Differential effects of artificial lighting on flight and foraging behaviour of two sympatric bat species

Abstract

Human habitation in deserts can create rich novel resources that may be used by native desert species. However, at night such resources may lose attractiveness when they are in artificially lit areas. For bats, attraction to such manmade habitats might be species specific. In an isolated village in the Negev desert that is known for its high bat activity we investigated the effects of artificial lighting on flight behaviour of two aerial insectivorous bat species: Pipistrellus kuhlii, a non‐desert synanthropic bat, common in urban environments and Eptesicus bottae, a desert‐dwelling species. Using an acoustic tracking system we reconstructed flight trajectories for bats that flew under artificial lights [Light treatment (L)] versus in natural darkness [Dark treatment (D)]. Under L both P. kuhlii and E. bottae flew significantly faster than under D. Under L, P. kuhlii also flew at significantly lower altitude (i.e. away from a floodlight) than under D. Whereas P. kuhlii foraged both in L and D, E. bottae only foraged in D. In L, activity of E. bottae decreased and it merely transited the illuminated area at commuting rather than foraging speed. Thus, under artificially lighted conditions the non‐desert synanthropic species may have a competitive advantage over the native desert species and may outcompete it for aerial insect prey. Controlling light pollution in deserts and keeping important foraging sites unlit may reduce the synanthropic species' competitive advantage over native desert bats.

Full text

Differential effects of artificial lighting on flight and
foraging behaviour of two sympatric bat species
in a desert
T. Polak
1
, C. Korine
1
, S. Yair
2
& M. W. Holderied
3
1 Mitrani Department of Desert Ecology, Jacob Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Midreshet Ben-
Gurion, Israel
2 Department of Evolution, Systematic and Ecology, Alexander Silberman Institute of Life Sciences, The Hebrew University of Jerusalem,
Jerusalem, Israel
3 School of Biological Sciences, University of Bristol, Woodland Road, Bristol, UK
Keywords
light pollution; desert bats; Eptesicus bottae;
flight behaviour; Pipistrellus kuhlii.
Correspondence
Marc Holderied, School of Biological
Sciences, University of Bristol, Woodland
Road, Bristol BS8 1UG, UK. Tel: +44 (0)117
331 8049; Fax: +44 (0)117 331 7985
Email: marc.holderied@bristol.ac.uk
Editor: Virginia Hayssen
Received 29 March 2010; revised 10 February
2011; accepted 17 February 2011
doi:10.1111/j.1469-7998.2011.00808.x
Abstract
Human habitation in deserts can create rich novel resources that may be used by
native desert species. However, at night such resources may lose attractiveness
when they are in artificially lit areas. For bats, attraction to such manmade
habitats might be species specific. In an isolated village in the Negev desert that is
known for its high bat activity we investigated the effects of artificial lighting on
flight behaviour of two aerial insectivorous bat species: Pipistrellus kuhlii, a non-
desert synanthropic bat, common in urban environments and Eptesicus bottae,a
desert-dwelling species. Using an acoustic tracking system we reconstructed flight
trajectories for bats that flew under artificial lights [Light treatment (L)] versus in
natural darkness [Dark treatment (D)]. Under L both P. kuhlii and E. bottae flew
significantly faster than under D. Under L, P. kuhlii also flew at significantly lower
altitude (i.e. away from a floodlight) than under D. Whereas P. kuhlii foraged both
in L and D, E. bottae only foraged in D. In L, activity of E. bottae decreased and it
merely transited the illuminated area at commuting rather than foraging speed.
Thus, under artificially lighted conditions the non-desert synanthropic species may
have a competitive advantage over the native desert species and may outcompete it
for aerial insect prey. Controlling light pollution in deserts and keeping important
foraging sites unlit may reduce the synanthropic species’ competitive advantage
over native desert bats.
Introduction
The past century and especially the last 40 years saw a sharp
increase in artificial lighting (Frank, 1988), and today a
significant and ever increasing portion of the Earth’s surface
is directly affected by electrical lighting (Cinzano, Falchi &
Elvidge, 2001). Impacts of lighting on animal behaviour are
best documented in two main respects (Longcore & Rich,
2004): (1) orientation ability can either be improved (e.g. ants,
Klotz & Reid, 1993) or impaired (e.g. newly hatched sea
turtles, Salmon, 2006); (2) species distribution patterns can be
modified directly by phototaxis (e.g. moths, Frank, 2006) or
indirectly by predators following phototactic prey (e.g. insec-
tivorous bats foraging at street lights, Rydell, 1991). Conse-
quently, prey respond to this increased predation pressure
(e.g. scorpions forage less under lit conditions, Skutelsky,
1996). Illumination can have subtle effects such as changing
territorial singing in birds (Bergen & Abs, 1997).
