First published online February 12, 2007
Journal of Experimental Biology 210, 865-880 (2007)
Published by The Company of Biologists 2007
doi: 10.1242/jeb.02707
Escape behavior and neuronal responses to looming stimuli in the crab Chasmagnathus granulatus (Decapoda: Grapsidae)
Damián Oliva,
Violeta Medan and
Daniel Tomsic*
Laboratorio de Neurobiología de la Memoria, Depto.
Fisiología, Biología Molecular y Celular, Facultad de Ciencias
Exactas y Naturales, Universidad de Buenos Aires, IFIBYNE-CONICET, Buenos
Aires 1428, Argentina
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 1. Experimental set-up and generation of visual stimuli. (A) The set-up for
delivering visual stimuli is located inside a sealed Faraday cage. Four
monitors are located at 20 cm to the sides, in front and above the animal (the
upper monitor is not shown). Stimulus signals generated by PC1 were directed
to any one of the four monitors through a selector switch located outside the
Faraday cage. A second computer (PC2) was used in combination with PC1 to
record the response of the crab. (B) The looming stimulus was a simulated
projection of an approaching object from one of the monitor screens. The
simulation corresponded to a black square object of 5 cm that approached from
a distance of 70 cm at a constant velocity of 20 cm s–1 (see
Materials and methods).
|
|
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 2. Measurement of the escape response. Locomotor activity was studied in a
walking simulator device consisted of a styrofoam ball that could be freely
rotated by the animal. The crab was held in position by a weightless rod
attached to its carapace that could move freely up and down within a vertical
guide located above the crab. Both the rod and the guide sleeve had square
cross-sections, which prevented the animal from rotating around its yaw axis.
The horizontal position of the floating ball was stabilized by four contact
points separated by 90° and provided by two optical mice and two flexible
sheets. Locomotion was assessed by recording the rotations of the ball with
the two mice according to the method described in the text (see also
Fig. 1).
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3. The escape response to looming stimuli. (A) Each trace corresponds to the
distance covered by a different animal during 10 s, starting from the
beginning of the stimulus presentation at the right side. Different animals
run different distances. However, when normalized to the maximum value, the
time courses of all responses are remarkably similar (B). The response then
can be divided into four distinguishable phases (for further details see the
text and Movie 1 in supplementary material). The dotted line signals the time
at which the virtual object would collide with the animal. N=18, one
trial per crab.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4. Threshold of escape response to looming stimuli. Individual responses from
different crabs are displayed from the beginning of stimulus expansion up to
0.5 s following the time of virtual collision (same dataset as shown in
Fig. 3). The time at which each
response was initiated (response latency) was projected onto the curve that
describes the angular size of the stimulus, which identifies the apparent
stimulus size at the time the crab initiated the escape response. The inset
shows at higher magnification the part of the responses corresponding to the
time enclosed by the horizontal rectangle. See the text for further
details.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 5. Responses to repeated looming stimulation. (A) Repeated presentations of
the same looming stimulus at 1 min intervals produce a progressive reduction
of the escape distance (open squares, left axis) and an increase in response
latency (filled circles, right axis) (N=14 crabs). (B) Repeated
stimulation with longer inter-trial intervals (3 min) causes less changes to
the escape distance and did not affect the response latency (N=15
crabs). Broken lines indicate the value of the mean response latency at first
trial.
