|
|
Training |
Test(1) |
Test(2) |
Test(3) |
Test(4) |
|
Tested association |
All (composite) |
Classical (CS-US) |
Operant (B-US) |
B-US vs. CS-US |
|
|
Mode |
Switch |
Switch |
Flight Simulator |
Yaw torque learning |
Switch |
|
Choice |
CShBh vs. CScBc |
CShBh vs. CScBc |
CSh vs. CSc |
Bh vs. Bc |
CShBc vs. CScBh |
|
Behaviours and stimuli used |
Yaw torque domain and colours |
Yaw torque domain and colours |
Flight direction and colours |
Only yaw torque domain |
Contingency between colours and yaw torque domain reversed. |
|
Figure |
1c |
1d |
1e, g |
1f, h |
1i |
| Table 1: Experimental design. All animals are trained to avoid CSh (e.g. blue) and Bh (e.g. the ‘left’ yaw torque domain) and to prefer CSc (e.g. green) and Bc (e.g. the ‘right’ yaw torque domain) during switch-mode training. Subsequently, the flies are divided into four different test groups: Test(1) through (4). During test, the flies’ choice of yaw torque domain (B) and/or arena coloration (CS) is recorded without reinforcement. Each test is designed in a complementary manner to assess the relative contributions of the classical (CS-US) and the operant (B-US) associations in comparison to all possible associations in the control test(1). CSh – the fly is heated during training if the arena is illuminated by this colour; CSc – the fly is not heated during training if the arena is illuminated by this colour; Bh – the fly is heated during training if it produces yaw torque of this range; Bc – the fly is not heated during training if it produces yaw torque of this range. (back) |