sidan har flyttat till: http://www.fladdermus.net/thesis.htm
Eklöf, J. Vision in echolocating bats
the page has moved to http://www.fladdermus.net/thesis.htm
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
copybat.webb - www.fladdermus.net
Eklöf, J. Vision in echolocating bats
Zoology Department, Göteborg University
För Svensk sammanfattning, klicka här
Key words: acoustic clutter, foraging tactics, Microchiroptera, perception,
sensory ecology, ultrasound, visual acuity
ABSTRACT
The use of ultrasonic echolocation (sonar) in air is seriously constrained
by the attenuation of high frequency sounds and unwanted echoes from the background
(called clutter). Therefore, in many situations, echolocating bats have to rely
on other sensory cues. The aim of this thesis is to investigate the use of vision
by echolocating bats.
Bat eyes are generally small, especially among aerial hawking insectivores,
with the exception of members of the family Emballonuridae. In gleaning, and
in frugivorous species, however, the eyes tend to be larger and more prominent.
The eyes of all bats are well adapted to low illumination, having mainly rod-based
retinas, large corneal surfaces and lenses, and generally large receptor fields.
Bats can easily detect small differences in brightness on clear nights, and
the visual acuity remains relatively good in dim illuminations. The visual resolving
power (as obtained from counts of retinal ganglion cells or by optomotor response
tests) varies considerably among the different species of bats, from less than
0.06° of arc in Macrotus californicus (Phyllostomidae) to almost 5°
in aerial hawking Myotis species (Vespertilionidae). Generally, the visual acuity
is similar to that of rats and mice, suggesting that cm-sized object can be
discriminated at ranges less than a few metres. Studies on pattern discrimination
have yielded highly variable results. Fruit and nectar eating species respond
to patterns to a larger extent than aerial insectivores.
One of the most fundamental roles of the eyes is to register the amount of
ambient light, in order to establish photoperiodic cycles. Some tropical bats
avoid too bright conditions, i.e. moonlit nights probably due to increased predation
risk, a behaviour not found in high latitude species.
As sonar only works well at short ranges, vision is primarily used for detection
of landmarks and to avoid objects when moving over long distances, for example
during seasonal migration and when commuting between feeding sites. In these
situations, there seems to be precedence of vision over sonar. At short range,
within that of echolocation, bats may defer to visual cues in addition to sonar
and spatial memory to solve different tasks of orientation, especially when
there is conflicting information. Light conditions and time of the day may determine
the behaviour of the bats and thus which sensory cues will be used.
There is an increasing amount of data suggesting that vision might be of importance
in some situations and some aspects of foraging, especially for frugivorous
and nectarivorous bats, which can make use of differences in brightness and
spectral composition, to find different food items. But even in species traditionally
considered to rely heavily on echolocation, such as most insectivorous bats,
vision seems to play a more important role than has been recognised previously.
The gleaning brown long-eared bat (Plecotus auritus, Vespertilionidae), known
to forage mostly by using passive listening, detects prey more readily by using
vision than by using sonar, and the aerial hawking northern bat (Eptesicus nilssonii,
Vespertilionidae), use visual information in addition to sonar to find large
stationary prey in clutter.
Although echolocation is the key innovation that have made it possible for bats to fly at night, vision is retained as an important complement; and indeed bats use an array of different sensory inputs to solve the different tasks of life.
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Bats (Order: Chiroptera) are among the most diverse and abundant mammals on earth and the thousand or so species comprise about one fourth of all mammalians. Bats occur throughout the world, except the Polar Regions, and show a remarkable wide range of habitat use, behaviour, morphology, and diet. Most bats feed on insects but there are also bats that feed on fruit, nectar, fish, small vertebrates, and blood. Bats are the only mammals that have evolved active flight, and they can navigate through complete darkness by using echolocation or sonar (SOund Navigation And Ranging). Bats live almost everywhere, in tropical jungles as well as in cities; they inhabit caves, trees, houses, churches, bridges, coiled banana leafs, bamboo canes, and some species even build their own tents by using large leaves. Bats have a remarkable spatial memory and are quick learners. They can form colonies of up to 20 million individuals, eat hundreds or thousands of insects every night and migrate across continents. Many bats hibernate through a cold winter and some can reach more than 40 years of age. Despite this, bats are seldom people’s number one choice of favourite animal. Instead, bats have become symbols of darkness, doom and occultism in the western world. They often appear in not so flattering contexts, such as in myths, scary movies, heavy metal lyrics, and are often one of the most important ingredients in witches’ brews. Being called an old bat is not a compliment, and having a bat in one’s belfry is not very often socially accepted. In the eastern world, however, bats are considered as symbols of fortune and a long, prosperous life. Nevertheless, the bats’ leathery wings and their ability to navigate through the night are presumably two reasons behind their often somewhat scary reputation, as well as the two main reasons behind their success as a group. But how do they perceive the world; or as Thomas Nagel (1974) put it in his classic paper: what is it like to be a bat?
All information about the surrounding world is filtered through our senses and processed in our brains in order to give us just the right kind and amount of information to help us make proper decisions. This is true for all animals, although the senses receiving the information and the brains that process it differ across the animal kingdom. The type and amount of information that is needed obviously varies considerably depending on life style. What humans cannot perceive tend to be called ultra-, infra-, or extra-something. We do not know what ultraviolet light looks like, only that it gives us a nice tan. We cannot hear infrasounds although elephants can, which is why we and not the elephants invented the telephone. Many animals live entirely in the world of ultra-, or infra-, making it hard for us to relate to their every day life, or as Thomas Carlyle (1837) elegantly put it: “In every object there is inexhaustible meaning; the eye sees in it what the eye brings means of seeing. To Newton and to Newton’s dog Diamond, what a different pair of universes.” Bat echolocation is different from any of the senses that we are familiar with, and therefore, we cannot know or even imagine how they experience the world; or as in the words of Thomas Nagel (1974): “Anyone who has spent some time in an enclosed space with an excited bat knows what it is to encounter a fundamentally alien form of life.” With this in mind, it may seem impossible to study sensory ecology, and still, we try. We accept that there is information outside our perception range, although we will never be able to fully understand those things. We may perhaps be able to understand how a bat collects and uses information from the environment, but never what this really is like for the bat. We may however, from a human point of view, describe behaviour and reactions of animals under defined conditions. For example, when studying bats flying, and recording and describing echolocation calls, we can tell that sonar is a high precision tool, as good as vision for perceiving and identifying objects, only entirely different. But we begin to understand that echolocation alone is not enough to fully experience the world as a bat. As high frequency sounds attenuate rapidly in air, the effective range of echolocation is limited to a few metres in practice. Background echoes, known as clutter, also impose severe constraints on the use of sonar, and for a bat to perceive distant objects or objects hidden in vegetation, other senses must be used. One of these senses is vision. However, looking at a typical bat eye gives little hope of any breathtaking visual adventures. The eyes are often small and inconspicuous, especially compared to the more fanciful ears and nose-leafs of many bats, and considering our own sensory limits and the fact that bats fly at night, it is not hard to imagine why an expression like “as blind as a bat” exists. But still, bats do have eyes and perhaps “as blind as we would be if we had bat eyes” would be a more suitable expression. As I will discuss in this thesis, bats do have eyes that function for bats. In the same way humans have a sense of smell that function for humans, although a dog, or a bat for that matter, probably would not be impressed!
