Bats don’t echo brain theory

Photobucket - Video and Image Hosting

The prevailing computational model of spatial memory in mammals is based mainly on investigations of hippocampal function in rats. This work has shown that the hippocampus contains a special class of pyramidal neurons called place cells, which fire when the animal is in a particular location of its environment (called the “place field”).

Place cells have also been identified in the human hippocampus, and in non-human primates, and there are now known to be other distinct types of cells with spatial firing properties. In monkeys, for example, there are so-called “spatial-view” cells which are sensitive to the direction in which the animal is gazing; very recently, researchers identified putative navigation neurons which fire when monkeys move in a specific direction of a given location, and movement-selective neurons which fire in response to a left or a right turn in any location.

The model of how hippocampal function produces spatial memory is based on a number of characteristic firing patterns generated by rat place cells. One of these patterns of brain waves is the short, high frequency “ripple” oscillation (120-200Hz) which occurs during slow-wave sleep, and which are thought to be involved in consolidating newly-formed memories by transferring the information to the neocortex. The other is the low frequency (5-10Hz) theta oscillation that is generated while the animal is exploring or moving; these theta waves are likely to play a role in navigation.

Both place cells and spatial view cells have been reported in the human hippocampus, and it is generally assumed that the hippocampal activity underlying spatial memory was similar in all mammals. But new research published in the February issue of Nature Neuroscience shows that the current model of spatial memory may not be representative of all mammals. Nachum Ulanovsky and Cynthia Moss, of the University of Maryland’s Auditory Neuroethology Laboratory, have, for the first time, performed in vivo recordings of the activity of hippocampal cells in freely moving echolocating bats. Their findings show that the brain waves generated by echolocating bats differ from those produced by rats, suggesting that the current model of how hippocampal function generates spatial memory may need to be reconsidered.

Bats use echolocation to navigate and to forage for food. These flying mammals have an excellent spatial memory, which they use to migrate long distances and to re-use “flyways” from their roosts to foraging sites. Typically, on arrival to a foraging site, they find food on the wing by emitting ultrasonic bursts and detecting the echoes produced by prey. But can also forage while crawling, using space-based (or allocentric) navigation; in rodents, this type of foraging is wholly dependent upon the functions of the hippocampus.

Ulanovsky and Moss trained two big brown bats (Eptesicus fuscus) to forage for mealworms suspended within an open field arena.The bats were then anaesthetized, and approximately 1.4mm-squared openings were made in the parts of the bats’ skulls overlying the hippocampus. Miniature electronic devices, consisting of an array of 4 independently movable microelectrodes and a micro-hard drive, were then implanted onto the surface of the bats’ brains. Over the course of a week after the surgery, these 4-tetrode microdrives were slowly lowered into position into the CA1 region of the bats’ hippocampi, so that the activity of hippocampal cells could be recorded while the animals foraged for food during the behavioural tests, and light-emitting diodes were placed on their heads so that their movements could be tracked.

In both bats and rodents, place cells are contained within the thin CA1 layer of the hippocampus. It is because of this similarity in hippocampal tissues that Ulanovsky and Moss were able to insert the microelectrode array directly into appropriate location – this is the technique that has been used to record the activity of place cells in freely moving rodents. Their results showed that the bats produced echolocation calls continuously, except while they ate the mealworms they found or while they slept. As in rats, about one third of the pyramidal cells in the hippocampus were characterized as place cells. Their abundance, and many of their properties, were very similar to the place cells found in the rodent hippocampus. As in the rat, some of the the bats’ cells had diffuse place fields, but most were activated only in a spatially restricted manner, and less than 50% of the place cells were active at any one time. And the location and amplitude of the high-frequency ripple oscillations were similar to those in mice.

There were, however, some significant differences between the activity of the cells in bats and rodents. The theta oscillations recorded from the bats’ hippocampi were not continuous, as they are in rodents, but occurred about once every minutes in short bouts which lasted 1-2 seconds, rather like the activity that has been observed in cats. Furthermore, theta oscillations are generated when rodents are moving freely, while in bats they were emitted only when the environment was explored during crawling. While foraging in this way, bats were almost stationary.

The differences in theta oscillations between rodents and bats may be explained by differences in behaviour. Rodents have poor vision and instead rely heavily on smell and touch to explore their immediate environment, and, although bats also have poor vision, they can use echolocation to explore their environment from a distance. The oscillations were recorded only during the bats’ stationary foraging, so it is possible that continuous theta waves, like those observed in rodents, can be generated during aerial foraging. The authors speculate that the size of the theta wave oscillations could increase with flight speed. However, the bats’ theta waves recorded probably occur too infrequently to be involved in navigation, as they are in the rodent, and may instead play a role in the processing of spatial information. Whatever the reasons for the differences, it appears that the current model of spatial memory is inadequate.


Ulanovsky, N. & Moss, C. F. (2007). Hippocampal cellular and network activity in freely moving echolocating bats. Nature Neurosci. 10: 224- 233.

O’Keefe, J. & Nadel, L. The hippocampus as a cognitive map. Oxford University Press, Oxford: 1978.