First published online January 3, 2006
Journal of Experimental Biology 209, 327-342 (2006)
Published by The Company of Biologists 2006
doi: 10.1242/jeb.01982
Neural responses of goldfish lateral line afferents to vortex motions
Boris Phillippe Chagnaud1,*,
Horst Bleckmann1 and
Jacob Engelmann1,2,*,
1 Institute for Zoology, University of Bonn, Poppelsdorfer Schloss, 53115
Bonn, Germany
2 Institute Alfred Fessard, Centre National de la Recherche Scientifique,
91198 Gif sur Yvette, France
View larger version (40K):
[in a new window]
|
Fig. 1. (A) Schematic drawing of the experimental set-up, a goldfish and a vortex
ring. The pipette tip used to generate the vortex is at the left. The
elevation of the pipette (y-axis) was always at the height of the
fish's trunk lateral line canal. The rostro-caudal distance (x-axis)
between the pipette tip and the neuromast was adjusted to 5 cm. The lateral
distance (z-axis) of the pipette tip to the operculum of the fish was
0.5 cm, if not stated otherwise. A laser sheet was used to illuminate the
particles added to the water. In all cases in which we obtained PIV data and
neuronal data, the laser sheet illuminated the x–y
plane (yellow), i.e. a plane that was perpendicular to the fish's
rostro-caudal axis. The x–y plane was placed as close
to the skin as possible. For a further characterization of the water motions
caused by a vortex ring, the laser sheet was also positioned in the
z–x plane (red) or in the z–y
plane (gray). (B) Vectors (arrows) and vorticity (blue, counterclockwise; red,
clockwise) of the particle motions recorded 400 ms after valve opening
measured in the horizontal (z–x) plane (see A) in the
absence of a fish. (C,D) Velocities (v) directed parallel to the
x-axis, calculated from the vectors highlighted by the longitudinal
boxes in B. The plot in C transects the vortex through its midline and reveals
the central flow. The plot in D transects the laser sheet in the
z-axis. Here, the central flow is bordered by two vortex cores of
minimal velocity and a region where the flow direction is opposite to the jet
flow. The diameter of the vortex ring is approximately 3 cm.
|
|
View larger version (54K):
[in a new window]
|
Fig. 2. Vector plots of particle motions in the vertical plane recorded 80 ms
(top), 200 ms (middle) and 440 ms (bottom) after valve opening. Vector plots
are based on (A) individual and (B) averaged trials (n=9). Arrows
indicate the direction and velocity of particle motions. Note that the
averaged plots are similar to the plots based on individual trials, i.e. that
the vortex stimulus was reproducible from trial to trial. Scale vector, 12.5
cm s-1. In this figure, as well as in Figs
7,
8,
9,
11,
12, the fish schematic
indicates the size, location and orientation of the fish relative to the
vector plots.
|
|
View larger version (53K):
[in a new window]
|
Fig. 3. (A) Averaged (n=10) vector plots based on particle motions
recorded after valve opening at the times indicated. The orientation of the
laser sheet was orthogonal to the propagation direction of the stimulus
(z–y plane in Fig.
1A). The distance between the pipette and the laser sheet was
adjusted to 5 cm. Scaling vectors (large horizontal arrows) represent 12.5 cm
s-1. For better visualization, the vectors are enlarged in the
first three images. For parallel pressure measurements, a hydrophone was
placed 5 cm away from the pipette tip. (B) Mean changes in hydrodynamic
pressure (n=10) caused by the vortex ring. Inset: pressure changes,
measured 100–400 ms after valve opening, at an expanded time scale. (C)
Mean water velocity (n=10) measured with an anemometer. In B and C,
zero indicates the time of valve opening. Gray shaded areas in B and C
represent ±1 s.d.
|
|
View larger version (53K):
[in a new window]
|
Fig. 4. Discharge patterns of goldfish posterior lateral line nerve
(PLLN)-afferents to the vortex stimulus passing the fish laterally from
anterior to posterior. Afferents responded with an initial increase
(E-afferents, A) or decrease in discharge rate (I-afferents, B). In A and B,
raster plots of the responses to 10 (A) and 8 (B) stimulus presentations and
their corresponding peri-stimulus–time histograms (PSTHs; binwidth 20
ms) are shown. Top traces in A and B show original recordings that correspond
to the respective first trace in the raster plots. Horizontal bars below the
PSTHs indicate the time of valve opening. The initial response components were
fairly reproducible from trial to trial. This was quantified by calculating
the correlation coefficients between individual trials (n=5) for the
first (A and B, top right) and second (A and B, bottom right) 300 ms following
valve opening.
|
|
View larger version (19K):
[in a new window]
|
Fig. 5. The responses of an E-afferent (A) and an I-afferent (B) to the vortex
stimulus passing the fish laterally at the distances indicated. Each
peri-stimulus–time histograms (PSTH) is based on eight trials.
