Territory and Navigation

1
Territory and Navigation

• Part II Navigation

2
Navigation

• We talked about territoriality
• To get around a territory an animal must know
  where it is going
• Indeed, for most any animal that moves, it needs
  to have a way of knowing where it has been and
  where it is going
• There are many ways this can be accomplished,
  from complex cognitive mechanisms to simple odour
  trails

3
Path Integration

• Simplest form of navigation that uses memory
• Cataglyphis, the Long Legged Desert Ant
• Twisting outgoing path, but a direct path home.

4
Path Integration

• Animal Stores direction and distance
• Simple vector mathematics
• Animal must maintain a running calculation
• Error will be cumulative
• How could it be improved?

5
Path Integration and Landmarks

• While integrating the animal could,
  periodically, take a fix.
• Probably from the stars or the sun
• Clock shift experiments show this to be true!
• Same thing sailors used to do with a sextant (or
  a GPS today)
• But stellar position changes over time, animal
  cannot have stellar positions hardwired!

6
Beacons and Landmarks

• A beacon directs behaviour towards it
• A landmark points toward a goal, along with other
  cues
• Both are used by many animals
• There have been some great strides made in
  understanding various species use of landmarks
  in the last 20 years

7
Bees!!

• So, the bee seems to be matching the SIZE of the
  retinal image with the size of the image in
  memory
• Colour change has no effect
• Making the landmark a wireframe has no effect

Training
Half size test
Double Size test
8
Two Landmarks
Training
Stretch Test

• Bees are sort of half using angular information
• 3 Peak places of search in the Stretch test
• Collett, Cartwright, Cheng and their colleagues

Rotation Test
9
Landmark Use in Pigeons

• Ken Chengs (1989) work

Landmark
Goal
10
Tests of Pigeon Landmark Use
Landmark
Search

• Animal searches along the same axis of landmark
  shift
• Does not COMPLETELY follow the landmark
• But, does not shift search in the other direction

11
Further tests of.

• Now the shift is up down, so the animal searches
  in the up/down axis

Landmark
Search
12
The Vector Sum Model

• Cheng concluded that the pigeons must be adding
  self goal and goal to landmark vectors.
• This is the only model that explains the search
  patterns
• Eureka!

13
Conclusions

• Animals can use many sources of information to
  navigate
• Multiple sources usually point to the same place
• Hierarchical representations I think