Polarization sensitivity in the red swamp crayfish

Procambarus clarkii enhances the detection of a moving transparent object

John Tuthill

Abstract

We tested the hypothesis that polarization sensitivity enhances the detection of a moving, transparent object by examining the escape response of the red swamp crayfish ( Procambarus clarkii ) from a visual threat. A transparent, birefringent target trans-illuminated by either partially linear polarized or unpolarized light was advanced toward individual crayfish. The optical axis of the target was aligned such that it would be conspicuous to a viewer with polarization sensitivity when trans-illuminated by polarized light. Under polarized light, significantly more crayfish retreated from the target than under unpolarized light of identical intensity (p<0.0001, Chi-square). While the potential for polarization sensitivity has been shown in neurophysiological and structural studies of the visual system of P. clarkii and the signal crayfish Pasifastacus leniusculus , our results provide the first behavioral evidence for polarization sensitivity among crayfish. The ecological function of this ability is unclear, but may enhance the detection of fish with mirror-like scales, transparent zooplankton, or macroinvertebrates. Because escape responses are generally more reliably induced than other behaviors, the method employed in this study may prove useful for examining sensory capabilities in other species.

2 Introduction

Polarized light in nature

In addition to wavelength and intensity, light possesses a third fundamental property that

depends on the plane of polarization, often represented by the angle of the electric vector. The

polarization of an electromagnetic wave is defined as the distribution of the planes of vibration

of the electrical fields within the wave. A light beam oscillating in more than one plane is

considered unpolarized, or depolarized. Linearly polarized light contains only a single e-vector

direction, meaning that its oscillation occurs within a single plane. Natural light is partially

polarized, consisting of a mixture of fully linearly polarized light with a common e-vector angle,

and fully depolarized light with indiscriminate e-vector orientation. The percentage of the total

field that is comprised by the fully polarized component is defined as the percent polarization. A

partially polarized light field is therefore characterized by three properties: intensity, percent

polarization, and dominant e-vector.

Despite the fact that the sun produces fully depolarized light, partially linearly polarized light is abundant in nature. Rayleigh scattering of sunlight in the atmosphere and water column produces a robust polarization pattern that varies with solar position and direction of view.

Underwater, the illumination pattern of the sky is limited to within 48 degrees from the vertical, a region called Snell’s window. Inside Snell’s Window, the polarization pattern is influenced by the same factors that determine the sky polarization, such as sun position, cloudiness, distance from the zenith, and depolarization due to anisotropy of air molecules (Shashar et al., 2004).

Outside of Snell’s window, e-vector orientation depends primarily on Rayleigh scattering by submicroscopic particles in the water (Waterman, 2006). The polarization pattern in this region is perpendicular to depolarized light from the sun that is refracted at the air/water interface and

3 scattered in the water column. This e-vector distribution varies with the sun’s refracted zenith angle as well as the viewer’s line of sight, resulting in a complex polarized light field where the perpendicular e-vectors tilt varies with solar position and direction of view (Figure 1). For example, polarization is generally strongest in the horizontal viewing direction with an approximately horizontal e-vector (Shashar et al., 2004) when the sun is at the 0 ° zenith position

(at noon). In addition to polarization by scattering, the reflection of light from shiny, dielectric surfaces produces strong polarization in some circumstances (Wehner, 2001).

Figure 1. Effect of the solar zenith angle on sky and underwater e-vectors, as the sun moves from sunrise to ideal noon. As the sun continued across the sky, the figure would reflect the right mirror image of the diagram. The tilt of the sky and underwater polarization planes are shown for three different solar zenith angles (adapted from Waterman, 2006).

Sensitivity to polarized light in animals

Many animals have specific membrane structures that differentially analyze the polarization of incoming light. The ability to distinguish the e-vector of light, or polarization

sensitivity, occurs in an eclectic variety of both terrestrial and aquatic species, and serves a

variety of functions. Karl von Frisch was the first to demonstrate that honeybees could sense the

4 polarization pattern of the sky, and that they used the e-vector pattern as a compass for navigation while foraging (Frisch, 1967). In honeybees, desert ants, and certain other insects, the dorsal rim of the eye functions specifically to detect the overhead e-vector pattern (Labhart and

Meyer, 2002). The same consistent polarization cues are exploited by aquatic organisms for orientation and navigation. In the characteristic “shore-flight” behavior, Daphnia swim toward light with a higher degree of polarization, regardless of intensity (Schwind, 1999).