The aim of this introductory chapter is to put my work into perspective by
summarizing current knowledge of the role of vision in the lives of the echolocating
bats.
Footnote 1. The Microchiroptera includes ca 800 species of echolocating bats
but excludes the generally non-echolocating Megachiroptera or flying foxes,
which are not considered in this thesis.
VISION IN ECHOLOCATING BATS
The microchiropteran eye
The eyes of Microchiroptera1 rank among the smallest in mammals (Tab 1), although there are considerable differences in both eye size and morphology across species, reflecting a great ecological diversity (Chase 1972; Hope & Bhatnagar 1979a; b; Marks 1980; Suthers & Bradford 1980; Bell & Fenton 1986; Paper IV). In general, the eyes of frugivorous and nectarivorous Microchiroptera are larger than those of insectivorous species. Bats roosting in relatively exposed sites, and those that sometimes are active in dusk- and daylight conditions such as many members of the family Emballonuridae also have relatively large eyes. Hence eye size seems to reflect how much bats are exposed to light in their daily life.
The microchiropteran eyes are shaped for nocturnal conditions in that they
have large corneal surfaces and lenses relative to the size of the eye. They
also have relatively large receptor fields, which give them good light gathering
power at the expense of acuity, i.e. the ability to resolve fine spatial details
(Suthers 1970; Suthers & Wallis 1970). The bat retina, which is relatively
thin (91-126 µm) compared to that of voles (178 µm) and rats (198
µm), for example, consists mainly of rods, which are arranged loosely
in visual streaks (Chase 1972; Marks 1980; Pettigrew et al. 1998). However,
cones or at least cone like structures (receptor cells with pedicles) are present
at least in the fruit-eating bats Artibeus lituratus and Phyllostomus hastatus
(Phyllostomidae) and the aerial hawking insectivorous Saccopteryx bilineata,
Saccopteryx leptura and Rhynconycteris naso (Emballonuridae) (Suthers 1970;
Chase 1972).
Suthers and Wallis (1970) studied the eyes of two species of Vespertilionidae
(Myotis sodalis and Pipistrellus subflavus) and four species of Phyllostomidae
(the vampire bat Desmodus rotundus, and the fruit-eating Carollia perspicillata,
Anoura geoffroyi and Phyllostomus hastatus), and concluded that the visual capabilities
of all the species tested would allow the bats to see well at ranges beyond
that of echolocation. Due to the more or less spherical lenses (small species
tend to have more asymmetric lenses; Chase 1972), it also follows that most
Microchiroptera have a short focal distance and hence a great depth of focus
(Suthers & Wallis 1970). In fact, microchiropteran bats seem to be farsighted,
indicating that vision is used predominantly at long ranges, which is where
echolocation does not work so well.
The brain and the retinal pathways
The relative size of the internal brain structures of bats differs between
insectivorous, sangivorous and carnivorous species on one hand and frugivorous
and nectarivorous species on the other (Jolicoeur & Baron 1980; Barton et
al. 1995; Barton & Harvey 2000). Whereas insect eating bats have enlarged
echo-acoustic brain structures, fruit eating species have relatively large olfactory-
and visual bulbs, clearly reflecting the different feeding strategies in the
various species.
The size and lamination of the main targets of retinal projections in the brain:
the superior colliculus (which transmits visual information and controls head-
and eye movements) and the lateral geniculate body (a processing station on
the way from the retina to the visual centre, which e.g. serves to enhance contrasts)
have been studied in Artibeus, Eptesicus (Cotter 1985), Myotis (Cotter &
Pentney 1979; Crowle 1980) and Pteronotus (Covey et al. 1987). Megadermatids
and fruit eating phyllostomids show the thickest and most developed layers in
the superior colliculus, at least in the superficial ones, which receive exclusively
visual input. Also gleaning species tend to have relatively large superior colliculi.
Open-air insectivorous species on the other hand, seem to have superior colliculi
consisting almost entirely of the deeper layers, which receive a variety of
different sensory inputs (including visual stimuli). However, some insectivorous
bats, like the Emballonuridae (especially Saccopteryx and Cyttarops) have relatively
large superior colliculi and resemble frugivores in this respect, although their
total brain volume is smaller than in most other microchiropteran families (Baron
et al. 1996b). This may perhaps reflect the fact that most emballonurid species
roost in exposed sites and therefore live in bright light conditions. However,
considering that the Emballonuridae form a basal clade in the phylogenetic tree,
it may just as well suggest that bat ancestors had a well-developed visual system.
(Simmons & Geisler 1998). The projections to the superior colliculus are
similar to those of most mammals, in that they have no binocular overlap, and
thus the left superior colliculus receives input only from the right eye and
vice versa (Pettigrew 1986; Neuweiler 2000). In Megachiroptera and in primates,
both superior colliculi receive input from both eyes, and hence these animals
have better stereoscopic vision than Microchiroptera. (Interestingly the microchiropteran
family Rhinolophidae, which contain highly specialized echolocators, show similarities
to Megachiroptera in this respect (Reimer 1989). This may reflect phylogenetic
relationship rather than visual adaptation, however (e.g. Springer et al. 2001)).
The lateral geniculate body consists of two parts, the ventral lateral geniculate,
which has connections with several other brain structures, and the dorsal lateral
geniculate, which connects to the visual cortex. In most Microchiroptera, a
larger proportion of the nerves are projected to the ventral side of the lateral
geniculate body, suggesting that vision is important for orientation rather
than for cognitive tasks (Neuweiler 2000). However, the sizes of the retinal
pathways vary between genera. The nerves are generally larger in frugivores
(Phyllostomus hastatus, Anoura geoffroyi, Suthers & Bradford 1980; and Artibeus
jamaicensis, Cotter 1985) than in insectivores (Eptesicus fuscus, Cotter 1985;
and Pteronotus parnellii, Covey et al. 1987), although, again, insectivorous
Emballonuridae and Megadermatidae are exceptions. Both have relatively large
visual pathways projecting through the dorsal lateral geniculate to the visual
cortex. This suggests that vision is more important in these species, and they
show similarities to the visually oriented Megachiroptera in this respect (Neuweiler
2000), and may reflect phylogenetic relationship (Springer et al. 2001). For
a comparison of different brain structures between all groups of Microchiroptera,
see Baron et al. (1996a; b; c).