Horizontal bars represent time of valve opening. Gray vertical lines indicate
the time of response onset of the top traces (lateral distance 0.5 cm).
|
|
View larger version (23K):
[in a new window]
|
Fig. 6. Response latencies (A) and peak latencies (B) as a function of lateral
(z-axis) pipette distance. Black circles, E-afferents; gray circles,
I-afferents. Note that B contains only E-afferents. Linear regression lines
were calculated across the data (E-afferents, solid lines; I-afferents, dotted
line). See Results for the equation for each line.
|
|
View larger version (37K):
[in a new window]
|
Fig. 7. E-afferent response and PIV. (A) Averaged (n=8) vector plots based
on particle motions–measured in the vertical plane–caused by the
vortex stimulus. Particle motions were recorded (from left to right)
immediately before first threshold crossing, at first and second threshold
crossing, and 200 ms after second threshold crossing. Gray bars indicate the
rostro-caudal position of the neuromast. (B) Raster plots of the responses to
a vortex stimulus and (C) peri-stimulus–time histograms (PSTHs; bin
width 20 ms) based on these raster plots. In this figure and in Figs
8,
9, the vertical gray bars
indicate the neural activity recorded during the 40 ms time windows on which
the respective PIV images are based.
|
|
View larger version (30K):
[in a new window]
|
Fig. 8. I-afferent response and PIV. (A) Averaged (n=8) vector plots based
on particle motions caused by the vortex stimulus. Particle motions were
recorded (from left to right) immediately before first threshold crossing, at
first and second threshold crossing, and 200 ms after the second threshold
crossing. (B) Raster plots of the responses and (C) peri-stimulus–time
histograms (PSTHs; bin width 20 ms) based on these raster plots. See
Fig. 7 for further details.
|
|
View larger version (51K):
[in a new window]
|
Fig. 9. E-afferent response (A) and I-afferent response (B). (Top) Averaged
(n=8) vector plots based on particle motions caused by the vortex
stimulus. Particle motions were recorded (from left to right) at first, second
and third threshold crossing. (Middle) Raster plots of the responses and
(bottom) peri-stimulus–time histograms (PSTHs; bin width 20 ms) based on
these raster plots. Note that changes in the initial response components from
excitation to inhibition are associated with approximate 180° changes in
flow direction at the rostro-caudal level of the neuromasts (gray bars on fish
schematic). See Fig. 7 for
further details.
|
|
View larger version (32K):
[in a new window]
|
Fig. 10. Flow velocities (red and blue curves) calculated from interrogation windows
(see small squares on fish drawings). Interrogation windows with the red
vectors were at the position of the neuromast. The interrogation windows with
the blue vectors were slightly more rostral (A) or caudal (B), respectively.
The black traces in A and B show the discharge rates (spikes/40 ms) obtained
in a single trial of the E-afferent (A) and I-afferent (B) whose data are
shown in Fig. 9. Negative
velocities represent flow in the tail-to-head direction; positive velocities
represent flow in the head-to-tail direction. Gray bars indicate the times
when the first and third reproducible response components occurred. Zero
indicates the time of valve opening.
|
|
View larger version (32K):
[in a new window]
|
Fig. 11. (A) (Top) Averaged (n=8) vector plots based on particle motions
caused by the vortex stimulus passing the fish laterally. Particle motions
were recorded during the reproducible bursts marked by the first vertical gray
bar and during the following period of decreased neural activity marked by the
second vertical gray bar. (Middle) Raster plots of the responses to the vortex
stimulus and (bottom) peri-stimulus–time histograms (PSTH; bin width 20
ms) based on these raster plots. Note that the flow pattern during the bursts
marked by the first vertical gray bar was similar to the flow pattern that
occurred after the bursts (second vertical gray bar). (B) (Top) Raster plot of
the third trial shown in A. (Bottom) discharge rate (spikes per 40 ms, black
trace) calculated for that trial, and local water velocities (red trace)
obtained from the interrogation window indicated by the red vector in A. This
interrogation window is at the position of the neuromast. The blue trace shows
the first derivative of the velocity data (i.e. water acceleration). Note that
during the late bursts, small rhythmic fluctuations occurred in both local
water velocity and local water acceleration. (C) (Top) Fast Fourier
transformations (FFTs) of the discharge rate (trial 3 in A) and the
corresponding FFTs of the velocity (red; middle) and acceleration traces
(blue; bottom) shown in B. Note that both neural activity and water
acceleration have peaks at 3 and 8 Hz.
|
|
View larger version (41K):
[in a new window]
|
Fig. 12. (A) (Top) Averaged vector plots (n=7) based on particle motions
caused by the vortex stimulus, (middle) raster plots of the responses to the
vortex stimulus and (bottom) the peri-stimulus–time histograms (PSTH;
bin width 20 ms) based on these raster plots. Zero indicates the time of valve
opening. Averaged particle motions were recorded immediately before (left PIV
image) and during (right PIV image) the non-reproducible bursts marked by the
ellipses in the raster plot. (B) Discharge rates (black traces) and water
velocities (gray traces) measured during the first (left graph) and seventh
(right graph) trial shown in A. During the initial decrease in neural activity
(90–190 ms after valve opening), water velocity was positive in both
cases, i.e. water flow at the level of the neuromasts was from head to tail.
During the following increase in neural activity, the direction of water flow
reversed. No correlation between the occurrence of late bursts and water flow
direction was apparent. Asterisks indicate the bursts marked with ellipses in
A.
|
|
© The Company of Biologists Ltd 2006
