., Pellissier, V., Julien, J.-F., Bas, Y., Kerbiriou, C., 2013. Use of
large-scale acoustic monitoring to assess anthropogenic pressures on
Orthoptera communities. Conserv. Biol. 27, 979–987.
Puechmaille, S.J., Frick, W.F., Kunz, T.H., Racey, P.A., Voigt, C.C., Wibbelt, G., Teeling,
E.C., 2011. White-nose syndrome: is this emerging disease a threat to European
bats? Trends Ecol. Evol. 26, 570–576.
Racey, P.A., 2013. Bat conservation: past, present and future. In: Adams, R.A.,
Pedersen, S.C. (Eds.), Bat Ecology, Evolution and Conservation. Springer, pp.
Racey, P.A., Stebbings, R.E., 1972. Bats in Britain — a status report. Oryx 11, 319–327.
Ransome, R.D., 1989. Population changes of greater horseshoe bats studied near
Bristol over the past twenty-six years. Biol. J. Linn. Soc. 38, 71–82.
Raven, P.J., Holmes, N.T.H., Dawson, F.H., Everard, M., 1998. Quality assessment
using river habitat survey data. Aquat. Conserv. Mar. Freshw. Ecosys. 8, 477–
Razgour, O., Whitby, D., Dahlberg, E., Barlow, K., Hanmer, J., Haysom, K., McFarlane,
H., Wicks, L., Williams, C., Jones, G., 2013. Conserving grey long-eared bats
(Plecotus austriacus) in our landscape: a conservation management plan.
Available to download from the Bat Conservation Trust <http://
Razgour, O., Hanmer, J., Jones, G., 2011. Using multi-scale modelling to predict
habitat suitability for species of conservation concern: the grey long-eared bat
as a case study. Biol. Conserv. 144, 2922–2930.
Rebelo, H., Tarroso, P., Jones, G., 2010. Predicted impact of climate change on
European bats in relation to their biogeographic patterns. Glob. Change Biol. 16,
K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26 25
Roche, N., Langton, S., Aughney, T., Russ, J.M., Mardell, F., Lynn, D., Catto, C., 2011. A
car-based monitoring method reveals new information on bat populations and
distributions in Ireland. Anim. Conserv. 14, 642–651.
Rodhouse, T.J., Ormsbee, P.C., Irvine, K.M., Vierling, L.A., Szewczak, J.M., Vierling,
K.T., 2012. Assessing the status and trend of bat populations across broad
geographic regions with dynamic distribution models. Ecol. Appl. 22, 1098–
Russ, J., 2012. British Bat Calls. A Guide to Species Identiﬁcation. Pelagic Publishing,
Russo, D., Cistrone, L., Jones, G., Mazzoleni, S., 2004. Roost selection by barbastelle
bats (Barbastella barbastellus, Chiroptera: Vespertilionidae) in beech woodlands
of central Italy: consequences for conservation. Biol. Conserv. 117, 73–81.
Schmeller, D.S., Henry, P.-Y., Julliard, R., Gruber, B., Clobert, J., Dziock, F., Lengyel, S.,
Nowicki, P., Déri, E., Budrys, E., Kull, T., Tali, K., Bauch, B., Settele, J., van Swaay,
C., Kobler, A., Babij, V., Papastergiadou, E., Henle, K., 2009. Advantages of
volunteer-based biodiversity monitoring in Europe. Conserv. Biol. 23, 307–316.
Sewell, D., Guillera-Arroita, G., Grifﬁths, R.A., Beebee, T.J.C., 2012. When is a species
declining ? Optimizing survey effort to detect population changes in reptiles.
PLoS ONE 7, e43387.
Simmons, N.B., 2005. Order chiroptera. In: Wilson, D.E., Reeder, D.M. (Eds.),
Mammal Species of the World: a Taxonomic and Geographic Reference. Johns
Hopkins University Press, pp. 213–529.
Stebbings, R.E., 1988. Conservation of European Bats. Christopher Helm, London.