Most bats are nocturnal and may be affected by both
natural and artificial lights. One important conservation
issue is artificial lighting along bat commuting routes, which
has major instantaneous impact on bat activity (Stone,
Jones & Harris, 2009). The orientation ability of bats might
also be better at low light levels, because high levels can
impair bat vision (Fure, 2006). Bats may change their spatial
distribution pattern either as hunters following their prey
into the light (e.g. higher prey densities, Hecker & Brigham,
1999; higher prey vulnerability, Svensson & Rydell, 1998;
and larger prey, Rich & Longcore, 2006) or as prey fleeing
light to avoid their own predators (e.g. flying in shadows at
high levels of moonlight, Reith, 1982). These changes in
spatial patterns of prey abundance and predation pressure
can lead to shifts of foraging hours and duration (Fenton
et al., 1977), and changes of foraging routes (Reith, 1982).
Deserts habitats are resource limited and human irrigated
gardens and/or agriculture can provide valuable additional
resources for desert wildlife. Indeed, Bell (1980) described
significantly higher bat activity and diversity and also
significantly different insect communities in riparian versus
desert and scrub habitats. Bats that inhabit desert areas such
Journal of Zoology
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London 21
Journal of Zoology. Print ISSN 0952-8369

as the Negev desert might profit from settlements because
insect proliferation increases (Korine & Pinshow, 2004; see
also Feldman, Whitaker & Yom-Tov, 2000). At night
however, settlements are frequently lit, with light spilling
far into neighbouring pristine desert. Because many insects
are strongly attracted to lights (Bell, 1980; Rydell, 1992),
light may deplete neighbouring desert habitats of potential
prey (i.e. the ‘vacuum effect’, Eisenbeis, 2006), while gen-
erating local patches of prey superabundance (Salcedo et al.,
1995). Not all bat species found in the Negev desert take
advantage of such local prey patches to the same degree
(Korine & Pinshow, 2004). Hence the desert bat community
might change favouring more adjustable generalist species
(Arlettaz, Godat & Meyer, 2000). In addition, new synan-
thropic species might enter desert habitats following human
habitations (Korine & Pinshow, 2004) thereby increasing
exploitation competition.
We studied the effect of artificial lighting on the flight
behaviour of two insectivorous bat species, which were
chosen because they are associated with human activities in
deserts to a different degree: Kuhl’s Pipistrelle Pipistrellus
kuhlii (Kuhl, 1817) is a small bat of 5 g and Botta’s
Serotine Eptesicus bottae (Peters, 1869) is slightly larger at
9 g (Korine & Pinshow, 2004). Pipistrellus kuhlii is a
synanthropic species that has expanded its distribution
range, from the Mediterranean into deserts, following hu-
man settlements (Yom-Tov & Kadmon, 1998; Mendelssohn
& Yom-Tov, 1999). It often is the only Microchiropteran
species in city centres and roosts mainly in man-made
structures (Mendelssohn & Yom-Tov, 1999). In contrast,
E. bottae in Israel is found in desert habitats, it does not
roost in man-made structures and is not as common in man-
made habitats in the Negev desert (Korine & Pinshow,
2004). Both species belong to the same foraging guild of
aerial insectivores, they co-occur in the study area and
forage in the same habitats including around street lamps,
where insect prey is abundant (Korine & Pinshow, 2004).
Thus, one species is a native desert bat taking advantage of a
new resource, and the other is a new arrival already well
adapted to exploiting such man-made patches of prey super-
abundance. We hypothesize that changes in lighting level
affect flight behaviour in both species and predict that such
behavioural changes will be immediate and reversible. Based
on differences in the degree of synanthropism of these two
species, we further hypothesize that P. kuhlii will be better
than E. bottae to take advantage of artificial lights as a
foraging resource. We predict (1) that P. kuhlii will be more
likely to forage under artificial lights; (2) that E. bottae will
show more prominent changes in its foraging behaviour.
Methodology
Research area
The research was conducted at Midreshet Ben-Gurion in the
Negev desert, Israel in July 2008. The experimental site was
located at the Ben Gurion tomb area (30150052.9700 N,
34146055.4500 E). The area consists of a park and an open
square with several trees. The tomb is lit every night from
dusk until c.11PM by two strong floodlights that are
positioned 7.6 m above the ground on the roof of an
adjacent building. Recording equipment was located at a
known flight corridor (Grodzinski et al., 2009) between trees
and an artificial wall just below the floodlights hence
recordings covered an area of maximum illumination levels.
