Escape manoeuvres in damsel-fly larvae: kinematics and dynamics
Department of Anatomy, University of Cambridge, Downing Street, Cambridge CB2 3DY, UK
(e-mail: jhb1000{at}cam.ac.uk)
Accepted 15 October 2002
![]() |
Summary |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: larva, escape manoeuvre, unsteady swimming, caudal fin, kinematics, vortex wake, control of locomotion, damsel-fly, Enallagma cyathigerum
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In a previous study on the hydrodynamics of steady swimming in damsel-fly
larvae (Brackenbury, 2002), the
author noted a rapid escape response similar in kinematics and dynamics to the
rapid C-start of fish. Subsequently, a related, although more complex, rapid
escape manoeuvre has also been identified that involves specialised movements
within the three-lobed caudal fin. The present study was undertaken to analyse
the dynamics of the escape responses of damsel-fly larvae against a background
of the knowledge of steady swimming gained from the earlier investigation.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
During experiments, individual larvae were placed in a transparent
container measuring 16 cmx8 cmx8 cm (width x depth x
height) containing water to a depth of 6-7 cm. The method of flow
visualisation follows that previously described
(Brackenbury, 2002) using fresh
cow's milk as a dye tracer. Both in the wild and in the laboratory,
Enallagma cyathigerum larvae adopt a bottom-dwelling lifestyle
because their bodies are denser than water. Steady locomotory behaviour is of
two kinds: horizontal swimming, with the body held just clear of the bottom,
and open-water swimming, which may be upwards, downwards or horizontal. In
addition, swimming may be `fast', with the legs trailing to minimise drag, or
slow, with the legs performing rhythmical paddling movements. Escape
manoeuvres are nearly always followed up with fast, bottom swimming;
consequently, the open-water dye-streamer technique employed in the earlier
study (Brackenbury, 2002
) was
not appropriate and all results were obtained using the bottom-layer
technique. A thin layer of milk (approximately 1 mm) was carefully laid down
on the bottom of the container using a syringe, and larvae initiating an
escape response left behind a trail as evidence of the wake. The trails were
filmed from directly above the container, care being taken to reject any
examples in which there was evidence of direct contact between the body and
the tracer once the larva had begun its movement. In a second, shorter series
of experiments, kinematic data were collected with the camera filming directly
from the side of the container. In total, approximately 40 individuals were
used to collect data and the results are presented as mean values ± 1
S.D. with the number of observations (N) in brackets.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The second manoeuvre can be termed the `flex and twist'. As with the simple
flex, a larva reacting to a stimulus applied to its left side appears to flex
to its right (Fig. 2A,
3A) before extending in the new
direction. In fact, the insect flexes to its left [an ipsilateral or
`strong-side' flexure to adopt the terminology of Drucker and Lauder
(2001a)], but this movement is
masked by a simultaneous helical twisting of the body that induces pitching
and rolling instability as well as yaw instability. The sequence of events, as
viewed from above, is as follows. The first observable movement is a rapid
dorsiflexion that elevates the tail and begins to pull the anterior end of the
body and the legs away from the substrate (20 ms stage in
Fig. 2A). Simultaneously, the
caudal end of the body, including the tail fin, begins to twist clockwise, as
viewed from behind (20 ms stage in Fig.
2A, curved arrows), and this twist is relayed cranially along the
body axis, rolling the larva onto its right side (40 ms and 60 ms stages in
Fig. 2A). Thus, the impression
of right-side flexure is created by a combination of dorsiflexion plus the
fact that the larva has rolled onto its right side. Already by this stage, the
body is beginning to flex along its left side. Note that this is the opposite
to the simple flex response where lateral bending is to the right not the
left. The direction of the force experienced by the tail plates, as inferred
from their angle of attack at this stage, is roughly opposite to the line of
motion of the head (Fig. 2B,
3B), suggesting that the
anterior part of the body is being pivoted about its centre of gravity into
its new alignment. At the end of the re-alignment stage, the insect lies on
its side, facing in the intended direction of escape, with its tail fin in
contact with the head (60 ms stage in Fig.
2A). Extension then drives the body in the direction of escape
115.8±25.3° (N=20) away from the original heading
(Fig. 7B). The latter stage is
kinematically identical to the extension phase of the simple flex manoeuvre,
and both culminate in the shedding of a thrust jet (see below) behind the
retreating body. Front and side views of the twisting larva
(Fig. 4) show the vertical
displacements of the body and also give a clearer view of the onset of the
left-side body flexure as soon as the insect has rolled onto its side
(Fig. 4A, 60 ms stage).
