INSIGHTS | PERSPECTIVES
rate two-dimensional optical fields, which
comfortably allows us to reconstruct even
complex scenes of our three-dimensional
environment. In vision, two-dimensional
object directions are directly accessible,
whereas distances have to be reconstructed
through correlation or accommodation. By
contrast, using echolocation, bats only receive two one-dimensional acoustic signals
as echoes to their echolocation calls, which
are not emitted continuously, but often
stroboscope-like at a high repetition rate.
Echo delays directly encode reflector distances, but the corresponding two-dimensional directions of each reflector must
be reconstructed through more advanced
mechanisms (3, 4).
The mammalian inner ear transforms
the incoming acoustic signal into a blurred
field of spectral information over time, in-
herently trading frequency versus time res-
olution. Scientists have suggested multiple
sophisticated strategies through which bats
can build a three-dimensional perception
based on only two one-dimensional input
signals (5). These strategies include the use
of spectral cues (6, 7), sequential scanning
(8), and the use of the directivity of their
sonar system (9). In simple environments,
these strategies allow unambiguous recon-
struction of three-dimensional reflector po-
sitions. However, in complex scenes, such
as when navigating through the foliage of
trees (see the photo), hundreds of short
echo glints—discernible single reflections,
for example, of a single leaf—superimpose
virtually at the same time, and matching all
corresponding echo glints from both ears is
impossible; ambiguities are inevitable. It is
therefore not feasible to create a complete
three-dimensional perceptional reconstruc-
tion of complex scenes of the real environ-
ment purely from echolocation.
Because of this difficulty, bats are forced
to also use innate knowledge and experience
when interpreting echo-acoustic scenes. This
makes bats prone to sensory deceptions, for
example, if they encounter a previously unexperienced complex environmental scene
that echo-acoustically closely resembles a
different, frequently experienced, and previously unambiguous situation.
Greif et al. now show that vertical mir-
roring structures, which have been built
abundantly into the bats’ environment as
windows and glass surfaces, are such sen-
sory traps for echolocating bats and could
thus pose a serious threat. Horizontal mir-
roring structures have been regularly pres-
ent in the bat environment as flat water
surfaces, and bats thus perceive them as
water surfaces (10). But vertical, perfectly
flat mirroring structures rarely occur in
the natural environment. Because such
structures reflect sound away at the same
angle as the incident sound, no conspicu-
ous echo returns from the central area of
the mirroring structure as long as the bat
is outside the space perpendicularly in
front of the mirror. In a complex environ-
ment, this acoustic behavior is similar to
an unobstructed area, because most natu-
ral structures reflect at least a weak echo
back to the echolocating bat. Bats may
wrongly interpret this situation as an un-
obstructed flyway (see the figure).
Greif et al.’s findings indicate that even
if the bats receive perpendicular echoes
from the vertical plane at the last moment,
they frequently do not avoid the obstacle
or can no longer prevent collision. The authors merge different approaches to prove
this deception: In field experiments, they
demonstrate the presence of this sensory
deception for three bat species. In lab experiments, they analyze flight behavior to
better understand the mechanism of the
misinterpretation. Through examination
of the acoustical properties of the scene,
they verify and visualize the acoustic background of this situation.
Having proven the existence of this sensory trap in bat perception in the laboratory
and in field experiments, Greif et al. suggest
more frequent monitoring of relevant locations to assess the importance of this trap,
because potential fatalities caused by it
may have been missed in surveys that were
unaware of this problem. Given the abundance of vertical mirroring structures like
windows or glass-front buildings built by
humans into the bat environment in recent
decades, it would be desirable to determine
whether these structures represent a relevant ecological threat, how well bats can
learn to deal with this threat, and whether
potential casualties can be avoided. j
1. D. R. Griffin, Listening in the Dark (Yale Univ. Press, 1958).
2. S.Greif,S.Zsebők, D.Schmieder, B.M.Siemers, Science
357, 1045 (2017).
3. H.-U. Schnitzler, C. F. Moss, A. Denzinger, Trends Ecol.Evol .
18, 386 (2003).
4. A.Surlykke, J.A.Simmons, C.F.Moss,in Bat Bioacoustics,
M. B. Fenton, A. D. Grinnell, A. D. Popper, R. R. Fay, Eds.
5. C. F. Moss, A. Surlykke, Front.Behav.Neurosci. 4, 33
6. H.Peremans,J.Hallam, J.Acoust.Soc.Am.104,1101
7. J.A.Simmons, J.E.Gaudette, IE T Radar Sonar Nav.6,556
8. A. Surlykke, K. Ghose, C. F. Moss, J. Exp. Biol. 212, 1011
9. M. Aytekin, E. Grassi, M. Sahota, C. F. Moss, J.Acoust.Soc.
Am. 116, 3594 (2004).
10. S. Greif, B. M. Siemers, Nat. Commun. 1, 107 (2010).
Vertical mirroring structures
Because there are no vertical, perfectly smooth surfaces in
their natural environment, bats tend to interpret ambiguous
situations with vertical mirroring structures as open fyways
and may collide with them.
Bats encounter horizontal
mirroring surfaces in
their natural environment
and interpret them as
978 8 SEPTEMBER 2017 • VOL 357 ISSUE 6355
How bats are deceived
Most natural structures, like foliage, return some echo back to the echolocating bat. Plane mirroring surfaces
like ponds or window panes reflect the echo away at incident angle. Bats cannot discern if the sound is just
traveling away or if it is reflected away by such a mirror.