This was the location for both Light and Dark treatments.
Acoustic tracking setup and analysis
We used two arrays of four Knowles BT1759 microphones
each to record the bat’s calls and obtain their spatial
position. The four microphones in each array were arranged
in one plane as a symmetrical triangular star with one
microphone in the centre with 58 cm distance between
microphones. Distance between arrays was 5 0.1 m
(mean SD). Both arrays were facing 451upwards and
towards the corridor but pointed slightly inwards such that
their survey volumes overlapped. This overlap determined
the three-dimensional (3D) space in which the bats’ flight
trajectories could be reconstructed (tracking range). The
tracking range did not cover individuals foraging directly in
front of the floodlights, but rather those in a lit area more
than 5 m from them. Arrays were aligned using a Stanley 77-
153 CL2 FatMax Cross Line Laser (Stanley, Sheffield, UK).
Distance between the arrays, height above ground and their
distance from certain objects were measured using a Bosch
DLE 50 Laser Rangefinder (Bosch, Uxbridge, UK). Calls
were sampled at 500 kHz with 11-bit resolution on a custom-
made digital recorder. We manually started a recording
every time a passing bat appeared on an amplitude display.
A four second pre-trigger compensated for the reaction time
of the experimenter. Individual recordings were ended
manually when all bats had left the tracking range. The
custom-made software BatSonar (v1.0, Aubauer & Holder-
ied, Darmstadt, Germany), was used to derive the bats’ 3D
flight trajectory from the eight microphones’ recordings. The
bat’s position at the moment of calling can be derived from
the time of arrival differences of the bat’s call between these
microphones. Because bats call about 10 times every second,
the individual localizations string together like a pearl chain
tracing the bat’s flight trajectory. The potential length of
trajectories is not limited, but it is impossible to decide
whether two consecutive trajectories are from two indivi-
duals or from one individual that left and then reentered the
tracking range. Maximum tracking errors are between 0.2
and 2% of the distance from the arrays. For further details
on methodology see Holderied & von Helversen (2003).
We calculated three flight variables from each trajectory:
(1) mean flight speed per trajectory (m s
1
) as mean of all
segment speeds within the trajectory, that is, the spatial
distances between pairs of consecutive localizations divided
by the corresponding time periods between calls. Trajec-
tories of nindependent localizations provided n1 indepen-
dent flight speed measurements; (2) mean flight height per
trajectory (m), the mean of all independent individual height
measurements within a trajectory relative to the ground; (3)
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London22
Light pollution in deserts and bat foraging T. Polak et al.

tortuosity per trajectory (an indicator of straightness of the
trajectory, Grodzinski et al., 2009) calculated as the overall
spatial distance (sum of call-to-call distances) divided by the
displacement (straight spatial distance from first to last
localization in trajectory). Analysis was done in Matlab v.
6.5 (The MathWorks, Natick, MA, USA).
Bat species were identified based on mean call end
frequency of all calls per trajectory. The two bat species in
this study have no overlap in search call end frequencies (P.
kuhlii 40.8 0.8 kHz, reanalysed after Berger-Tal et al.,
2008, and E. bottae 32.5 0.87 kHz, Holderied et al., 2005).
No other species were recorded in the tracking range, but
Tadarida teniotis and Taphozous perforatus flew at much
greater height in the area.
Experimental design
We manipulated lighting in the experimental area by blocking
both rooftop floodlights using cardboard screens
(henceforth called D as in Dark treatment). Floodlights were
left unobstructed during light treatment (L). L conditions were
at 12.77 1.93 lx and D at 0.27 0.058 lx, note that 0.2 lx was
the minimum sensitivity of our equipment (Gossen GO 4068
Mavo-Monitor light meter, Gossen, N ¨
urnberg, Germany); the
actual ambient light level in D was lower. The experiment was
run over four consecutive nights, for a 2-h period from
20:00–22:00 h. A mbient t emperature (P=0.07), wind speed
(P=0.59), and relative humidity (P=0.54) did not change
between the experimental nights. Sunset was at 19:35h and
moonrise always after 22:00 h and accordingl y ambient light
levels in D did not change within or between nights. Each
night was divided into four half hour periods in which we
recorded the bats’ behaviour for the first and the last 5 min,
that is, in eight 5-min intervals every evening. Light regimes:
night 1 had continuous L and night 4 continuous D. In nights
2 and 3, we alternated between treatments every 30min,
starting with L on night 2 and with D on night 3, resulting in
an inverse pattern between the two nights (compare Table 1).