Throughout the manoeuvre, the geometric centre of the body remains close to
the substrate and the larva exits from the manoeuvre (100 ms stage in Figs
2,3,4)
with its body parallel to the bottom and its body centre separated from the
bottom by a distance of 0.25±0.09 body lengths (N=12). Thus,
the twist manoeuvre, like the simple flex, prepares the body for sideways
escape with little or no vertical component.
|
|
|
|
The duration of flexion phase, measured to the point where the tail fin makes contact with the side of the head, was 77±20 ms (N=20). The duration of the extension phase, measured to the point where the tail fin reaches its maximum excursion to the opposite side of the body to the initial flexure, was 79±12 ms (N=20).
In contrast with the simple flex manoeuvre, in which all three constituent lobes of the tail fin bend uniformly in response to largely unidirectional side forces applied to the fin, a degree of independence of movement between the plates is manifest during the twist and flex. At rest, the lateral plates diverge to the sides of the body with little indication of any flexion between apical and basal segments (Fig. 5, 0 ms stage). Tail elevation (Fig. 5, 20 ms and 40 ms stages) leaves the median plates unaffected but flexes the apical segment of the lateral plates outwards (i.e. in the direction of abduction) relative to the basal segments. During the twist (Fig. 5, 60-100 ms stages), lateral forces flex all the apical segments in one direction relative to the base but are evidently too weak to force the plates together in the median plane. Finally, during the flex and extension (Fig. 5, 120-180 ms stages), strong lateral forces, similar to those experienced during steady swimming, result in compression of the plates together and identical bending at the base and the flexion line.
|
Hydrodynamics
Within 20-40 ms of the completion of the extension phase of both
manoeuvres, a thrust jet could be seen impinging on the bottom layer of the
tracer and propagating away from the body
(Fig. 6). Quantitative data
were collected for the twist manoeuvre only, which was the commoner of the two
responses (Fig. 7A,B). The
momentum angle of the jet in the horizontal plane was 24±16.5°
(N=20) measured relative to dead-aft. The jet was associated with a
single, discrete ring vortex, the opposite cores of which could often be
identified (Fig. 3C) although
images were not detailed enough to allow accurate measurement of the core
radius. The measured ring radius R, the external radius orthogonal to
the ring plane Ra, and the external radius parallel to the
ring plane Rp (Fig.
7) were 0.0025±0.0003 m (N=14),
0.0028±0.0003 m (N=14) and 0.004±0.001 m
(N=14), respectively. The velocity of the axial jet
(Vjet) measured at the ring plane within 20-40 ms of the
shedding of the vortex was 13.1±2.0 cm s-1 (N=16;
Fig. 7B). The velocity of the
head (V) in the direction of escape measured over the same period was
21.6±5.4 cm s-1 (N=15;
Fig. 7B). There were
insufficient data to allow a velocity profile to be established along the
vortex axis.
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It is clear from the present study that the behaviour of the tail fin during steady swimming represents only a fraction of the potential flexibility of this organ. The twist manoeuvre, in particular, highlights the degree of independence of the constituent lobes of the fin. Hinge-like movements of the plates in the transverse plane of the body can occur both at the base and along the flexion line. The simple flex manoeuvre is based on the same kind of sideways motion of the tail that occurs during steady swimming, which imposes mainly lateral forces on the fin elements and causes them to bend as a unity throughout their length. The twist manoeuvre, by contrast, subjects the tail to torque (during the twist) and tangential (during elevation) forces. Tangential forces, produced by movement of the tail fin within the median plane, tend to splay the lateral elements outwards from the base as well as along the flexion line (20 ms and 40 ms profiles in Fig. 5B), while torque imposes a combination of tangential and weak lateral forces resulting in the profiles depicted in the 40-100 ms stages shown in Fig. 5B. The strong lateral forces experienced during the final stages of the flex (120 ms and 140 ms stages) and during extension (160 ms and 180 ms stages) result in a grouping and bending regime similar to that shown during steady swimming. The above is a very simplified analysis of tail fin function based solely on kinematic data and without the benefit of detailed information about the flow field around the tail fin. Nor can the possibility of some degree of active control of the tracheal lobes from the base be discounted.