Thogmartin, W.E., Sanders-Reed, C.A., Szymanski, J.A., McKann, P.C., Pruitt, L., King,
R.A., Runge, M.C., Russell, R.E., 2013. White-nose syndrome is likely to extirpate
the endangered Indiana bat over large parts of its range. Biol. Conserv. 160,
Uhrin, M., Benda, P., Obuch, J., Urban, P., 2010. Changes in abundance of hibernating
bats in central Slovakia (1992–2009). Biologia 65, 349–361.
Van der Meij, T., Van Strien, A.J., Haysom, K.A., Dekker, J., Russ, J., Biala, K., Bihari, Z.,
Jansen, E., Langton, S., Kurali, A., Limpens, H., Meschede, A., Petersons, G.,
Presetnik, P., Prüger, J., Reiter, G., Rodrigues, L., Schorcht, W., Uhrin, M., Vintulis,
V., 2014. Return of the bats? A prototype indicator of trends in European bat
populations in underground hibernacula. Mamm. Biol. http://dx.doi.org/
10.1016/j.mambio.2014.09.004, in press.
Verboom, B., Spoelstra, K., 1999. Effects of food abundance and wind on the use of
tree lines by an insectivorous bat, Pipistrellus pipistrellus. Can. J. Zool. 77, 1393–
Walsh, A.L., Catto, C.M.C., Hutson, A.M., Langton, S., Racey, P.A., 2003. The United
Kingdom National Bat Monitoring Programme: Turning Conservation Goals into
tangible results. In: O’Shea, T.J., Bogan, M.A., (Eds.), Monitoring Trends in Bat
Populations of the United States and Territories: Problems and Prospects. U.S.
Geological Survey, Biological Resources Discipline, Information Technology
Report USGS/BRD/ITR-2003-0003, pp. 103–117.
Walsh, A.L., Catto, C.M.C., Hutson, A.M., Racey, P.A., Richardson, P.W., Langton, S.D.,
2001. The UK’s National Bat Monitoring Programme, Final Report 2001. DETR
Contract Ref. CR018. The Bat Conservation Trust, London.
Walsh, A.L., Harris, S., 1996a. Factors determining the abundance of vespertilionid
bats in Britain: geographical, land class and local habitat relationships. J. Appl.
Ecol. 33, 519–529.
Walsh, A.L., Harris, S., 1996b. Foraging habitat preferences of vespertilionid bats in
Britain. J. Appl. Ecol. 33, 508–518.
Walsh, A.L., Harris, S., Hutson, A.M., 1995. Abundance and habitat selection of
foraging vespertilionid bats in Britain: a landscape-scale approach. Symp. Zool.
Soc. London 67, 325–344.
Walters, C.L., Freeman, R., Collen, A., Dietz, C., Fenton, M.B., Jones, G., Obrist, M.K.,
Puechmaille, S.J., Sattler, T., Siemers, B.M., Parsons, S., Jones, K.E., 2012. A
continental-scale tool for acoustic identiﬁcation of European bats. J. Appl. Ecol.
Warren, R.D., Witter, M.S., 2002. Monitoring trends in bat populations through roost
surveys: methods and data from Rhinolophus hipposideros. Biol. Conserv. 105,
Waters, D., Walsh, A.L., 1994. The inﬂuence of bat detector brand on the
quantitative estimation of bat abundance. Bioacoustics 5, 205–221.
Wright, L.J., Newson, S.E., Noble, D.G., 2013. The value of a random sampling design
for annual monitoring of national populations of larger British terrestrial
mammals. Eur. J. Wildl. Res. 60, 213–221.
Zeale, M., Stone, E., Bennitt, E., Newson, S., Parker, P., Haysom, K., Browne, W.J.,
Harris, S., Jones, G., 2014. Improving mitigation success where bats occupy
houses and historic buildings, particularly churches. Defra Research Project
WM0322, Final Report. Defra, London, pp. 125.
Zuur, A., Ieno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects
Models and Extensions in Ecology with R. Springer, pp. 574.
26 K.E. Barlow et al. / Biological Conservation 182 (2015) 14–26