Statistical analysis
Statistical analysis was performed on trajectories with seven
or more independent localizations to reduce problems with
unbalanced sample sizes and increase reliability of mean
flight variable measurements. Activity was measured as
number of flight trajectories within the 5-min intervals, and
we used w
2
-tests to test for effects of lighting treatment on
bat species. We tested for potential effects of the time of
night on flight behaviour for nights 1 and 4 (continuous L or
D) by linear regression. We used Student’s t-test to compare
flight variables in D and L. To test for differences between
short and medium term effects and any potential distur-
bance effects we grouped the 5-min interval data according
to the sequence of the change in lighting condition: The
categories were light after light (LAL), light after dark
(LAD), continuous light (LC) (entire night 1 and first
half hour of night 2), dark after light (DAL), dark after
dark (DAD), and continuous dark (DC) (night 4). We used
one-way ANOVA to test for differences in the flight vari-
ables between these six categories. In the flight data for
P. kuhlii, mean flight speed and mean flight height violated
the homogeneity of variance because one of the categories
had a larger variance in flight speed and another group
had smaller variance in flight height. Therefore, we per-
formed a randomization test (resampling, Manly, 2007).
We created 10 000 datasets by drawing randomly from the
original dataset (without replacement) and then conducted
one-way ANOVAs on all of the 10 000 datasets. This gave
us a probability mass function of F-values for the
randomized datasets. The placement of the F-value of our
original data in this distribution gave us the P-value.
The randomization tests confirmed the ANOVA results. The
differences between the categories were tested post hoc with a
randomization of mean differences (for method see descrip-
tion above – execution of the random database was the same
but we used mean differences between the categories instead
of F-values). The minimum significant level for all statistics
was Po0.05. Data are presented as mean SD.Randomiza-
tion tests were created specifically for our dataset in MATLAB v.
6.5. All other statistical analyses were performed in SPSS 13.0
(SPSS Inc., Chicago, IL, USA).
Results
Foraging activity
We recorded 437 trajectories of P. kuhlii and 199 trajectories
of E. bottae over four recording nights (Table 1). Pipistrellus
kuhlii activity (trajectories per minute) under L was lower than
under D (w
1
2
=4.63, P= 0.031, N
(Light)
=196, N
(Dark)
=241)
while E. bottae hadmuchloweractivityinLthaninD
(w
1
2
=140.15, Po0.001, N
(Light)
=16,N
(Dark)
=183).
Table 1 Experimental design and number of trajectories per species per sampling period: rows represent the four sampling nights with different
light regimes in four half hour segments (columns).
Time night
20:00–20:30 h
PK/EB
20:30–21:00 h
PK/EB
21:00–21:30 h
PK/EB
21:30–22:00 h
PK/EB
Per night
PK/EB
Night 1 21/0 7/0 5/0 25/0 58/0
Night 2 27/1 69/36 35/2 14/7 145/46
Night 3 27/33 29/2 29/11 47/11 132/57
Night 4 29/35 36/37 32/24 5/0 102/96
Grey backgrounds represent Dark phase periods; white backgrounds represent Light phase periods. Numbers of trajectories for the two species
Pipistrellus kuhlii (PK) and Eptesicus bottae (EB) are given for each half hour segment and for the total of each night.
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London 23
Light pollution in deserts and bat foragingT. Polak et al.

The dependence of flight behaviour on time
of night
On nights 1 and 4 (constant L or D) we found some
significant changes with time of night, yet they all had very
low coefficients of determination. This applied to flight
speed of P. kuhlii on nights 1 and 4 (r
2
=0.13, P= 0.004
and r
2
=0.076, P= 0.005, respectively) and flight height of
P. kuhlii and E. bottae on night 4 (r
2
=0.15, Po0.001 and
r
2
=0.07, P= 0.009, respectively). The overall effect of time
of night was always below 0.1%, which is negligible com-
pared with the effects of our treatments. We hence pooled
our data according to lighting conditions without correcting
for time of night.
Effects of lighting conditions on
flight behaviour
In P. kuhlii, all flight variables except tortuosity
were significantly different between lighting conditions.
Flight speed was significantly higher in L (Fig. 1a; L 7.51 
1.23 ms
1
vs. D 5.84 0.98 ms
1
,Po0.001; N
(Light)
=196,
N
(Dark)
=241), while flight height was lower in L (Fig. 1b; L
5.2 1.29 m vs. D 6.52 1.36 m, Po0.001, N
(Light)
=196,
N
(Dark)
=241). In E. bottae only flight speed was signifi-
cantly different between lighting conditions (Fig. 1c;
L 9.25 1.19 ms
1
vs. D 5.99 1.11 ms
1
,Po0.001,
N
(Light)
=16, N
(Dark)
=183). Flight height (Fig. 1d) and
tortuosity showed no significant differences.