What bats can see
Brightness discrimination and light sensitivity
At the most basic level, vision is involved in the establishment of photoperiodic
cycles, and serves to distinguish daylight from darkness. It was previously
believed that this was the sole purpose of the microchiropteran eye (Eisentraut
1969 cited in Dietrich & Dodt 1970). The bat’s activity cycle is controlled
by an endogenous circadian rhythm, which is synchronized with the daylight cycle
by light sampling behaviour. This means that, before they emerge from the roost
to feed, the bats move from the darker areas in their roosts to lighter areas
near the entrance, in order to test the outdoor light level (Erkert 1982). Cloudiness
and moonlight can thus affect the time of emergence. On moonlit nights, many
tropical microchiropterans typically reduce their foraging activity, presumably
due to increased predation risk (Morrison 1978; Usman et al. 1980; Fleming 1988)
or perhaps lower availability of food (Lang et al. 2002). In contrast, bat activity
at high latitudes is not influenced by moonlight to any high extent (Paper VI).
On twelve nights in August-September 2000, the impact of moonlight on bat swarming
activity (associated with mating season) was studied at an abandoned mine in
southern Sweden. Bat activity at and near the mine entrance did not vary with
moon phase, or cloud cover, suggesting that moonlight had no effect on the bats’
behaviour. It seems likely that insectivorous bats at high latitudes may not
have been exposed to significant nocturnal predator pressure, leading to the
evolution of lunar phobia, as many tropical bats. In contrast to high-latitude
bats, the latter have to face specialized bat predators such as bat falcons
(Falco rufigularis). Furthermore, high latitude bats are exposed to relatively
bright light conditions throughout the summer. They do react to light, but not
by decreasing their activity, instead, they fly closer to protective vegetation
or sometimes high in the air (Rydell et al. 2002). This kind of behaviour is
also seen in species that migrate during the day, such as the noctule, Nyctalus
noctula (Ahlén 1997). Both types of behaviour may have the purpose of
avoiding predatory birds (e.g. small hawks and falcons).
The ability of bats to detect small differences in brightness, i.e. brightness
discrimination, was first studied by Eisentraut (1950), who found that Plecotus
auritus and Eptesicus serotinus (Vespertilionidae) could easily distinguish
black cards from white. Curtis (1952) trained the vespertilionids Eptesicus
fuscus and Myotis lucifugus to search for food at the illuminated end of a box,
and found that the bats’ ability of brightness discrimination is similar
to that of rats and mice. Brightness discrimination performance in Eptesicus
fuscus peaks around 10 lux, which is equivalent to the light level prevailing
at dusk and dawn, but remains good in illuminations as low as 0.001 lux, conditions
which resembles darkness to a human eye adapted to low light intensity. As a
comparison, a light level of 0.1 lux is equivalent to light levels at full moon,
and on overcast nights the amount of light drops to 0.0001 lux (Ryer 1997).
Based on focal distance and diameter of the dilated pupil, Dietrich and Dodt
(1970) calculated that the light gathering power of Myotis myotis is 4-5 times
that of man. This suggests that bats can readily use visual cues at dusk, when
they normally emerge from their roosts, and probably also under nocturnal conditions
(Ellins & Masterson 1974).
As may be expected from a retina consisting predominantly of rods, the visual
sensitivity generally declines as the ambient illumination increases towards
daylight (Hope & Bhatnagar 1979b). This indicates that the bat eyes work
better in dim light than in bright light. This has been verified behaviourally
by Bradbury & Nottebohm (1969), who found that Myotis lucifugus avoids obstacles
better under ambient illuminations resembling dusk, than they do in bright daylight.
These findings may explain why early studies, which were made in room illumination,
usually failed to prove any major visual capacity in microchiropteran bats (e.g.
Eisentraut 1950; Curtis 1952).
Light tolerance has been estimated in three species of Vespertilionidae (Myotis
myotis, Dietrich & Dodt 1970; Eptesicus serotinus, Bornschein 1961; and
Eptesicus fuscus, Hope & Bhatnagar 1979b) and three species of Phyllostomidae
(Desmodus rotundus, Carollia perspicillata, and Artibeus jamaicensis, Hope &
Bhatnagar 1979b) by measuring the luminance of light stimuli required to provoke
electroretinogram responses. Among the vespertilionids, Eptesicus fuscus showed
the highest light tolerance, and among the phyllostomids, which generally responded
to lower luminance levels than the vespertilionids, Artibeus jamaicensis showed
the highest tolerance. This presumably reflects the relative importance of vision
in the different species, but perhaps more importantly the time at which these
species normally emerge in the evening, and to what extent they are exposed
to bright light (Hope & Bhatnagar 1979a; b). The Emballonuridae Emballonura
spp. and Saccopteryx spp., some of which roost at exposed sites and often fly
in daylight (Lekagul & McNeely 1977; Bradbury & Vehrencamp 1976; Kalko
1995), would thus be expected to be more light tolerant than other bats. Although,
light tolerance levels have not been measured in these bats directly, the small
receptive fields and the low receptor-to-ganglion ratio (ca 1:10) in Saccopteryx
spp., compared to that of other microchiropteran species (ca 1:100), indicate
a high light tolerance and good resolving power as expected. In fact they resemble
diurnal mammals in this respect (Chase 1972). Nevertheless, the eyes of Microchiroptera
work well under low ambient illumination, although the sensitivity to different
light levels and the ability of brightness discrimination vary considerably
between the different families and species.
Spatial resolution
The eyes of microchiropterans are primarily adapted to function in low light levels. This carries the disadvantage of a relative poor ability to resolve fine spatial details (acuity). The ability of spatial resolution of the bat eye can be estimated either anatomically, by calculating the density of retinal ganglion cells (Marks 1980; Pettigrew et al. 1998; Heffner et al. 2001) or behaviourally, by presenting the bats with striped patterns of different fineness (Suthers 1966; Bell & Fenton 1986; Paper IV). When the visual acuity is measured with the latter method, it is often referred to as grating acuity and is expressed as degrees of arc or as cycles per degree, where one cycle is one pair of black and white stripes. The two methods give indications of the minimum separable angles, i.e. the minimum distance between two points that an animal needs in order to separate them.