As Drucker and Lauder
(2001a) have argued, animals
probably spend much more time executing short-term manoeuvres such as
stopping, starting, turning, braking and negotiating obstacles than they do in
performing steady movement. Their investigations in the sunfish showed that
turning is not simply a modification of steady swimming but involves
specialised manoeuvres by the pectoral fins. A stimulus administered to one
side of the body led to rapid abduction of the strong-side fin, which produced
a rotation with little or no translation. The contralateral or weak-side fin
produced a delayed thrust jet that accelerated the body away from the
stimulus. The rapid C-start manoeuvre of fish, which often involves large
turning angles (Domeninci and Blake, 1993, 1997;
Blake et al., 1995
), also
involves a re-orientational stage (rapid flex) followed by a translational
phase (extension), and parallels have already been drawn between this
manoeuvre and the flex response of damsel-fly larvae
(Brackenbury, 2002
). Whilst in
the sunfish these sequential movements are brought about by independent
activity in the pectoral fins, in C-starting fish and rapid-flexing damsel-fly
larvae re-orientation and translation represent a continuum in the movements
of caudal fin and body. Although nothing can be said directly in the present
study concerning the neuromotor programmes responsible for executing rapid
evasion manoeuvres, significant questions arise from a simple comparison of
kinematic events. The common factor linking C-starting fish and rapid-flexing
damsel-fly larvae is an initial weak-side response [the `away response' of
Domenici and Blake (1997
)]:
contraction of the weak-side flexors produces the C-shaped cavity into which a
thrust jet is drawn and eventually shed `into the face' of the threatening
stimulus. Analysing the hydrodynamics of the C-start, Wolfgang et al.
(1999
) have shown how fluid is
drawn into the concavity as two oppositely signed vortices form simultaneously
around the head and tail ends of the body. Rapid straightening then massages
the vortices caudally until they are shed into the wake as a ring vortex with
a powerful central jet (Fig.
8). Although the translational phases of the rapid flex and twist
responses of damsel-fly larvae appear to be functionally equivalent, this
cannot be said about the preparatory re-orientational phases, which involve
different muscle-activation patterns driven by independent neuromotor
programmes: the C-shaped cavity of the twisting larvae at the functionally
equivalent stage (Fig. 8, stage
2) has been formed via an initial strong-side, not weak-side,
flexure. The cavity faces forward in the direction of escape only because the
body has rolled and yawed away from the stimulus. The twist is self-evidently
a much more complex manoeuvre than the rapid flex, involving what is
tantamount to a helical contraction of the body to bring about the required
re-orientation.
|
Any biomechanical advantages of the twist over the simple flex are not clear because, as will be shown in the following section, the impulse provided by the thrust jet during a rapid twist escape is only approximately 50% more than that produced during a half-stroke of normal steady swimming. Nor is the twist capable of producing an appreciably faster response, or a greater turning angle, than the rapid flex. The real difference may lie in a more subtle area of the insect's biology viz. behaviour. For example, in some situations, it might benefit the insect to present its ventral side, including its legs, to the potential enemy as it makes its escape, rather than its flank. In these circumstances, the insect would choose to initiate the twist and not the simple flex. For the moment, this is merely speculation, but such ideas may be useful in framing the questions to be posed in any future behavioural study of escape responses in these insects.
A study on rapid C-start behaviour in larval Chinook salmon
(Oncorhynchus tschawytscha; Hale,
1996) makes it possible to compare the kinematics of a virtually
identical manoeuvre in a vertebrate and invertebrate of similar body length
(2-3 cm). With such a complex movement, it is not possible to assign a single,
whole-body Reynolds number. In the case of the final-stage damsel-fly larva
(body length approximately 2 cm), an upper estimate can be arrived at during
the extension phase at the end of which the head velocity reaches 0.22 m
s-1. Assuming a characteristic length of 0.02 m, the resultant
Reynolds number is approximately 4.5x103. This is somewhat
higher than Hale's estimate for juvenile salmon, although both animals clearly
lie in the zone of `intermediate' Reynolds numbers. In other respects, C-start
performance is remarkably similar between the insect and the fish. The total
duration of the manoeuvre (stage 1 plus stage 2) was approximately 150-200 ms
in each case, although stage 1 in the fish (60 ms) was shorter than in the
insect (77 ms). The C-start in the fish is technically equivalent to the flex,
described in the present study, as opposed to the `flex and twist', although,
as was shown earlier, these are very similar. One of the most interesting
findings of Hale's study was that the relationship between fast-start
performance parameters and body length differs between adult and juvenile
fish. Experimental studies have shown that high-acceleration manoeuvres are
important for predator avoidance in both young
(Eaton et al., 1977
) and
mature (Webb, 1982
;
Fuiman, 1993
) fish. However,
in adults, C-start duration increases with body length when comparisons are
made both within and between species (Webb,
1976
,
1978
), whereas in larval
Chinook salmon C-start duration decreases with age and size. Hale
(1996
) suggests that the
transition from the larval to the adult relationship probably coincides with
the point where the yolk sac reserves have become depleted and the larval fish
begins active foraging, exposing it to heightened predator threat. According
to this analysis, C-start performance peaks during early larval development.