Short-term effects of the change in
lighting conditions
Pipistrellus kuhlii flight speed (Fig. 2a) was significantly
higher in all three L categories (LC, LAL and LAD) than
in all D categories (DAL, DAD and DC) (one-way ANO-
VA: flight speed, Po0.001). Randomization post hoc tests
for these variables showed no differences (P40.05) among
the three L nor among the three D categories. However,
significant differences in speed were found between all L and
D categories (Po0.001). Flight height (Fig. 2b) displayed an
opposite trend, with D categories having higher values than
L categories (one-way ANOVA: Po0.001). LC differed
significantly in flight height from all others including the L
categories (randomization post hoc tests, Po0.05). The total
number of trajectories for E. bottae in L was considered too
small (L 16 vs. D 183 trajectories) for further statistical
analysis. Tortuosity was removed from this analysis because
it did not differ between L and D.
Discussion
We found that light level affected flight behaviour of both
bat species. Previous studies on artificial lighting found
positive effects on bat activity and/or abundance (Rydell,
1991; Svensson & Rydell, 1998; Ekl ¨
of & Jones, 2003; Stone
et al., 2009). Recent studies also show how artificial lights
can delay bat emergence (Boldogh, Dobrosi & Samu, 2007;
Stone et al., 2009) and reduce physiological condition
(Boldogh et al., 2007). Only Winter (1999) quantified effects
Speed (m / s)
0
4
5
6
7
8
9
10
Light Dark
Flight Height (m)
0
4
5
6
7
8
9
10
Speed (m / s)
0
4
6
8
10
12
Light Dark
Flight Height (m)
0
4
5
6
7
8
9
10
Pipistrellus kuhlii Eptesicus bottae
(a) (c)
(d)
(b)
Figure 1 Mean speed ( SD) and mean flight
height per trajectory in Light (open circles) and
Dark conditions (closed circles): (a) speed and (b)
flight height of Pipistrellus kuhlii; (c) speed and
(d) flight height of Eptesicus bottae.P-values
were o0.001 for light and dark for all comparis-
ons except for the flight height of E. bottae.
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London24
Light pollution in deserts and bat foraging T. Polak et al.

of artificial lights on bat flight speed (captive bats in flight
tunnels), and found a positive relationship between artificial
illumination and flight speed as in the present study.
Reith (1982) found that bats tend to fly lower and in tree
shadows under full moon, while Hecker & Brigham (1999)
found that bats flew higher when the moon was fuller. This
inconsistency might be due to differing foraging strategies of
different bat species (compare Rydell, 2006), yet species in
both studies belong to the same foraging guild of aerial
insectivores. Our finding of a decrease in flight height under
artificial lighting is in agreement with Reith (1982), yet levels
of ambient moonlight and artificial lighting might not be
fully equivalent.
Adaptive value of changes in
flight behaviour
Bats may react to changes in the light regime primarily
because of (1) predation risk; (2) spatial orientation abilities;
(3) prey availability in the area. First, because higher light
levels increase exposure to visual predators (e.g. owls), flying
faster could avoid capture. As bats are more manoeuvrable
than their nocturnal predators (Baker, 1962), flying closer to
obstacles might provide cover and block predator attacks.
Second, bats like Plecotus auritus use vision for prey capture
(Ekl ¨
of & Jones, 2003), and vision may improve a bats’
ability to orient (Fure, 2006). In our experiment, the addi-
tional lighting might have perceptually permitted both bat
species to fly faster and closer to the ground and vegetation.
Third, insect density tends to be higher under illumination
(Bell, 1980; Rydell, 1992; Salcedo et al., 1995). Many
nocturnal insects can hear bat calls and show evasive flight
manoeuvres (Triblehorn & Yager, 2005), yet street lights
seem to impair this avoidance behaviour making them more
vulnerable (Svensson & Rydell, 1998). Strong lights can also
impair insect vision while illuminated surfaces can improve
it (Frank, 1988). Nocturnal insects may use the same
predator avoidance strategies as bats and hence an increase
in flight speed could be adaptive for bats when foraging
around lights (Salcedo et al., 1995), yet slower more man-
oeuverable flights might be adaptive due to the increased
prey encounter rate. As insects might be attracted both to
the floodlight itself and to lit structures, predicting to what
degree bats should fly at the height of the lights or lower that
is closer to the illuminated structures is difficult.