Polarization sensitivity may also be used to enhance the visibility of transparent or otherwise-camouflaged animals in water (Lythgoe and Hemmings, 1967). Octopods are able to recognize contrast produced by e-vector variation within a single imaged object, indicating that they use object detection based on polarized recognition as a means of contrast enhancement

(Shashar and Cronin, 1996). Cuttlefish use polarization-based contrast enhancement for detecting transparent prey (Shashar et al., 1998), as well as for breaking the countershading camouflage of light-reflecting silvery fish (Shashar et al., 2000). Some species, such as cuttlefish and stomatopods, may exploit polarization signals for intraspecific recognition and communication (Marshall et al., 1999; Shashar et al., 1996). These experiments demonstrate that animals use patterns of polarized light for image formation rather than just orienting stimuli, and that polarization sensitivity provides some selective advantage for the animals that possess it.

Dichroic visual pigment molecules are inherently sensitive to the e-vector orientation of linearly polarized light due to the fact that pigment molecules are maximally excited when the e- vector is parallel to the linear absorption dipole of the chromophore (Waterman, 1981). Because the visual pigments are not fixed in a specific orientation within the flat membranes of rods and cones, most vertebrate photoreceptors are unable to discriminate the e-vector of light. The mechanisms by which some vertebrate species distinguish e-vector is poorly understood

5 (Flamarique et al., 1998). In contrast, the visual pigments in invertebrate compound eyes with rhabdomeric photoreceptors are aligned such that they preferentially absorb light of a specific e- vector.

Polarization sensitivity in crayfish

Among invertebrates, the structural and neural basis for polarization sensitivity is particularly well-documented in crustaceans, especially the stomatopods (Marshall, 1988) and crayfish (Glantz, 2001). The ommatidia of many crustacean compound eyes possess adjacent photoreceptors with orthogonally oriented microvilli, establishing the potential for discriminating the orientation of the plane in which the e-vector of light oscillates (Waterman, 1981).

The red swamp crayfish, Procambarus clarkii , appears to have all the anatomical and neural structures required to distinguish differences in the dominant e-vector of polarized light.

The responses of both the long wavelength-sensitive and short wavelength-sensitive retinular cells are affected by the polarization of the incident light (Waterman and Hernandez, 1970;

Muller, 1973; Waterman, 1984). In addition, the lamina and medulla externa of the visual system demonstrate polarization sensitivity in four of the neuronal classes that make up the earliest stages of the visual pathway (Glantz, 1996a; Glantz, 1996b). Inputs from photoreceptors and visual interneurons of P. clarkii converge at the medulla externa into two pathways with

orthogonal e-vector sensitivity: the sustaining and the dimming fibers (Figure 2) (Glantz and

McIsaac, 1998). Vertically and horizontally polarized light activate the sustaining and dimming

fibers respectively (Glantz, 1996b). In addition, activation of one pathway inhibits its

complement. Antagonistic inputs from these orthogonally oriented polarization analyzers project

to both the brain and the medulla terminalis, where polarization opponency between

6 photoreceptors may be integrated (Glantz, 2001). Structural investigations of connectivity patterns between receptors and interneurons have indicated that opponency between these cell groups may provide a mechanism for analyzing and enhancing the sensitivity to the polarization of the signal (Glantz, 1996b). In most of the cells of this pathway, the polarization response is substantially higher for changing e-vector orientation than for fixed orientation, suggesting that polarization discrimination may enhance sensitivity to moving stimuli, a function that is particularly well-developed in crayfish (Glantz, 2001).

