Home - Back to bats - Visual prey detection in vespertilionids
Talk on vision in microbats: NA Annual symposium on bat research, USA 2002
Johan Eklöf
(no slides)
Northern bats (Eptesicus nilsonii) use vision but not flutter-detection when searching for large prey in clutter
My talk this morning is about northern bats, and how they use vision to detect non moving prey in a cluttered environment. It is based on results from field studies in Sweden, over the last two summers.
The northern bat, Eptesicus nilsonii is one of the most common bats throughout
Scandinavia. It is an aerial hawking bat, and is often the first seen at dusk.
This means that in summertime they often forage in relatively bright ambient
light.
If you are lucky and close to a hayfield in southern sweden, preferably in Late
June, you may see these guys. They are Ghost swifts: conspicuous large (ca.
6cm), silvery white moths that swarm in a stationary display flight over and
among the grass. There can be hundreds and even thousands of moths in one field
and they are heavily exploited by northern bats. I have to say that it is quite
spectacular to watch, ....Anyway.....As all hepialid moths, the ghost swifts
lack ultrasonic hearing, and the only protection they have is to be stationary
and close to the grass panicles.
...When searching for prey in the vicinity of vegetation, bats have to separate the echoes from targets, from the background clutter, which seems to be a serious problem for many bats. Presumably, bats feeding close to vegetation have to use additional sensory cues.
For example, the importance of prey generated sounds have been shown in severel species, including long-eared and mouse eared bats, just to mention a few.....
Bats using constant frequency pulses which are doppler shifted against fluttering targets such as flying insects, can use this to separate the prey from the background, so called flutter detection. Some lab experiments suggest that also vespertilionid bats can use flutter detection, although they use only Frequency modulated signals (and also lack the auditory specialistaions). However, this has not been verified in the field.
The bats visual sensitivity is relatively good in low illumination, suggesting that they may use visual cues at night. Vision has previously been shown important for long distance navigation and for detection of large objects, at least for some species. In some situations visual cues even seem to override the acoustic information. However, only few bats are shown to use vision in prey detection, where I guess the most famous example is the californian nose leafed bat, Macrotus californicus which can catch prey on the ground using vision alone. It is also suggested that some species may use the twilight sky to backlight prey, so that the insects are seen as silhouettes, especially for bats flying early at dusk, like some emballonurids, but also vespertilionids...
So the purpose of this study was to examine which additional sensory cues, if any, northern bats use to detect ghost swifts hovering among high grass (clutter).
We made our experiments in the field, and we made use of the natural situation,
where ghost swift leks occured. We put two sticks in the middle of the field,
and mounted moths on top of them. One live, fluttering and one dead and spread
moth, both of them of same size and colour. We noted which moth was attacked
first. Then we changed the position of the sticks and again noted which one
was attacked...and so on.
It was light enough to observe the bats visually without artificial lights or
light tagging.
So, what happened? as you can see, we had several attacks on the mounted moths. However, there was no significant difference between the two. The control in this case consisted of an empty stick.
From these results, we draw the conclusions that the bats do not make use of the wing flutter when detecting the prey in this case (as is suggested from previous laboratory experiments).
To test if the bats instead used vision, we again mounted two moths on sticks, but this time we used two dead moths beside each other. We turned one of them upside down, because on the ventral side the ghost swifts are not white, but dark grey to black. We assumed that the acoustic image of the two sides were the same to the bats, that is, the only difference to the bats would be the colour. Again we compared the attacks in the same way as in the first experiment.
And again, we found that the bats attacked both moths, but this time we found a diference between the two targets. There were significantly more attacks on white ones.
Therefore, we suggest that the northern bats use vision to detect the lekking ghost swifts
But is it really that simple? I guess not. As you probably noticed, there were a number attacks on the dark moths, and throughout the study we used bat detectors and could detect echolocation calls from circling bats, as well as distinct feeding buzzes from bats attacking moths.
We therefore believe that vision alone is not enough to catch the moths. We think that vision plays a major role in detecting the prey initially, but that echolocation takes over in the terminal attack.
Similar has been suggested by Joermann and his collegues in 1988, when they studied phyllostomid bats during a landing performance.. The bats in that study were presented with one landing grid and one visual projection of a the grid and the conclusions were that the bats seemed to discover the landing grid visually, but that the actual landing was always under the control of sonar.
So, We have concluded that Eptesicus can use vision to detect ghost swifts, but as I said, the ghost swifts are big, around 6-7 cm, so the question is, how large must the prey be for the bats to see it ..... Again, we have our two sticks, and this time we cut the wings of the dead and spread moths and presented the bats with 6, 5, 4 and 3 cm large prey. We compared the number of attacks as previously.
At 6 the cm and 5 cm large moths, the results are the same as in previous experiment.
But at 4 cm, there is no longer any significant difference in attack frequency between the white and dark moths, although there are slightly more attacks on the white ones
...and at three cm, there are almost equally many attacks on the two moths.
So, as we presented the bats with smaller and smaller prey, we got relatively
speaking less attacks on white moths, and more attacks on dark ones
...and when the prey was smaller than 4 cm, the bats could not make use of the
visual information.
Now we know that the bats are able to detect at least 4 cm large objects...
...and we know that the bats circle consistently on 3-4 meters above the field, before diving almost vertically towards the prey.
This means that they need a visual acuity of at least 0.7 deg of arc to discover the prey . And this is also the measured spatial resolution of the north american relative to the northern bat, Eptesicus fuscus, and presumably Eptesicus nilsonii has a resolution similar to this, although it has not been measured... (yet anyway, I’m working on it...)
So far, I have been discussing prey that is stationary, and not moving in space. But what happens if we present the bats in a similar manner, with objects moving rapidly over the grass panicles, simulating flying insects. Will the bats still use vision for detection?
This time we did not use moths, but instead paper marbles of sufficient size, which we shot across the field using a sling shot. We shot black and white marbles in random order and counted the number of attacks The results show that there were equally many attacks on the two colours, and opposite to when the prey was still, the bats ignored the visual information from the white. The movement and sonar seem to give the bats more reliable information of a prey, than the visual cues in this case..
It is about time I sum up what we have learned today.
First: Eptesicus nilssonii, the northern bat, do not make use of wing flutter from ghost swifts, (contrary to what have been suggested from laboratory results)
Second: the bats use vision to get additional information about prey hiding in clutter, at least as long as it is conspicuous and of a certain size
And third. The resolution of the nortern bat eye is presumably around 0.7 deg of arc, which is similar to what have been found in Eptesicus fuscus
thank you