Whether or not the relative speed of escape response varies throughout the
development of larval damsel-flies is not known, as the present study deals
only with the final instar. Clearly, the ontogeny of rapid escape manoeuvres
would be an important area to address in further studies on this insect.
Hydrodynamics
Although the bottom-layer technique is suitable for visualising flow events
associated with an organism in close proximity to the substrate, its main
weaknesses are that it presents only a two-dimensional impression of a
three-dimensional wake structure and it is difficult to quantify with
accuracy. These deficiencies were addressed in a previous study where wake
modelling from bottom-layer and open-water dye-streamer techniques was
compared during steady swimming
(Brackenbury, 2002). Both were
consistent with the picture of a wake consisting of single, discrete ring
vortices shed alternately to the left and right side of the swimming path with
consecutive half-strokes. The momentum angle of the vortex jets was
approximately 70° relative to dead-aft, implying considerable side-forces
generated by the body and tail. Wake momentum and thrust estimates were
consistent with values predicted from kinematic data using a standard
bulk-momentum model for large-amplitude, elongated-body fish locomotion
(Lighthill, 1971
). The ring
vortices produced during steady swimming, as visualised by the bottom-layer
and open-water techniques, were slightly smaller and had a jet velocity
slightly lower than the thrust vortex identified at the end of the twist
manoeuvre in the present study.
These comparisons are important in view of the lack of open-water data in
the present study against which to check the results of the bottom-layer
technique. The larva exits from the twist manoeuvre with its body held within
0.5 cm of the bottom, making it almost inevitable that the wake will impinge
on the bottom: after all, that is the basis of the technique and, regardless
of possible ground effects on the wake, this is the normal wake pattern
produced by the larva in these circumstances. In some cases, the measured
external diameter of the vortex was based on direct impressions within the
milk layer; in others, such as that illustrated in
Fig. 3C, the geometry of the
vortex was clearly shown as milk was drawn into its trailing edge. In the
latter instances, the lower edge of the vortex was just (approximately 1 mm)
clear of the bottom. This kind of variability precludes any precise
quantification of ground effects on the shape, size, velocity and symmetry of
the wake, but comparison with the previous study
(Brackenbury, 2002) suggests
that the error in the measurement of the dimensions of the vortex is probably
less than 25%. The momentum M of an axi-symmetric ring vortex can be
estimated as:
![]() | (1) |
![]() | (2) |
The rapid flex and twist manoeuvres are single-stroke movements distinct
from the rhythmical sinusoidal movements involved in steady swimming, but
there are also clear dynamic and kinematic similarities. As in the individual
half-strokes of steady swimming, the escape manoeuvre generates a single
thrust vortex. In the former case, there was clear evidence of body as well as
well as tail-generated flows that combined to produce the final vortex shed
from the caudal fin. The very large lateral undulation of the body and the
tail resulted in vortices being shed at nearly 70° away from dead-aft. The
wake structure of continuously swimming damsel-fly larvae differed from the
high-efficiency reverse Karman vortex sheet, with its caudally directed
zigzagging jet, characteristic of some caudal fin swimming fish
(Blickan et al., 1992;
Muller et al., 1997
;
Wolfgang et al., 1999
), and
was much more like that of the intermittently swimming Zebra Danio
Brachydanio rerio (McCutchen,
1977
) and the continuously swimming eel Anguilla anguilla
(Muller et al., 2001
). The
digital particle image velocimetry images of Muller et al.
(2001
) showed a double vortex
shed to the side of the swimming path with each half-stroke, which was
interpreted as a ring vortex formed by contributions from a body wave (termed
the `protovortex' by these authors) and movements of the tail. As discussed by
Brackenbury (2002
), the
rapid-flex manoeuvre of startled damsel-fly larvae can be modelled on the fast
C-start of fish (Fig. 8). Flow
visualisation studies by Wolfgang et al.
