Dr Kate Jeffery is one of our heroes and has a lab at University College London where she experiments with what the brain is doing when rats navigate around a maze. In her experimental rig, she can see neurons fire up as the rat faces in different directions (head direction cells) while it is exploring. These head direction cells as the rat moves its head to each direction (each cell has its own preferred direction). So this provides an on-board compass that is not magnetic, but rather uses the visual world combined with information about movements. As we have described elsewhere landmarks are critical to navigation as they recalibrate the other navigational inputs and give a fix, while self-motion information allows the compass to be updated.
Her latest paper covers the problem of a rat navigating in three dimensions.
In a nutshell, the question is whether an animal (or indeed a human) can get away with a flat compass in a 3D world, given that a compass only gives you 360 degrees of direction but 3 dimensional space is 360 x 360 x 360 degrees (one for each plane – horizontal, North-South vertical and East-West vertical), which would be a lot more neurons for the brain to have to make. In addition, most of the neurons would also be silent most of the time, which is wasteful of the energy needed to keep cells alive. However, if you only have a flat compass that tracks yaw (head direction) rotations, then if you go around vertical corners, the compass ends up pointing the wrong way. Jeffery’s team showed that you can avoid this error if you adjust the compass to account for rotations around vertical corners (or indeed, any type of rotation around a vertical axis – not necessarily a corner).
So, the team propose two rules, one rule for adjusting the compass when you turn left or right in the body plane in the usual way, and a second one that adds an additional correction of the whole body plane itself rotates around a vertical axis. In this way, one could, for example, walk right around the side of a hill without turning left or right, and yet still have an appropriately updated compass.
Prof Jeffery’s team did some simulations to show this works in principle. They then recorded head direction cells from real rats navigating around the vertical corners of a cube and showed that they indeed seem to update according to this rule. Thus, it’s possible that rats, and maybe other animals, can use a flat compass when moving over non-flat surfaces, provided they use this dual-axis rule, without needing to make a lot of neurons and without accruing errors.
In addition to a compass, the rats, as all good navigators, must have a way of measuring distance (odometer). To do this they probably use several things – footfall, optic flow, vestibular signals. These are processed in another part of the brain by a different type of neuron called grid cells – more on these another time.
Please follow the link : http://jn.physiology.org/content/jn/early/2017/10/06/jn.00501.2017.full.pdf
Prof Kate Jeffery has kindly helped me understand some further truths about animal navigation. These are not proved but seem to stand up to investigation. For humans, it isn’t just head-on-neck turning, it is anything that rotates the whole head in space that is what helps guide navigation. You’ll find that in the course of a day you face pretty much all directions without becoming disoriented, even when going over and around hills, so that her work investigates how this is achieved.
In summary we are beginning to tease out how animals navigate but we are still far from understanding how migratory species and long distance travellers know where they are going.