Comparisons between the two methods should be treated carefully because the
acuity values estimated by counting retinal ganglion cells tend to be higher
than those estimated from behavioural studies. This suggests that the anatomical
method gives a theoretical minimum, rather than an indication of what the bats
actually respond to. Nevertheless, Table 2 should give an idea of the wide range
of spatial resolution ability that has been documented in different species
of microchiropteran bats, from the coarse vision of the small Myotis spp. (Vespertilionidae)
(3-5º arc, Paper IV) to the relatively fine visual ability of Macrotus
californicus (Phyllostomidae) (0.06° arc, Bell & Fenton 1986). Macrotus
californicus has by far the best resolving power found in any microchiropteran
bat studied so far, and is comparable to that of a dog in this respect (Heffner
& Heffner 1992). It is also the only microchiropteran known to be capable
of detecting insects, using vision alone (Bell 1985).
The visual resolving power is never a fixed value, but depends on the ambient
light intensity. In the common vampire bat Desmodus rotundus, for example, the
acuity drops from 0.8° arc at a light intensity of ca 310 lux to over 2°
arc in ca 0.004 lux (Manske & Schmidt 1976). Other bats, such as Macrotus
californicus (0.06° arc) and Antrozous pallidus (0.25° arc) retain their
visual acuity down to light levels as low as ca 0.002 lux (Bell & Fenton
1986). In comparison, species of Megachiroptera, which do not echolocate, has
been shown to respond to striped patterns of 0.8° in light levels of ca
0.0005 lux, whereas humans responds only to patterns of 1.3° arc under the
same conditions (Neuweiler 1967). Hence, in very dim light, bats can see better
than humans.
Pattern discrimination
Bats can visually distinguish patterns and shapes of objects. The nectarivorous
Anoura geoffroyi (Phyllostomidae) distinguishes rectangles from solid discs
of the same surface area, when trained to seek food at the discs (Suthers &
Chase 1966; Suthers et al. 1969). This species is also able to distinguish outlines
of erected triangles from inverted ones, as long as the baselines of the triangles
are intact. However, when the bats were presented with two sides of a triangle,
i.e. an outline of a triangle without a base, the shape was no longer distinguished
from other shapes. This indicates that Anoura geoffroyi does not possess a concept
of form, but rather perceive the relative position of horizontal lines. Similar
conclusions were drawn from studies of common vampire bats Desmodus rotundus
(Phyllostomidae). This species is able to separate vertical stripes but not
horizontal stripes from circles of the same area (Schmidt & Manske 1978;
Manske & Schmidt 1979). In contrast, the insectivorous species Vespertilio
superans (Vespertilionidae) cannot distinguish objects of different shapes but
equal size, and responds only to the size of the surface areas (Chung et al.
1990). The only bat that has been shown unambiguously to respond to shapes alone
is the frugivorous phyllostomid Carollia perspicillata. This species can discriminate
squares from circles, even if the squares are rotated (Suthers et al. 1969).
In conclusion, studies on pattern discrimination have yielded highly variable
results, but in general it seems as if fruit- and nectar-eating microchiropterans
respond to patterns and shapes more readily than insectivorous species. This
may perhaps reflect that plants are more easily detected by vision, and less
detectable by sonar than insects, and that frugivores therefore may use a different
search image when foraging.
Perception of colour
Given that microchiropteran bats are all more or less nocturnal, true colour vision seems unlikely to occur in these animals, as it would probably be of minor importance. Nevertheless, cones occur in the retinas of some species, although most authors report only rods (reviewed by Suthers 1970; Chase 1972). Nevertheless, there is evidence that at least two different photo pigments occur in the eyes of Microchiroptera (Chase 1972; Hope & Bhatnagar 1979a). Electroretinogram response tests have shown sensitivity peaks around 500 nm and 570 nm in the vespertilionid species Myotis myotis (Dietrich & Dodt 1970) and Eptesicus fuscus (Hope & Bhatnagar 1979a) and the phyllostomid species Artibeus jamaicensis, Desmodus rotundus and Carollia perspicillata (Hope & Bhatnagar 1979a). There is also preliminary evidence that there is a spectral sensitivity peak in the near UV-range (around 390 nm) in the nectarivorous phyllostomid Glossophaga soricina (Lopez et al. 2001). It is thus possible that this species is able to perceive ultraviolet light reflected from fruits and plants.
Vision in orientation and navigation
Long distance navigation
The fact that the eyes of most bats function better beyond than within the
range of echolocation (Suthers & Wallis 1970) suggests that visual cues
may preferably be used in preference to echolocation for navigation and orientation
over longer distances.
Several species of Microchiroptera make long distance movements and some even
perform seasonal migration (Griffin 1970). It seems unlikely that ultrasonic
echolocation plays any major role in orientation over long distances, as it
works only at short range. For example, insect sized targets can be detected
a few metres away at best (Kick 1982), although trees, hillsides or the ground
obviously may be detected much further away. However, even during the most favourable
conditions, bats do not pay attention to echoes returning from more than 100
m or so away (Altringham 1996) and therefore, migration over long distances
is almost certainly guided by other senses, including vision (Griffin 1970).
Bats can use distant low frequency sounds for orientation over moderate distances,
indicating that passive hearing may also be involved in navigation over longer
distances (Griffin 1970; Buchler & Childs 1981). There is also some evidence
that bats possess magnetic material (Buchler & Wasilewski 1985), but if
they possess a magnetic sense like birds (Wiltschko & Wiltschko 1995) or
not, is still unknown.
When migrating at night, it is possible that stars can serve as navigational
cues for some species of bats. For example, Eptesicus fuscus can see point light
sources, which simulate bright white and blue stars against the night sky, if
these are located at > 6° angle (Childs & Buchler 1981). This species
is also able to orient and navigate in relation to the post-sunset glow in the
west (Buchler & Childs 1982).
In homing experiments with bats released within 10 km from their roost, the
bats have been demonstrated to do well, using echolocation alone. This suggests
that they are acoustically familiar with a relatively large home territory (Williams
et al. 1966; Williams & Williams 1967; Davis & Barbour 1970). Nevertheless,
blinded bats tend to fly slower and closer to the ground than non-treated bats
(Mueller 1968), indicating that they change their orientation behaviour when
they no longer are able to see. Bats also seem to rely heavily on their spatial
memory, as they often follow the same paths night after night (Höller 1995).
When commuting between roost- and feeding sites during dusk and dawn periods,
bats often follow outlines in the landscape, such as river banks, forest edges,
hedgerows and hillsides (Racey & Swift 1985; Limpens & Kapteyn 1991;
Verboom & Huitema 1997). The reason may be to minimise predation risk (Swift
1998), or to use outlines as acoustic landmarks, which perhaps facilitate navigation
by sonar (Verboom et al. 1999). More likely, however, landscape outlines and
silhouettes provide the bats with visual cues, contrasting against the twilight
sky, and such cues are probably essential for orientation and navigation along
travelling routes (Davis 1966; Layne 1967; Griffin 1970; Manske & Schmidt
1979; Höller and Schmidt 1996).