Flight speed
Flight speeds were similar to those measured at the same site
previously. Grodzinski et al. (2009) found P. kuhlii foraging
at 6.74 1.16 ms
1
, which is between flight speeds in this
study, but without giving lighting levels. Mean foraging
speeds differed between foraging sites over a comparable
range (Grodzinski et al., 2009), which might partly be due to
lighting levels. As commuting speed for this species is
9.3 1.22 ms
1
(Grodzinski et al., 2009) we are confident
that all bats in our study were foraging. Holderied et al.
(2005) found E. bottae foraging speeds of 5.7 1.3 ms
1
,
which is very close to the speeds in D (5.99 1.11 ms
1
). In
L however, E. bottae flew much faster (9.25 1.19 ms
1
)
and close to its predicted commuting flight speed of
8.7 m s
1
(Holderied et al., 2005). This implies that E. bottae
was foraging only in D, but in L merely passed the area at
commuting speeds while presumably being present in adja-
cent darker areas. This is further corroborated by its drastic
drop in activity in L. Flight speed of P. kuhlii changed
equally in the long term (nights 1 and 4) and the short term
(nights 2 and 3) being instantaneous and reversible. Eptesi-
cus bottae showed the same long term pattern and had a
similar trend in the short term, which was however not
analysed due to its small sample size. The instantaneous
reversibility confirms that observed changes were caused by
changes in light regime and not by any potential disturbance
by the physical process of covering the floodlights.
Flight height and tortuosity
Pipistrellus kuhlii flew on average 41 m lower in L than in D
(Figs. 1 and 2), while E. bottae showed a similar yet
Speed (m / s)
0
4
5
6
7
8
9
10
LC LAL LAD DAL DAD DC
Flight Height (m)
0
4
5
6
7
8
9
10
a
aa
bbb
a
b
b
ccc
(a)
(b)
Figure 2 Mean ( SD) speed and flight height of Pipistrellus kuhlii
according to lighting categories: (a) flight speed and (b) flight height.
Lighting conditions are L, light (open shapes), D, dark (closed shapes),
LAL, light after light; LAD, light after dark; LC, continuous light; DAL,
dark after light; DAD, dark after dark; DC, continuous dark). Small
letters (a–c) represent the post hoc results and the differences
between the groups, different letters represent significant difference
between the groups of at least Po0.05.
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London 25
Light pollution in deserts and bat foragingT. Polak et al.

insignificant trend (Fig. 1). Flight height of P. kuhlii was
about 2.5 m below the floodlights in L and still about 1 m
below in D. Eptesicus bottae flew higher than P. kuhlii and
mainly at or just below the height of the floodlights. For
P. kuhlii short-term reactions were more gradual, with an
increase in height as conditions became darker (Fig. 2b).
This suggests it might be caused by a gradual change in
insect distribution.
Tortuosity did not differ between L and D, suggesting
that most trajectories were foraging flights with similar
straightness. The commuting flights of E. bottae in L were
not straighter than the foraging flights in D. This is surpris-
ing, but might be due to a general straight search pattern
during foraging in this species and also due to the low
sample size for L.
Insect distribution
We could not quantify how our treatment affected the prey
density perceived by bats, yet argue this was not causing the
observed behavioural changes because they were so immedi-
ate. The fact that most foraging flights did not include
attacks means suitable prey was not encountered at every
pass. Consequently, several passes are needed to notice
changes in density, yet the observed behavioural changes
were instantaneous. We cannot rule out however that bats
changed their behaviour in anticipation of lighting depen-
dent changes in insect distribution.
Experimental design
We are aware that general conclusions from a single location
study should be drawn with greatest caution. However, our
study site is one of the most important foraging site in about
a 50 km radius. The local behavioural changes can thus
affect desert bat community over a large region. The second
limitation is its short time span (4 days), yet weather and
ambient light did not differ between sampling nights, thus
did not bias our results. Also, our analysis is based on the
number of trajectories and not sampling nights. However,
this measure is prone to pseudoreplication, as individuals
may be recorded repeatedly.
In conclusion, we found marked differences between
the two species’ activity, flight behaviour and flight mode.
P. kuhlii used the study site under both lighting conditions
with slightly less activity in L. Flight speed and frequent
feeding buzzes indicate that this species foraged in both
lighting conditions. Eptesicus bottae avoided the foraging
site during L and foraged only in D. The finding that it
reappeared instantaneously in D implies it was always
present yet avoided the most brightly lit areas. Eptesicus
bottae indeed repeatedly intercepted insects in the compar-
able darkness several tens of metres from the floodlights.