(1999
) show that rapid
curvature of the backbone into a C-shape draws fluid into the C-shaped cavity
as two oppositely signed vortices develop simultaneously around the head and
tail ends of the body. Rapid straightening then leads to these vortices being
shed into the wake as a thrust jet that drives the fish in the direction of
the initial flexure. By forming the body into a C-shape, the damsel-fly larva
ensures that the resultant thrust jet is directed approximately due aft of the
realigned anterior part of the body.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Blake, R. W., Chatters, L. M. and Domenici, P. (1995). Turning radius of yellow fin tuna (Thurnos albacures) in unsteady swimming manoeuvres. J. Fish. Biol. 46,536 -538.
Blickan, R., Krick, C., Zehren, D. and Nachtigall, W. (1992). Generation of a vortex chain of a subundulatory swimmer. Naturwissenschaften 79,220 -221.
Brackenbury, J. H. (2001). The vortex wake of
the free-swimming larva and pupa of Culex pipiens (Diptera).
J. Exp. Biol. 204,1855
-1867.
Brackenbury, J. H. (2002). Kinematics and
hydrodynamics of an invertebrate undulatory swimmer: the damsel-fly larva.
J. Exp. Biol. 205,627
-639.
Domenici, P. and Blake, R. W. (1991). The kinematics and performance of the escape response in the angelfish (Pterophyllum eimeki). J. Exp. Biol. 156,187 -205.
Domenici, P. and Blake, R. W. (1993). The effect of size on the kinematics and performance of angelfish (Pterophyllum eimeki) escape responses. Can. J. Zool. 71,2319 -2326.
Domenici, P. and Blake, R. W. (1997). The
kinematics and performance of fish fast-start swimming. J. Exp.
Biol. 200,1165
-1178.
Drucker, E. G. and Lauder, G. V. (2001a). Wake
dynamics and fluid forces of turning manoeuvres in sunfish. J. Exp.
Biol. 204,431
-442.
Drucker, E. G. and Lauder, G. V. (2001b).
Locomotor function of the dorsal fin in teleost fishes: experimental analysis
of wake forces in sunfish. J. Exp. Biol.
204,2943
-2958.
Eaton, R. C., Farley, R. D., Kimmel, C. B. and Schabtach, E. (1977). Functional development in the Mauthner cell system of embryos and larvae of the zebrafish. J. Neurobiol. 8, 151-172.[Medline]
Fuiman, L. A. (1993). Development of predator evasion in Atlantic herring, Clupea harengus. Anim. Behav. 45,1101 -1116.[CrossRef]
Hale, M. E. (1996). The development of fast-start performance in fishes: escape kinematics in the Chinook salmon (Oncorhynchus tshawystscha). Am. Zool. 36,695 -709.
Lauder, G. V. (2000). Function of the caudal fin during locomotion in fishes: kinematics, flow visualisation and evolutionary patterns. Am. Zool. 40,101 -122.
Lighthill, M. J. (1971). Large-amplitude elongated-body theory of fish locomotion. Proc. R. Soc. Lond. B 179,125 -138.
McCutchen, C. W. (1977). Froude propulsive efficiency of a small fish, measured by wake visualisation. In Scale Effects in Animal Locomotion (ed. T. J. Pedley), pp. 339-363. London, New York, San Francisco: Academic Press.
Muller, V. K., Van den Heuvel, B. L. F., Stamhuis, E. J. and
Videler, J. J. (1997). Fish foot-prints: morphometrics and
energetics of the wake behind a continuously swimming mullet (Chelon
labrosus Risso). J. Exp. Biol.
200,2893
-2900.
Muller, V. K., Smit, J., Stamhuis, E. J. and Videler, J.
(2001). How the body contributes to the wake in undulatory fish
swimming: flow fields of a swimming eel (Anguilla anguilla).
J. Exp. Biol. 204,2751
-2762.
Spedding, G. R., Rayner, J. M. V. and Pennycuick, C. J. (1984). Momentum and energy in the wake of a pigeon (Columba livia) in slow flight. J. Exp. Biol. 111,81 -102.
Webb, P. W. (1976). The effect of size on the fast-start performance of rainbow trout Salmo gairdner: and a consideration of piscivorous predatorprey interactions. J. Exp. Biol. 65,157 -177.[Abstract]
Webb, P. W. (1978). Fast-start performance and body form in seven species of teleost fish. J. Exp. Biol. 74,211 -226.
Webb, P. W. (1982). Avoidance responses of flathead minnow to strikes by four teleost predators. J. Comp. Physiol. 147,371 -378.
Wolfgang, M. J., Anderson, J. M., Grosenbaugh, M. A., Yue, D. K. P. and Triantophyllou, M. S. (1999). Near-body flow dynamics in swimming fish. J. Exp. Biol. 202,2302 -2327.