The frequent observation that bats have a tendency to crash into windows of
buildings when released indoors (Fenton 1975), during migration (Timm 1988),
or commuting (Test 1967), suggests that they predominantly rely on vision rather
than on echolocation in situations when both acoustic and visual cues are available.
The performance is greatly improved, i.e. there are fewer collisions, when the
bats are blinded (Davis & Barbour 1965) or when they are flown under dark
conditions, and hence are “forced” to rely on echolocation alone.
The insectivorous Balantiopteryx plicata (Emballonuridae) was studied at different
times of the day in an empty mesh greenhouse (Paper V). At night they flew smoothly
and could easily avoid the ceiling and the walls of the greenhouse, but during
the day and at dusk and dawn they often tried to fly through the mesh and thereby
crashed into it. The bats used echolocation consistently and without any dramatic
change in echolocation call structure that could be related to the prevailing
light conditions. The study indicates that emballonurid bats trust their eyes
over their ears when exposed to contradictory auditory and visual cues.
Close range orientation and navigation
When moving towards resting places and specific sites within roosts, bats sometimes
face extremely unfavourable conditions for orientation, such as darkness, acoustic
clutter from the walls of the roost, and simultaneous echolocation calls from
many individuals. It is therefore likely that arrays of different sensory cues
are used in such situations, and also that a good spatial memory is of great
importance (Höller & Schmidt 1996). When introduced in a dark flight
cage, Nyctophilus spp. (Vespertilionidae) ceased to echolocate after 6-8 hours
of flight (Grant 1991), suggesting that they can learn to orient inside the
cage, using spatial memory alone. In the same way, Megaderma lyra (Megadermatidae)
remembers the positions of narrow openings with an accuracy of 2 cm, and if
an obstacle is removed from the flight path, the bats may continue to avoid
that position for days (Neuweiler & Möhres 1966). However, bats do
not trust their spatial memory exclusively, but can compare stored data with
new echo-acoustical and visual information (Joermann et al. 1988; Schmidt et
al. 1988; Höller 1995). When flying in a room of subdued daylight, the
two frugivores Carollia perspicillata and Phyllostomus hastatus (Phyllostomidae)
are able to see and avoid obstacles consisting of 30 cm wide strips of cloth
in their flight path (Chase & Suthers 1969). Those that were deafened with
earplugs avoided the obstacles significantly better than those that were both
deafened and blindfolded, showing that they could obtain visual information
of features in the environment during flight. These results are consistent with
those of Bradbury and Nottebohm (1969), who found that Myotis lucifugus (Vespertilionidae)
avoided collisions in a string maze better in dim light than in total darkness.
Rother and Schmidt (1982) noted that Phyllostomus discolor (Phyllostomidae)
uses fewer sonar pulses in adequate illumination than in darkness. When flying
the bats in a string maze, the same authors also showed that fewer pulses were
used if the obstacles exceeded 0.25 mm in width. The results suggest that vision
can shorten the bats’ reaction time for avoiding obstacles in a flight
path, as long as there is enough ambient light and the obstacles are of sufficient
size (given by the visual acuity threshold and the range).
Joermann et al. (1988) studied landing performance in two captive species of
Phyllostomidae (Desmodus rotundus and Phyllostomus discolor). The bats were
presented with visual illusions of landing grids, which thus gave them conflicting
acoustic and visual information. Although the grids were not detectable by echolocation,
the bats seemed to aim for them, and only ca 30 cm in front of the illusions
the bats interrupted the approach and turned away. The authors concluded that
bats rely mainly on echo-acoustical cues at close range, but in some situations
they defer to visual cues in an early phase of detection, even within the range
of echolocation.
To investigate what sense the Anoura geoffroyi (Phyllostomidae) (Chase 1981;
1983) and the Tadarida brasiliensis (Molossidae) (Mistry 1990) would defer to
when escaping from a roost, the bats were flown in a Y-maze, in which one exit
was blocked with Plexiglas and illuminated with a light source. The other exit
was open but dark. When tested in daytime, nearly all bats chose the illuminated
“exit”, thus indicating that they believed the light was an opening.
However, when releasing bats at night, the escape behaviour was the opposite,
the bats choosing the dark exit. It was suggested that the synchrony of light
schedules to the bats’ circadian rhythm might determine the use of the
appropriate sense (Mistry 1990).
Vision in foraging and prey detection
At close range, echolocation usually gives more detailed information about the prey than vision (Suthers & Wallis 1970; Pettigrew 1980). However, in some situations, it may be favourable to change the modality with which to search for prey, and indeed, many bats use a variety of sensory cues, including smell (Hessel & Schmidt 1994; Kalko et al. 1996; Helversen et al. 2000), passive listening for prey generated sounds (Fiedler 1979; Ryan & Tuttle 1987; Arlettaz et al. 2001), tactile information (Baron et al. 1996c), visual cues (Bell 1985), and vampire bats possess the ability of thermo-perception (Kürten & Schmidt 1982).
Insectivores and carnivores
For bats that search for insects within or near vegetation, separation of prey
echoes from the background clutter is usually a severe problem when using sonar
alone (Jensen et al. 2001). In such situations bats have to rely on additional
sensory cues to locate the prey. Nevertheless, few studies have addressed the
obvious possibility that visual cues may be used for detection of prey in acoustically
complex environments. However, when northern bats (Eptesicus nilssonii) search
for stationary targets among high grass (clutter), this seems indeed to be the
case (Paper II, Paper III). During early summer in Sweden, ghost swift moths
Hepialus humuli (Lepidoptera: Hepialidae) swarm in stationary display flight
over and among grass at dusk. These moths are large (ca 6 cm wingspan) and conspicuously
silvery white (Andersson et al. 1998), and in contrast to most other moths,
they lack ultrasonic hearing (Rydell 1998), and are intensively exploited by
northern bats patrolling in the air over the field (Andersson et al. 1998; Rydell
1998; Jensen et al. 2001). In an experimental set-up, making use of this natural
foraging situation, Hepialus humuli were presented to the bats, either with
their white dorsal side up or with their dark ventral side up. It was found
that the white moths were attacked more frequently than the dark ones, indicating
that the bats were guided by visual cues (Paper II).