These differences indicate that P. kuhlii forages in both
conditions while E. bottae does not. Because both species
are part of the same foraging guild (Korine & Pinshow,
2004), we conclude that the observed differences are due to
their level of synanthropism and respective light tolerance
and not to their flight abilities. Eptesicus bottae’s exclusion
from the lit and hence particularly valuable foraging sites
may give P. kuhlii a competitive advantage. Our results
concur with Korine & Pinshow (2004) who found that
artificial foraging sites, which include street lights, have
increased bat activity but only for two non-desert bat
species, P. kuhlii and T. teniotis.
The different behavioural responses to lighting condi-
tions may represent a more pressing conservation problem.
P. kuhlii is the most common insectivorous bat in Israel
(Mendelssohn & Yom-Tov, 1999). Its range expansion into
desert habitats (Mendelssohn & Yom-Tov, 1999; Korine &
Pinshow, 2004) and its competitive advantage over local
desert species may change the bat community (compare a
possible competitive exclusion due to street-lamps, Arlettaz
et al., 2000). Our results suggest that controlling light
pollution in deserts is essential to grant native species access
to important foraging sites and to reduce possible exploita-
tion competition by non-desert synanthropic species.
Acknowledgements
This research held in the Mitrani Department of Desert
Ecology (MDDE) was supported by the Blaustein Center
for Scientific Cooperation (to M.W.H.) and the Office of
Donor and Associates Affairs, BGU; The Open University
of Israel, Tel Aviv University and The Zoological Society of
Israel. We thank Avisoft Bioacoustics for continuous sup-
port. This is paper number 728 of the MDDE.
References
Arlettaz, .R., Godat, S. & Meyer, H. (2000). Competition for
food by expanding pipistrelle bat populations (Pipistrellus
pipistrellus) might contribute to the decline of lesser horse-
shoe bats (Rhinolophus hipposideros). Biol. Conserv. 93,
55–60.
Baker, J.K. (1962). The manner and efficiency of raptor
depredations on bats. Condor 64, 500–504.
Bell, G.P. (1980). Habitat use and response to patches of prey
by desert insectivorous bats. Can. J. Zool. 58, 1876–1883.
Bergen, F. & Abs, M. (1997). Verhaltens ¨
okologische Studie
zur Gesangsaktivit ¨
at von Blaumeise (Parus caeruleus),
Kohlmeise (Parus major) und Buchfink (Fringilla coelebs)
in einer Großstadt. J. Ornithol. 138, 451–467.
Berger-Tal, O., Berger-Tal, R., Korine, C., Holderied, MW.
& Fenton, M.B. (2008). Echolocation calls produced by
Kuhl’s pipistrelles in different flight situations. J. Zool. 274,
59–64.
Boldogh, S., Dobrosi, D. & Samu, P. (2007). The effects of the
illumination of buildings on house-dwelling bats and its
conservation consequences. Acta Chiropt. 9, 527–534.
Cinzano, P., Falchi, P.F. & Elvidge, C.D. (2001). The first
World Atlas of the artificial night sky brightness. Mon.
Not. Roy. Astron. Soc. 328, 689–707.
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London26
Light pollution in deserts and bat foraging T. Polak et al.

Eisenbeis, G. 2006. Artificial night lighting and insects:
attraction of insects to streetlamps in a rural setting in
Germany. In Ecological consequences of artificial night
lightning: 281–304. Rice, C. & Longcore, T. (Eds). Wa-
shington: Island Press.
Ekl ¨
of, J. & Jones, G. (2003). Use of vision in prey detection by
brown long-eared bats, Plecotus auritus.Anim. Behav. 66,
949–953.
Feldman, R., Whitaker, J.O. & Yom-Tov, Y. (2000). Dietary
composition and habitat use in a desert insectivorous bat
community in Israel. Acta Chirop. 2, 15–22.
Fenton, M.B., Boyle, N.G.H., Harrison, T.M. & Oxley, D.J.
(1977). Activity patterns, habitat use, and prey selection by
some African insectivorous bats. Biotropica 9, 73–85.
Frank, K.D. (1988). Impact of outdoor lighting on moths.
J. Lepid. Soc. 42, 63–93.
Frank, K.D. (2006). Effects of artificial night lighting on
moths. In Ecological consequences of artificial night light-
ning: 305–344. Rich, C. & Longcore, T. (Eds). Washington:
Island Press.