Figure 2. A simplified schematic diagram of connections in the crayfish visual system related to polarization sensitivity. Stars represent retinular cells sensitive to horizontal and vertical e-vector. Diamonds represent lamina monopolar cells, which innervate the medulla externa (octagons). Excitatory and inhibitory synapses connect the medulla externa, which projects to the brain through the sustaining and dimming fibers. Adapted from Glantz and McIsaac, 1998).

While the existence of polarization sensitivity in crayfish has been shown in neurophysiological and structural studies, specific functional hypotheses for polarization sensitivity have not been tested. In this study we test whether the neural evidence for polarization sensitivity in P. clarkii translates into a behavioral response. In particular, we test

7 whether polarization information in an otherwise cryptic moving stimulus increases the ability of

P. clarkii to detect it.

Methods

Male and female red swamp crayfish, Procambarus clarkii , were bought from a biological supply company (Carolina Biological Inc., NC, USA). All crayfish studied were of medium size (6-9 cm in length; measured from rostrum to telson). They were maintained in two aerated and filtered 120 liter aquaria on a 12:12 h light:dark cycle at 20 ° C, and fed peas weekly in separate 40 liter aquaria.

Figure 3. Schematic diagram of the experimental apparatus used in the study. The PVC half-section is transparent in this figure to show the inside of the device. In the actual setup the polarizer/diffuser sheets were sandwiched together against the side of the tank.

The experimental apparatus consisted of a half-section of white plastic pipe (10 cm diameter, 48 cm length) that was partitioned into two chambers separated by a sheet of acrylic

8 (Fig. 3). The 13 cm long rear section, where each crayfish was isolated to prevent mechanical stimulation, had a second wall of opaque white plastic. A longitudinal cut (33 cm length) was made down the top of the front section of the pipe, along which a transparent target could be advanced toward the crayfish chamber. The target was attached via a short rod and monofilament to a 300 g mass. When dropped, the mass pulled the target down the length of pipe at an average velocity of 0.3 m/s over 0.65 s. The cut ended 2 cm before the crayfish chamber, thus stopping the target before it struck the chamber wall.

The transparent target was made of a sandwich of 0.3 cm thick clear acrylic and an optically anisotropic, colorless sheet of mylar (Dura-lar Clear Overlay Film, Medium Weight

0.003, Graffix, Cleveland, OH, USA). The mylar was aligned so that it converted the linear polarized light to nearly circularly polarized light. Thus, while inconspicuous under transmitted unpolarized light, the target's optical properties rendered it highly visible when viewed under transmitted polarized light by a viewer with polarization sensitivity (Fig. 4).

Figure 4. The target was constructed of clear acrylic covered with colorless, polarization-active mylar. (a) Unmodified photograph of the transparent target viewed under polarized light. (b) Polarization contrast image of the target (trans-illuminated by vertically polarized light) generated by taking two photographs through a polarizing − Iv I h filter that was rotated by 90° between exposures. Each pixel brightness is equal to 255 ⋅ , where I v and I h are + Iv I h the pixel values when the transilluminated target is viewed through a vertical and horizontal polarizer respectively. (c) Same as in (b) but with wax-paper diffuser depolarizing the light. The bar on right shows the contrast scale.

9 The testing apparatus rested on a sheet of acrylic within a 40-liter aquarium filled with water. A 75-watt flood lamp was mounted 10 cm from the front of the tank. The light passed through a sandwich of a linear polarizer, (HN38S, Polaroid Co., Waltham, MA, USA) and a wax-paper diffuser/depolarizer. For the polarized light trials, the polarizer followed the depolarizer. For the unpolarized light trials, the polarizer preceded the depolarizer. Thus, the two lighting conditions differed only in polarization and not in intensity or spectral distribution

(Fig. 5).

Figure 5. Spectral irradiance inside the crayfish chamber under both polarization conditions. The irradiance probe faced the light source.

Experiments were conducted in June and July of 2005, between 10:00 and 16:00 DST.