The brown long-eared bat Plecotus auritus (Vespertilionidae) is a gleaning insectivore,
which usually uses its large and sensitive ears to passively locate its prey
by the noise they make (Anderson & Racey 1991). However Plecotus auritus
also has relatively big eyes (Cranbrook 1963, Tab 1), suggesting that they have
relatively good vision. We investigated if brown long-eared bats exploit visual
cues when searching for prey (Paper I). By using petri dishes, containing mealworms
that either were available to the bats or presented under glass, and presenting
these in different levels of illumination, we provided the bats with visual
cues, sonar cues or both. The bats did best in situations where both sonar cues
and visual cues were available, but the visual information seemed to be more
important than sonar.
The California leaf-nosed bat Macrotus californicus (Phyllostomidae), a gleaner
that normally searches for prey on the ground, has been shown to locate prey
by using auditory- and visual cues as well as by sonar. Indeed this bat shows
a particularly flexible hunting behaviour. In moonlight Macrotus californicus
can see well enough to hunt using vision alone (Bell 1985). This allows the
bat to hunt without alerting the prey with ultrasound (Fullard 1987; Rydell
1992a), and also to detect stationary targets, which otherwise would be hard
to detect (Arlettaz et al. 2001; Jensen et al. 2001; Paper II). In visual acuity
tests Macrotus californicus responded to stripes subtending 0.06° arc, (Tab
2), which is the best visual acuity found in any microchiropteran bat (Bell
& Fenton 1986). Moreover, the eyes of Macrotus californicus are relatively
large and have a much higher degree of binocular overlap (50°) than in other
bats (for example Antrozous pallidus 25° and Eptesicus fuscus 19°, Bell
& Fenton 1986). This suggests that Macrotus californicus has good stereoscopic
vision and that the near field distance perception is of great importance (McIlwain
1996), as would be expected in a species that forage visually. Macrotus californicus
exploits diurnal prey, that are stationary at night and therefore unavailable
to other bats (e.g. Howell 1920 cited in Bell & Fenton 1986).
The African yellow-winged bat Lavia frons (Megadermatidae) employs feeding tactics
that involve both gleaning and aerial hawking. This species is a sit-and-wait
predator, which scans the vicinity while hanging from a branch, waiting for
insects to pass by. Lavia frons is active in relative bright ambient illumination,
at dusk as well as late mornings, and is often seen catching prey against the
sky. It has large eyes and may be able to see insects against the bright sky
(Vaughan & Vaughan 1986). Nyctophilus gouldi and Nyctophilus geoffroyi (Vespertilionidae),
also combine aerial hawking with gleaning, and have been shown to use different
sensory cues according to circumstances. As in Lavia frons, visual cues are
preferentially used to detect prey in the air, whereas auditory cues are used
to detect prey on the ground (Grant 1991). The visual acuity of Nyctophilus
gouldi is nowhere near that of Macrotus californicus and Antrozous pallidus,
but rather similar to that of other aerial hawking Vespertilionidae (Tab 2),
which explains why they cannot find prey on the ground visually.
Eklöf & Anderson (unpublished) observed northern bats (Eptesicus nilssonii,
Vespertilionidae) feeding under midnight sun conditions in northern Norway.
The bats caught prey against the bright sky and sometimes without detectable
sonar signals. However, based on the performance of Eptesicus fuscus (Tab 2)
it seems unlikely that Eptesicus nilssonii has sufficient resolving power to
detect small airborne prey visually. A 2 cm insect is first detected at a distance
of ca 1 m using vision (considering a visual acuity of 0.7° -1° arc,
Tab 2), but the same object is first detected at ca 5 m using echolocation (Kick
1982), which thus suggests that echolocation would be the preferred sense. On
the other hand, when northern bats search for ghost swift moths (described above),
vision increases the chance of detection of the prey, only because they exceed
5 cm in wingspan and are detected at rather close range (3.5 m) (Paper III).
Smaller targets are detected using echolocation alone.
Little brown bats (Myotis lucifugus) have been observed to catch prey apparently
without using echolocation (D. R. Griffin personal comm.) This species’
visual resolving power is even poorer than that of the northern bat, and in
addition, its prey items are even smaller, so it is thus highly unlikely that
vision is involved in prey catching. In this species the apparent absence of
echolocation calls must have another explanation. In fact, earlier observations
of northern bats (Rydell 1992b) and little brown bats (Rydell et al. 2002) have
suggested that attempted insect captures are always associated with echolocation
calls, even in bright light conditions at high latitudes.
Under conditions that appear to us to be completely dark (0 lux), bats may still
be able to see conspicuous insects. For example, it has been reported that bat
activity is high where fireflies occur (Lloyd 1989), and it has been shown that
some fireflies stop flashing when approached by bats (Farnworth 1973). This
suggests that the light emitted by fireflies may guide the bats or at least
evoke their curiosity. More interestingly, fireflies are not eaten by bats and
were rejected by Eptesicus fuscus in feeding experiments (Vernon 1981). In the
same study, the bats were presented with flashing fireflies as well as with
artificial flashes. The bats responded to the flashes, although it was not clear
if they associated the flashes with food or with unpalatability. It seems possible
that firefly flashes may function as a visual aposematic signal to bats.
Frugivores and nectarivores
In general, fruit- and nectar feeding bats have larger eyes (Tab 1), better
visual resolving power (Tab 2) and enlarged visual and olfactory bulbs, compared
to insectivorous species (Jolicoeur & Baron 1980; Barton et al. 1995; Barton
& Harvey 2000). They also perceive and respond to different patterns more
readily than insectivorous species (Suthers & Chase 1966; Suthers et al.
1969), suggesting that vision may perhaps play a more important role in these
bats than in most insectivores.
Hessel and Schmidt (1994) investigated which sensory cues Carollia perspicillata
(Phyllostomidae) uses when orienting toward a food source. The bats were presented
with a triple choice of passive acoustic-, olfactory-, and visual cues. At least
initially, the visual cue was the most frequently preferred stimulus. But after
training the bats changed their behaviour and responded more to the olfactory
stimulus. The experiment suggests that Carollia perspicillata can detect new
sources of food using visual cues, and that they subsequently rely more on olfaction
as the food source becomes known. Indeed these bats seem to possess a remarkable
sense of olfaction (Fleming 1988; Laska 1990).
Kalko et al. (1996) showed that fig eating Microchiroptera do not use vision
when foraging, presumably because figs eaten by these bats are visually inconspicuous.
Instead, they rely mainly on olfactory cues, combined with broadband echolocation.
In fact, most bat-pollinated plants are greenish, pink and white, which presumably
reflect the fact that bats are most likely colour-blind (Suthers 1970; Faegri
& van der Pijl 1979). On the contrary, many species of bat pollinated Parkia
(Leguminosae: Mimosoideae) have bright red and yellow colours (Hopkins 1984).