|
|
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 6. Directional sensitivity of the escape response. The looming stimulus was
presented to each animal once in each of the four different visual regions
corresponding to the position of the monitors, and the sequence of
presentation was varied among animals. (A) Polar plots show the trajectories
of the escape responses to the stimulus approaching from each one of the four
screens: upper left, stimulus above; upper right, stimulus in front; lower
left, stimulus at left side, lower right, stimulus at right side. Black
circles denote the direction of escape for each animal. (B) Response distance
of the responses shown in A. (C) Latency to initiate the escape of the
responses shown in A. Values are mean response scores ± s.e.m.;
N=10 crabs, one trial at each stimulus location.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 7. Response of an M1 and an M2 type of movement detector neuron to the looming
stimulus appearing from the right visual field. Traces exemplify individual
responses from each neuronal type. The time of spike occurrence shown below
the traces illustrates the consistency of the response for each neuron to nine
presentations of the stimulus, separated by 1 min (the traces correspond to
the last trial of these series). Peristimulus time histograms show the mean
spike rate obtained from the nine trials. Data are divided into 100-ms bins
and are plotted as means ± s.e.m. Angular size of the looming object is
given in the lower traces of each panel. Arrowheads mark the beginning of the
stimulus expansion; long arrows mark the moment at which the synaptic activity
or the spike rate of neurons had clearly increased above resting level.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 8. Response of an M1 and an M2 neuron to (A) a flash of light, (B) a laterally
moving object and (C) an approaching object. (A) The horizontal open bar below
the traces represents 1 s duration of light stimulation (200 mW
m–2 at the crab eye). (B) The black bar represents 1 s motion
of a black square object moving laterally to the crab. (C) The curved line
represents the time course of expansion of the looming stimulus as described
in Fig. 7. Arrows in the traces
mark the approximate moment at which the neurons began to respond to the
looming stimulus by increasing their synaptic activity or spike rate above
resting level.
|
|
View larger version (49K):
[in this window]
[in a new window]
|
Fig. 9. Response of M1 and M2 neurons to different types of movement. (A–E)
Results with (A) the black looming stimulus, (B) the black receding stimulus,
(C) the white looming stimulus, (D) the laterally moving stimulus and (E) the
gradual darkening. Sample recordings from a single M1 and a single M2 neuron
illustrate the type of responses of these neurons to the five different
stimuli. Peristimulus time histograms show the mean spike rate recorded from
11 M1 and 13 M2 neurons from different animals (one trial per stimulus per
neuron). Data are divided into 100-ms bins and are plotted as means ±
s.e.m. The angular size of the looming and receding object is given in the
bottom traces of each panel. Arrowheads signal the beginning of the stimulus
expansion; black bars represent 1 s motion of a black square object moving
laterally to the crab. The darkening bar in E represents a gradual darkening
of the screen without motion components.
|
|
View larger version (10K):
[in this window]
[in a new window]
|
Fig. 10. Comparison of the behavioral and the neuronal responses to (A) a black
approaching object (B) a black receding object, (C) a white approaching
object, (D) an object moving laterally, and (E) a gradual darkening of the
screen. Behavioral responses are shown as the mean response distances covered
during 10 s following the beginning of the stimulus (left axis). Responses of
M1 and M2 neurons to the same set of stimuli were assessed by counting the
number of elicited spikes during the period of stimulation (right axis). In M2
neurons the spontaneous rate of firing was subtracted from the response. To
obtain the behavioral and the electrophysiological data, crabs were presented
with only one stimulus of each type separated by 3 min. In both behavioral and
electrophysiological experiments the sequence of presentation was varied among
animals. Bars represent mean response scores ± s.e.m. Behavior,
N=14 crabs; neurons: M1, N=11; M2, N=13.
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 11. Temporal relationship between the neuronal and the behavioral response to
the looming stimulus. Peristimulus time histograms show (A) the mean (±
s.e.m.) running speed of the animals and (B,C) the mean (± s.e.m.)
spike rate of M1 (B) and M2 (C) neurons, calculated over 100-ms bins and
plotted against time during the object approach. (D) Angular size and (E)
speed of expansion of the looming object. Arrowheads mark the beginning of the
stimulus expansion; the dotted line marks the moment at which the escape
behavior was initiated; the broken line indicates the moment when the stimulus
reached its maximum expansion and velocity; the vertical solid line marks the
point in time at which collision would have occurred; arrows in B and C point
to the neuronal activity 120 ms before the first movement of the animals.
Behavioral data are from 14 crabs (one trial per crab). Neurophysiological
data are from 19 trials recorded from a total of 10 M1 neurons and 18 trials
from 13 M2 neurons, each neuron from a different animal (i.e. no more than 2
records per crab were included).
|
|
© The Company of Biologists Ltd 2007