Before each day's trials, the experimental tank was rinsed and filled to approximately 20 cm depth with deionized water. Each trial then proceeded as follows. A crayfish was placed in the experimental chamber and allowed a 5-minute acclimation period. If, after 5 minutes, the crayfish did not face toward the light source, an extra period of up to 5 minutes was allowed for the crayfish to orient correctly. After 10 minutes, whether the crayfish was oriented towards the light or not (8 out of 80 did not), the target was advanced by dropping the mass from the top of

10 the aquarium. The animals were tested under two lighting conditions: (1) linearly polarized light, and (2) umpolarized light. After the initial treatment, the light environment was switched from either polarized to un-polarized or un-polarized to polarized, and the crayfish was again given at least a 5 minute acclimation period before the presentation of the visual stimulus. Once again, if the crayfish did not orient towards the light source, the stimulus was advanced after 10 minutes had elapsed. Each crayfish was tested twice, once under each lighting condition, within a period of 20 minutes. The initial lighting condition (polarized or unpolarized) was alternated between crayfish, and no differences were observed between the first and second trials ( (1, N = 80) =

.453, p ≤ 1).

The responses of the crayfish to the advancing target were recorded through the glass bottom of the aquarium using a digital video camera. The presence or absence of a response was recorded by a blind (ignorant of light condition) examination of the video recordings of each experimental trial. Based on preliminary observations prior to testing, a positive response was defined as a retreat by the crayfish of >2 cm. No animal retreated before the target moved.

Results

There was a highly significant difference in response between polarized and un-polarized treatments ( (1, N = 80) = 31.746, p < .001; Figure X). Out of 40 trials that took place under

diffuse, un-polarized light, 10 of the subjects reacted to the stimulus with a positive response

(retreat of >2 cm). Under polarized conditions, 35 out of 40 crayfish responded to the same

stimulus. Tail-flip escapes were observed only under polarized conditions, in 6 out of the 40

trials.

11

Figure 6 . Response of Procambarus clarkii to the approach of the target under polarized and unpolarized conditions (N = 40 for each treatment). Crayfish retreated from the target significantly more often in the polarized condition (p < 0.001).

In order to avoid pseudo-replication problems, the data from each initial trial was analyzed independently of the second trial, so that each crayfish was being exposed to the stimulus for the first time. Under partially linearly polarized light, P. clarkii was four times more

likely to retreat from an advancing, transparent, polarization-active object than under unpolarized

conditions ( χ 2 (1, N = 40) = 16.9, p < 0.0001; Figure 7). Out of 20 trials that took place under un-polarized light, 4 of the subjects retreated from the stimulus. Under polarized light, 17 out of

20 crayfish retreated. The retreat response generally consisted of the crayfish pushing off with the walking legs and chelipeds, and occasionally of a tail-flip escape response. Tail-flip escapes were observed only under polarized light conditions, in 4 out of the 20 trials.

12

Figure 7 . Response of Procambarus clarkii to the approach of the target under polarized and unpolarized conditions (N = 20 for each treatment). Crayfish retreated from the target significantly more often in the polarized condition (p < 0.0001).

Discussion

The highly significant difference in response between treatments suggests that crayfish were more threatened by the advancing target under polarized light conditions. These results provide the first behavioral demonstration of polarization sensitivity in crayfish and one of the few studies that show contrast enhancement. They also demonstrate one effect of sensitivity to polarized light, namely enhanced detection of a moving polarization-active object against a linearly polarized background.

Whether the increased number of retreats is due to enhanced detection of the transparent target itself, or enhanced detection of its motion (or both), is uncertain. P. clarkii are sensitive to moving visual stimuli over a velocity range of at least four orders of magnitude (Glantz, 2001), indicating that motion detection is a critical component of the crayfish visual system. A significant portion of the primary visual synapse (Glantz and Bartels, 1994) and the ascending

13 optic tract (Wiersma and Yamaguchi, 1966) are devoted to motion sensitivity. Glantz (2001) proposed that the polarization sensitive-neurons might also contribute to this pathway. This hypothesis is supported by the existence of tangential cells in the crayfish optic lobe that demonstrate multidimensional selectivity to contrast, motion, and e-vector, and the enhanced responses to changes in e-vector orientation versus responses to fixed orientations (Glantz,

1996b). The results of this study are consistent with, but do not confirm, the hypothesis that crayfish use such a polarization-opponency mechanism to enhance motion detection.