It is also suggested that dark flowers can be seen as silhouettes, against the
sky and that pale flowers appear conspicuous against dark foliage (Start 1974,
cited in Hopkins 1984). If the bats make use of such differences in contrast,
one would expect to find that the position of differently coloured flowers vary
accordingly in relation to the foliage, i.e., red flowers far from foliage and
yellow flowers closer, which in fact, seems to be the case.
The capitula of Parkia are also highly reflective under moonlight and starlight
conditions, and are therefore presumably visible to pollinating bats (Hopkins
1984). Many pollinators make use of a broad spectrum reflected from flowers,
fruits or seeds, including ultraviolet (UV) light (for example insects, Kevan
et al. 2001; and birds, Church et al. 2001). Ultraviolet vision seems, however,
to be absent in most mammals, although some rodents have been shown to have
UV sensitive retinas (Jacobs et al. 1991). Recently, it was suggested that bats
might perceive UV-light, as there is evidence for a spectral sensitivity peak
around 390 nm (i.e. in the near UV-range) in the nectarivorous Glossophaga soricina
(Lopez et al. 2001). However, if the bats actually use UV reflecting surfaces
as orienting cues is still uncertain, although Willson and Whelan (1989) have
shown that UV-reflectance is indeed relatively common throughout the plant kingdom.
The Passiflora species, Passiflora galbana and Passiflora mucronata, two plants
which flowers are exploited by the nectarivorous glossophaginae bats, reflect
light down to ca 400 nm and 370 nm (upper UV range), respectively. This should
be compared to the hummingbird pollinated Passiflora speciosa, which has its
main reflection above 570 nm (Varassin et al. 2001), perhaps reflecting the
spectral sensitivity of the pollinators. Furthermore, 80% of nocturnal Lepidoptera
have wing patterns that reflect UV, compared to ca 30% in diurnal species (Lyytinen
2001 cited in Honkavaara et al. 2002). On the other hand, this may imply that
bats cannot make use of the ultraviolet light, in contrast to birds, which usually
forage in daylight.
Predator surveillance and social behaviour
As discussed earlier, vision seems to be important in escape behaviour (Chase
1981; Chase 1983; Mistry 1990). Presumably it is also important in detection
of predators; it is much easier to approach a blindfolded bat than a non-blindfolded
individual (Chase 1972). Species of the family Emballonuridae often fly earlier
in the evening than most other bats, and sometimes even in the afternoon and
they often roost on exposed and well lit sites such as tree trunks (e.g. Bradbury
& Vehrencamp 1976). A Saccopteryx sp. will quite easily detect an approaching
person, and take flight without emitting any echolocation calls (Suthers 1970),
and Rhynconycteris naso seems to be disturbed more easily by seeing an approaching
figure at a distance, than by sudden sounds or vibrations at close range (Dalquest
1957). Vaughan and Vaughan (1986) noted that Lavia frons (Megadermatidae), which
also roosts exposed, seems to be constantly alert during the day, scanning its
surroundings for predators. In fact, the authors almost never saw a bat with
its eyes closed, and were never able to approach one undetected.
The evidence for the use of vision in social behaviour is mainly anecdotal.
Social grooming occurs in the vampire Desmodus rotundus (Phyllostomidae) and
may serve to identify individuals (Wilkinson 1986), although it is generally
rare (Fleming 1988). Goodwin and Greenhall (1961) noted that avian vampire bats
(Diaemus youngi) show grooming behaviour when seeing a mirror reflection, indicating
that vision might be involved in this behaviour. Sometimes bats are also observed
to imitate other individuals grooming themselves (Vaughan & Vaughan 1986).
Some bat species have distinct fur patterns, which may serve as visual recognition
signals (Fenton 2001), in addition to scents and sound, although fur patterns
may also serve as camouflage (Neuweiler 2000). Threat displays are common in
for example Carollia perspicillata (Phyllostomidae), and includes wing shaking,
harsh sounds, and aggressive looks such as extension of the tongue (Fleming
1988). Sexual displays are also common. The monogamous Lavia frons and Cardioderma
cor (Megadermatidae), perform stereotypical circular flights, described as aerial
ballets (McWilliam 1987; Vaughan & Vaughan 1986). Among Saccopteryx bilineata
(Emballonuridae), the males defend territories where they maintain harems. In
front of the females of the harem, they perform sexual displays, which include
stereotyped singing, and also shaking of wings and hovering. The wing shaking
presumably enhances the effect of olfactory glands by spreading pheromones,
but it may also function as a visual signal to draw the females’ attention
(Chase 1972).
Multimodality – vision and echolocation
The echolocation detection range of a 19 mm insect is around 5 m for Eptesicus
fuscus (Kick 1982), and the visual acuity of this species is 0.7°-1°
arc, Tab 2). This allows visual detection of the 19 mm object only when it is
closer than ca 1 m. This simple calculation strongly suggests that echolocation
is the more accurate sense at close range and for small objects. However, larger
objects can be detected visually at distances of hundreds of meters, far beyond
the range of echolocation. For example, an object of 5 m diameter can potentially
be detected visually by Eptesicus fuscus at a distance of ca 300 m. Using echolocation;
the same object is detected at a distance of only 25-30 m at most (depending
on call strength, attenuation etc., Lawrence and Simmons 1982; M. B. Fenton
personal comm.). This supports the general view that vision is used primarily
for detection of large objects and landmarks and for navigating over longer
distances (Davis 1966; Layne 1967; Griffin 1970; Höller and Schmidt 1996).
Nevertheless, for bats with better visual resolving power, vision can be used
and even replace echolocation, at short distances. The California leaf nosed
bat Macrotus californicus, referred to above, can visually detect a 19 mm insect
at a distance of ca 18 m. This presumably gives this bat a longer range of operation
if they use vision instead of echolocation, at least under conditions of moonlight
or bright starlight (Bell & Fenton 1986). Other bats, such as some Emballonuridae,
which have visual acuities below 0.4° arc (Tab 2), can visually detect insect
sized objects (1 cm) at distances less than 1 m, suggesting a range of operation
roughly similar for vision as for sonar. One could therefore assume that emballonurid
bats could use either vision or echolocation to detect prey, as suggested by
Pettigrew (1980). He observed one species of Emballonuridae (Craseonycteris
thonglongyai) catching prey against a bright sky apparently without using echolocation
and suggested that the bats could see the insects as silhouettes against the
sky.
The Australian ghost bat Macroderma gigas (Megadermatidae) also has a similar
prey detection range for vision as for echolocation. Since this species also
has good auditory sensitivity in the sonic range (Fiedler 1979; Kulzer et al.
