ABSTRACT
A dorsal light reflex is described in the squid Lolliguncula brevis. When illuminated from the side in visually homogeneous surroundings, a free-swimming squid rolls the dorsal side of its head and trunk 10–20 ° towards the light. With the trunk restricted in a holder, the squid rolls its head 4–5 ° towards the light; this reaction increases by about 50 % when the statocysts are bilaterally removed and increases further when the neck receptor organ is also destroyed. The results indicate a multi-modal interaction of visual, statocyst and proprioceptive inputs during postural control.
Introduction
The control of movement and posture in the three dimensions of space, especially during swimming and flight, requires sense organs and a reference system that allows such behaviour to be stabilized. In this respect, statocysts (invertebrates) and vestibular end organs (vertebrates) play the major role as sense organs, using the direction of gravity as a reference system (e.g. Platt, 1984; Budelmann, 1990). In contrast, most insects adjust their head and body positions during flight mainly according to visual cues (Hengstenberg, 1988, 1993), for example in response to the illumination field in the environment. They turn the dorsal surface of their head and trunk towards the direction of the brightest area of the overall illumination field (‘dorsal light reflex’; DLR). In most natural habitats, that direction is vertical and does not change. The same is true for the direction of gravity and, therefore, ‘dorsal light’ and gravity are ideal reference systems (see von Buddenbrock, 1915; Mittelstaedt, 1950; Stange, 1981; for further references, see Creutzberg, 1975; Wehner, 1981; Hengstenberg, 1993). A DLR also occurs in crustaceans, amphibians and fish, which use a combination of visual and gravity receptor inputs to stabilize posture and movement (von Holst, 1935; Schöne, 1951; Jahn, 1960; for further references, see Schöne, 1980; Meyer and Bullock, 1977). Some lower crustaceans show a reversed DLR: they turn their ventral side towards the light when swimming (‘ventral light reflex’; e.g. Seifert, 1932; for further references, see Wehner, 1981).
In cephalopods, a countershading reflex has recently been described that is driven not only by statocyst input but, in part, also by light (presumably by the light’s directional component; Ferguson et al. 1994). These results have revived the question of whether there is a DLR in cephalopods that is used in the stabilization of swimming and posture.
Materials and methods
The bay squid, Lolliguncula brevis (mantle length 60–80 mm), was used in this study. Its head and trunk position were videotaped while it swam in a glass aquarium (30 cm×30 cm×80 cm) supplied with recirculating sea water. To exclude visual cues, the animal was surrounded by an opaque cylinder (23 cm in diameter). The aquarium could be illuminated from either the left or the right side by a unit of four Philips 40 W fluorescent tubes, covered with a screen of milky Perspex. This arrangement created a light gradient with a centre of mean brightness inside the opaque cylinder on the left (or right) side of the animal and an angular light distribution, measured under water at the position of the animal, ranging from about 2800 lx (at the centre of illumination) to about 500 lx (at the opposite side).
In order to quantify the light-induced head roll response more precisely, a holder was designed that kept the trunk of the animal in its normal upright position (0 °) but allowed unrestricted movements of the animal’s head (for details of the holder, see T. Preuss and B. U. Budelmann, in preparation). Fixed to the holder, the animal was placed into the aquarium. Before each experiment, the animal was kept in dim red light (one Westinghouse 15 W darkroom bulb) for about 1 min. It was then exposed to 1 min periods of illumination (alternately from the left and right sides), followed by equal periods of red light to prevent visual adaptation. During illumination and red light exposure, the anterior view of the animal was videotaped for later analysis of the angular deviation of the animal’s head (head roll) from its normal (0 °) position. The head roll position was defined as the angular deviation from horizontal of a line drawn through the centre of both eyes. Each squid (N=3) was tested before and after bilateral statocyst and subsequent bilateral neck receptor organ destruction (for details of the operations, see Budelmann, 1990; T. Preuss and B. U. Budelmann, in preparation), with seven measurements of head roll position taken during each of the 1 min exposures to light.
Results and discussion
Free-swimming squid (N=5), when illuminated from the side, show a clear tendency to roll the dorsal side of their head and trunk 10–20 ° towards the light (Fig. 1A). The effect does not habituate for at least 1 h of constant light.
When restricted in a holder, intact animals, with their trunk in the normal (0 °) position, roll their head an average of 4–5 ° such that its dorsal side is inclined towards the centre of mean brightness (Fig. 2; roll to the right: 3.7±0.8 °, S.E.M., Student’s t=7.3, d.f.=16, P<0.01; roll to the left: 4.9±1.3 °, Student’s t=11.3, d.f.=24, P<0.01). This is much less than the 10–20 ° roll of the head (and trunk) shown by free-swimming squids. Part of this difference is due to the function of the neck receptor organ (T. Preuss and B. U. Budelmann, in preparation) and part presumably to the fact that the animals were restrained. No adaptation of the head roll response was ever seen during the 1 min periods of illumination.
Under the same experimental conditions but with the statocysts bilaterally destroyed, the animals roll their head about 7 ° towards the centre of mean brightness; this is about 50 % more than the roll of intact animal (Figs 1B, 2; roll to the right: 6.7±1.9 °; roll to the left: 7.2±2.4 °). Again, no adaptation of the head roll response was seen during the 1 min periods of illumination.
When the neck receptor organ, which controls head-to-trunk positions (T. Preuss and B. U. Budelmann, in preparation), is destroyed in addition to the statocysts, an even larger head roll towards the centre of mean brightness occurs (Fig. 2; roll to the right: 10.8±0.6 °; roll to the left: 7.9±0.5 °). Restrained animals sometimes not only roll their head but also bend their arms and funnel towards the direction of light (Fig. 1B); in a free-swimming squid, such a funnel position would result in further roll towards the light.
These experiments demonstrate that in the squid Lolliguncula brevis there is a DLR, and that the directional effects of light (most likely via the eyes) and gravity (via the statocysts), as well as input from the neck receptor organs, are integrated centrally to maintain normal orientation. The DLR of Lolliguncula was not very strong, at least compared with that of some fishes (von Holst, 1950); this might be squid-specific, but in fishes there is also great variation among different species (von Holst, 1935). It remains to be seen whether extra-ocular photoreceptors (Mauro, 1977) are involved in the DLR in cephalopods, as they sometimes are in insects, where the DLR is driven by the compound eyes and/or ocelli (see Wehner, 1981; Hengstenberg, 1993).
It is reasonable to assume that other squid species, and presumably also cuttlefishes, have a DLR to stabilize their swimming. It is more difficult to predict, however, whether octopods have such a reflex, since most species live in cavities and crevices of rocky habitats where the directional cue of light changes from one location to the next.
ACKNOWLEDGEMENTS
This work was supported in part by German LGFG and DAAD funds (T.P.) and in part by NIH grant EY 08312-02/05 (B.U.B.). Animals were provided through NIH grants RR 01024 and RR 04226 (Dr Roger T. Hanlon). The authors thank Dr J. B. Messenger (Sheffield) for comments on an earlier draft of the manuscript and T.P. thanks Professor D. Varjú (Tübingen) for additional support.