First published online October 5, 2007
Journal of Experimental Biology 210, 3616-3623 (2007)
Published by The Company of Biologists 2007
doi: 10.1242/jeb.004044
Visually guided obstacle avoidance in the box jellyfish Tripedalia cystophora and Chiropsella bronzie
A. Garm*,
M. O'Connor,
L. Parkefelt and
D.-E. Nilsson
Department of Cell and Organism Biology, Lund University,
Helgonavägen 3, 22362 Lund, Sweden
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 1. Experimental set-up. The animals were forced to encounter obstacles of
different color (black, red, blue, green, transparent, transparent with black
top) and size (1 or 2 cm diameter) in a flow chamber (A,B). In the case of
T. cystophora, three evenly spaced obstacles were present, and when
using C. bronzie, two obstacles were present (A). The brilliance of
the obstacle and the wall of the flow chamber was measured with a
spectrophotometer and weighted by the relative sensitivity of a 500 nm opsin
(C). The colors of the lines correspond to the colors of the obstacles; the
yellow line represents the wall of the flow chamber. The weighted brilliance
was turned into contrast between the obstacles and the wall of the chamber
(D).
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 2. Swim trajectories. During the 3 min in the flow chamber the medusae behaved
differently depending on the visual appearance of the obstacle. When high
contrast obstacles were present (A,B), the medusae swam against the current
and in general kept at a distance from them. When transparent obstacles with
or without a black top were present, the medusae did not respond to them and
either bumped into them and passed them (T. cystophora, C) or kept
swimming into them (C. bronzie, D). It should be noted that the
tracks represent the movements of the top of the bell and therefore contacts
with the obstacles by other parts of the bell can have taken place without the
track showing it.
|
|
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 3. Responses to large (2 cm wide) obstacles by T. cystophora. The
medusae performed clear avoidance responses to the colored obstacles but not
to the transparent ones (A–C, N=10). There were no significant
differences between the colored obstacles in avoidances per minute and
contacts per minute (A,B). The red obstacle induced avoidance responses
significantly further away than from any of the other obstacles (C,
**Fisher's PLSD, P 0.0001). Values are means and error
bars indicate s.e.m. Trans, transparent obstacle; Trans+B, transparent
obstacle with black top above water.
|
|
View larger version (28K):
[in this window]
[in a new window]
|
Fig. 4. Responses to differently sized colored obstacles by T. cystophora.
The thickness of the histogram bars follows the width of the obstacles. The
medusae responded more strongly to the wide obstacles than to the thin ones.
When comparing obstacles of the same color (e.g. red 2 cm vs red 1
cm), medusae had fewer contacts with (B) and performed avoidances further away
from (C) the wide obstacles. The rate of avoidance did not differ with
obstacle size (A). When converting the distances of avoidance into visual
angles none of the pairs differed significantly (D). N=10 except for
the thin green obstacle where N=6. Values are means and error bars
indicate s.e.m. *P values between 0.05 and 0.0001 from
ANOVA with Bonferroni–Dunn post hoc test.
**P values <0.0001.
|
|
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 5. Response to large obstacles by C. bronzie. As found for T.
cystophora (see Fig. 3),
there are major differences between the response to the colored obstacles and
the response to the transparent obstacles showing that the obstacle avoidance
response is visually guided (A–C). There are also indications that the
strength of the response follows the intensity contrast between the obstacle
and the chamber wall (A–C; see Fig.
1D). The medusae performed more avoidances (A) and had less
contact (B) when presented with the high contrast black obstacles (for
statistics see Results). There was no significant difference between the
responses to the two types of transparent obstacle (P>0.65).
N=10 for A and B, but since not all medusae displayed avoidances,
N values vary in C (shown as numbers in bars). Values are means and
error bars indicate s.e.m. Trans, transparent obstacle; Trans+B, transparent
obstacle with black top above water.
|
|
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 6. Response to differently sized colored obstacles by C. bronzie. The
thickness of the histogram bars follows the width of the obstacles. Again, the
wide obstacles provoked the strongest responses (A–C). Only a weak
response was seen when the medusae were presented with thin obstacles and the
medusae spent up to 31% of the time in contact with them (blue bar in B). The
thin obstacles evoked a total of only eight avoidances (C). When mean distance
for avoidance is transformed into visual angles similar values are seen across
obstacle size (D). There is some more variation in the visual angles but this
is probably due to the very low value for N (1 or 3) for the thin
obstacles. N=10 for wide obstacles and 6 for thin obstacles, unless
shown otherwise in the bars (repeats in A and B; recorded events in C and D).
Values are means and error bars indicate s.e.m. *Significant
difference at the 0.10 level from t-test, **significant
differences at the 0.05 level.
|
|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati 
Twitter What's this?
© The Company of Biologists Ltd 2007
