View larger version (30K):
[in this window]
[in a new window]
|
Fig. 2. Experimental setup and measurement procedure. (A) TrackFly. The free-flight
experimental setup consisted of a wind tunnel (only the working section is
shown) equipped with a real-time 3-D position tracking system (Trackit 3D)
(Fry et al., 2000 ;
Fry et al., 2004) and
custom-programmed graphical rendering software (based on the VisionEgg)
(Straw and O'Carroll, 2003;
Straw et al., 2006). Flies
were induced to fly upwind (red dotted arrow shows the flight direction of
fly; blue arrows indicate wind direction), while their position was tracked in
real time (green arrows pointing from cameras). Visual stimuli were projected
onto the sidewalls via pairs of mirrors (yellow arrows show the light
path from the projector to the screen for one side of the wind tunnel). The
virtual reality features of TrackFly were used to implement a one-parameter
visual open-loop paradigm [see text and Fry et al.
(Fry et al., 2008) for further
details]. (B) Definition of linear and angular coordinate systems. A plan view
of the screens (sine gratings are represented by stripes for clarity) is shown
to scale. The linear spatial frequency (SF, in m–1)
of the pattern displayed on the screen corresponds to the inverse of the
linear spatial pattern period ( , in m). In the example, SF=10
m–1 and thus  =0.1 m. The angular period (,
unit: deg.) of the displayed patterns depends on the azimuth, with
decreasing toward frontal and caudal positions (note red arrows). Angular
spatial frequency (sf, in deg.–1), the inverse of
, therefore increases toward frontal and caudal positions. The linear
wavelength (1, 2) and linear spatial frequency SF
remain constant. (C) Sample acceleration responses and parameter extraction.
The time course of body position along the wind tunnel (red traces) is shown
for 11 measurements performed under identical stimulus conditions. Flies were
first held near the middle of the wind tunnel by controlling the pattern speed
(t<0). At t 0, the flies were stimulated in open-loop
(TF=4 s–1; SF=12.5 m–1;
stimulus condition is marked with an asterisk in
Fig. 4A). Flies reacted to the
back-to-front image motion by accelerating forward, as indicated by the
exponential increase of the position function (red traces, right part of the
plot). Mean acceleration of each sample was measured from the fitting
parameters of a parabola (t>0.1 s, black traces). These values
were then averaged over the trials to obtain the response strength for a
single TF–SF combination (each marked with a dot in
Fig. 3 and
Fig. 4). Figure modified from
Fry et al. (Fry et al.,
2008).
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
0.6 m s–1;
SF<25 m–1) in presence of low (left: –0.29
m s–1; N=941) and high (right: –0.73 m
s–1; N=45) wind speeds. The measured accelerations
reveal a significant dependence on the wind speed, suggesting an
air-speed-dependent response saturation (Wilcoxon test,
P<0.001).