Fure, A. (2006). Bats and lighting. Lond. Nat. 85, 1–20.
Grodzinski, U., Spiegel, O., Korine, C. & Holderied, M.W.
(2009). Context-dependent flight speed: evidence for ener-
getically optimal flight speed in the bat Pipistrellus kuhlii?
J. Anim. Ecol. 78, 540–548.
Hecker, K.R. & Brigham, R.M. (1999). Does moonlight
change vertical stratification of activity by forest-dwelling
insectivorous bats? J. Mammal. 80, 1196–1201.
Holderied, M.W. & von Helversen, O. (2003). Echolocation
range and wingbeat period match in aerial-hawking bats.
Proc. Roy. Soc. Lond. B. 270, 2293–2299.
Holderied, M.W., Korine, C., Fenton, M.B., Parsons, S.,
Robson, S. & Jones, G. (2005). Echolocation call intensity
in the aerial hawking bat Eptesicus bottae (Vespertilioni-
dae) studied using stereo videogrammetry. J. Exp. Biol.
208, 1321–1327.
Klotz, J.H. & Reid, B.L. (1993). Nocturnal orientation in the
black carpenter ant Camponotus pennsylvanicus (DeGeer)
(Hymenoptera: Formicidae). Insectes Soc. 40, 95–106.
Korine, C. & Pinshow, B. (2004). Guild structure, foraging
space use, and distribution in a community of insectivorous
bats in the Negev Desert. J. Zool. 262, 187–196.
Longcore, T. & Rich, C. (2004). Ecological light pollution.
Front. Ecol. Environ. 2, 191–198.
Manly, B.F.J. (2007). Randomization, bootstrap and Monte
Carlo methods in biology, 3rd edn. Boca Raton: Chapman
& Hall.
Mendelssohn, H. & Yom-Tov, Y. (1999). Fauna Palaestina –
mammalia of Israel. Jerusalem: Keter Press.
Reith, C.C. (1982). Insectivorous bats fly in shadows to avoid
moonlight. J. Mammal. 63, 685–688.
Rich, C. & Longcore, T. (2006). Ecological consequences of
artificial night lighting. Washington: Island Press.
Rydell, J. (1991). Seasonal use of illuminated areas by fora-
ging northern bats Eptesicus nilssoni.Holarct. Ecol. 14,
203–207.
Rydell, J. (1992). Exploitation of insects around streetlamps
by bats in Sweden. Funct. Ecol. 6, 744–750.
Rydell, J. (2006). Bats and their insect prey at streetlights. In
Ecological consequences of artificial night lighting: 43–60.
Rich, C. & Longcore, T. (Eds). Washington: Island Press.
Salcedo, H.D., Fenton, M.B., Hickey, M.B.C. & Blake, R.W.
(1995). Energetic consequences of flight speeds of
foraging red and hoary bats (Lasiurus borealis and Lasiurus
cinereus; Chiroptera: Vespertilionidae). J Exp. Biol. 198,
2245–2251.
Salmon, M. (2006). Protecting sea turtles from artificial night
lightning at Florida’s oceanic beaches. In Ecological con-
sequences of artificial night lightning: 141–168. Rice, C. &
Longcore, T. (Eds). Washington: Island Press.
Skutelsky, O. (1996). Predation risk and state-dependent
foraging in scorpions: effects of moonlight on foraging in
the scorpion Buthus occitanus.Anim. Behav. 52, 49–57.
Stone, E.L., Jones, G. & Harris, S. (2009). Street lighting
disturbs commuting bats. Curr. Biol. 19, 1123–1127.
Svensson, A.M. & Rydell, J. (1998). Mercury vapour lamps
interfere with the bat defence of tympanate moths (Oper-
ophtera spp.; Geometridae). Anim. Behav. 55, 223–226.
Triblehorn, J.D. & Yager, D.D. (2005). Acoustic interactions
between insects and bats: a model for the interplay of
neuronal and ecological specialisations. In Ecology of
predator–prey interactions: 77–92. Barbosa, P. & Castella-
nos, I. (Eds). Oxford: Oxford University Press.
Winter, Y. (1999). Flight speed and body mass of nectar-
feeding bats (glossophaginae) during foraging. J. Exp. Biol.
202, 1917–1930.
Yom-Tov, Y. & Kadmon, R. (1998). Analysis of the distribu-
tion of bats in Israel. Divers. Distrib. 4, 63–70.
Journal of Zoology 285 (2011) 21–27 c2011 The Authors. Journal of Zoology c2011 The Zoological Society of London 27
Light pollution in deserts and bat foragingT. Polak et al.