1984), it switches between vision, echolocation and passive listening (Pettigrew
et al. 1986; Pettigrew et al. 1983). For frequencies below 20 kHz, the acoustic
axis (as defined from the directionality of the pinna and noseleaf) of Macroderma
gigas is aligned with the visual axis (defined by areas of highest ganglion
cell density), indicating that auditory cues help the bats to visually detect
the source of the sound (Pettigrew 1988). In fact, a major function of sound
localisation in animals is to direct the eyes toward the sound-source (Heffner
& Heffner 1992; Heffner et al. 1999). This reflex is even quicker than the
reaction to a flashlight (Whittington et al. 1981), and hence suggests that
hearing is closely co-ordinated with vision (Heffner 1997).
Sound localisation acuity is related to retinal organisation and the width of
fields of best vision (defined as the portion of the retina with at least 75%
of maximum ganglion cell density, Heffner et al. 2001). Animals with narrow
fields of best vision (foveae) have generally better localisation acuity than
animals with broad or elongated fields of best vision (visual streaks). The
retinas of microchiropteran bats are loosely arranged in visual streaks and
the density of ganglion cells falls irregularly toward the periphery. The field
of best vision is concentrated in the temporal part of the retina and seems
to be broader in frugivores than in insectivorous species (Heffner et al. 2001).
Overall, there is a higher ganglion cell density in the inferior part of the
retina than in the superior (Marks 1980; Pettigrew et al. 1988; Koay et al.
1998; Heffner et al. 2001; Eklöf unpublished). This means that the sharpest
image on the retina results from light reaching the eye from above, and consequently,
the bat eyes focus slightly upwards. Without moving their heads, bats are looking
up (Pettigrew 1988). The functional significance of this can be difficult to
establish, but it seems likely that vision and echolocation have evolved to
provide the bat with as little information overlap as possible. While echolocation
call emission and hearing is most effective in the flight direction and downwards
(Schnitzler & Grinnell 1977; b), vision serves as a complement by being
most effective upwards it thus gives additional information of obstacles and
landmarks further away. In Megachiroptera, which do not echolocate, one would
thus expect the fields of best vision to be above rather than below the optic
disk, which in fact seems to be the case (Pettigrew 1986).
All bats have well developed retinofugal projections (pathways of information
from retina to visual cortex) to the lateral geniculate nuclei as well as to
the superior colliculus (see above), which are the main targets for retinal
projections in mammals (Pentney & Cotter 1976; Suthers & Bradford 1980;
Reimer 1989). In the superior colliculus, different sensory modalities are integrated
and transformed, and the output may be perceived as a “new product”
(Stein & Meredith 1993). The capacity to deal with multisensory information
is however developed first after experience of multimodal inputs (Wallace &
Stein 2001). The superior colliculus controls for example eye movements, which
serves to keep objects of interest in the focal field. Auditory projections
to the superior colliculus are generally sparse in mammals. However, in the
mustache bat Pteronotus parnellii (Mormoopidae), at least three areas in the
brain stem contribute with well-developed auditory projections to the superior
colliculus. It has been shown that pinna movements can be controlled in the
same way as eye movements in other mammals (Covey et al. 1987), and thus that
orienting behaviour can be influenced through auditory as well as through visual
feedback. It is also known that auditory stimuli can trigger visuomotor neurones
and hence that the eyes can respond to sounds (Stein & Meredith 1993). Combined
sensory inputs can enhance perception and detection, but also cause behavioural
depression, for example when the cues are contradictory, as in the case with
bats and windows (discussed above). Cats have been shown to respond “half
way” between contradicting sounds and images (Stein & Meredith 1993),
but in most cases when animals have multiple cues to choose from, one can see
a clear sensory hierarchy (e.g. Dyer & Gould 1981), so also in bats (Chase
1983). However, the hierarchy can change depending on the behavioural context.
Visual cues have been shown to have precedence over auditory cues in for example
escape behaviour and when commuting (Chase 1981; Chase 1983; Mistry 1990; Paper
V). In cases where echolocation and visual cues are complementary rather than
contradictory, the bats may still rely on vision over sonar. In a study on brown
long-eared bats (Plecotus auritus), feeding from bowls presenting different
sensory cues (Paper I), the bats scored best in situations where both visual
and sonar cues were present. The visual information seemed however to be the
more important.
It has been suggested that there sometimes can be interference between the two
senses. For example, Simmons (cited in Chase 1981) has noted that some bats
have a problem learning acoustic discrimination when visual cues are present,
but can easily perform the same task in darkness. When trained to respond to
black or white triangles of different size, Myotis lucifugus (Vespertilionidae)
responded to brightness cues rather than the size of the triangles, although
these bats are capable of size discrimination by echolocation (Ellins 1970;
Masterson & Ellins 1974). This suggests that interference may have occurred,
or at least that the bats had a preference for visual cues in this case.
It is not yet known if bats can perform cross-modal recognition, i.e. learning
an object using one sense and then immediately recognising the same object by
using another sense, which is the case with for example bottle nose dolphins
Tursiops truncatus. These animals can integrate information from vision to echolocation
just as well as from echolocation to vision. Hence, what the dolphins perceive
from one sense is functionally similar of what it perceives from the other (Pack
& Herman 1995).
Although the question of how sensory inputs are combined in bats remains unsolved,
several authors have shown the importance of multimodality (Pettigrew et al.
1983; Schmidt 1988; Hessel & Schmidt 1994). In a two choice test, two phyllostomid
bats (Desmodus rotundus and Phyllostomus discolor) were trained to respond to
a combination of visual, olfactory and acoustic stimuli, and were then presented
with one of the three modalities separately (Schmidt et al. 1988). It was found
that Phyllostomus discolor chose the visual stimuli to a higher degree, whereas
Desmodus rotundus preferred the passive acoustic stimuli. However, both bats
were able to respond to all three modalities, although responses to the olfactory
stimuli needed additional training, as also noted by Hessel and Schmidt (1994),
when studying Carollia perspicillata. However when the Carollia had learned
to respond to the olfactory cue, this became the preferred stimuli, which was
not the case with Desmodus or Phyllostomus, which both used two other senses.
This clearly shows that bats use an array of different senses in the field,
and that ecology, feeding strategies and behavioural context all influence the
use of different modalities.
Echolocation may be the most important innovation throughout bat evolution,
allowing these animals to explore a niche of their own. But there is more to
the sensory ecology of microchiropteran bats, where vision is an important piece
of the puzzle and certainly needs further attention in the future.
------------------------------------------------------------------------------------------------------------------------------------------------------------
Get the thesis! Send an email to:
(C) 2005 Johan Eklöf