Ecological function of polarization sensitivity

The ecological function of polarization sensitivity in this species is unclear. P. clarkii is an abundant opportunistic omnivore found in seasonally flooded wetlands, lakes, and streams throughout North America, often concealing itself in burrows or under rocks and logs. It is threatened by a variety of predators including aquatic invertebrates, amphibians, reptiles, birds, mammals, fish, and con-specifics (Gherardi, 2002). At shallow depths in streams or lakes, the overhead polarization pattern is essentially a distortion of the polarization pattern of skylight

(Horváth and Varjú, 2004). At angles more than 48 ° from the vertical, outside of Snell’s

Window, this pattern is replaced by a complex but locally uniform pattern due to scattering of downwelling light that is dependent on the time of day and angle of the viewer (Figure 1;

Waterman, 2005). With increasing turbidity, the polarization vanishes. Levels of polarization in the upper photic zone are highest during crepuscular periods (Novales Flamarique and

Hawryshyn, 1997) when crayfish are most active (Gherardi, 2002), indicating that crayfish forage when they are most visually prepared to detect prey and avoid potential predators.

14 Transparent prey (e.g. zooplankton, macroinvertebrates) viewed against polarized backgrounds may be detectable because the scattering of light within their tissues can disrupt the background polarization pattern (Shashar et al., 1998). Transparent zooplankton may partially depolarize the background polarization pattern, shift the dominant e-vector by phase retardance, or if the object is birefringent, partially polarize background illumination (Sabbah and Shashar,

2006). The experimental design in this study was intended to mimic the approach of such an object, in this case a transparent object that exhibits phase retardance of transmitted light.

Fish, the chief predators of crayfish (Nyström, 2002), may be detected via the polarized reflections from their mirror-like scales (Shashar et al., 2000). Analysis of reflected polarized light may also allow crayfish to detect and recognize substrates, plants, and other underwater features (Novales Flamarique and Hawryshyn, 1997). It is also possible that polarization sensitivity could be used to increase the contrast of images within the stationary visual field of the crayfish by using a simple algorithm that removes the effects of haze from acquired images

(Schechner, 2004).

A polarization opponency mechanism

Based on the structural and neural properties of the crayfish visual system, Glantz (2001) proposed a polarization opponency mechanism for contrast enhancement of otherwise imperceptible moving objects. Consider a crayfish at the bottom of a murky lake in the early evening, when the predominate polarization signal is oriented perpendicular to the 48 degree incident refracted sunlight (Figure 1). If a polarization-altering transparent object, such as a small macroinvertebrate, were to cross the field of vision of the crayfish, then two retinal sections of the crayfish eye would transiently detect changes in e-vector distribution. The

15 portion of the eye initially imaging the transparent object will then be exposed to the background polarization signal, while a new section of the visual field will receive the polarization signal from the moving object. Each of these visual field patches will experience a temporal contrast of e-vector from the varying polarization signals. The information from individual retinal patches will then be encoded in the parallel fibers of the polarization opponency system , as the signal activates one of the polarization channels (e.g. sustaining fibers) and inhibit its complement

(dimming fibers; Figure 2). The mechanism is independent of specific e-vector angle and requires only that the e-vector orientation change over time, allowing detection of a moving object in the absence of an intensity contrast. This ability is analogous to the perception of movement using spectral differences in situations where intensity is constant (Bernard and

Wehner, 1977).

In addition to using differences in e-vector, the crayfish visual system may also exploit percent polarization levels for contrast enhancement. Just as temporal variation of dominant e- vector may produce unique signals in the crayfish optic tract, time-varying levels of percent polarization could potentially activate ascending optic fibers with different signal strengths. It is likely that animals with polarization sensitivity use both e-vector and percent polarization in parallel for polarization-based recognition (shabbah and shashar, 2006). Again, this ability is analogous to the interaction between hue and saturation in color vision.

Does polarization sensitivity provide contrast enhancement?

The results of this experiment provide evidence that crayfish may use polarization sensitivity to enhance contrast of transparent moving objects under certain conditions. However, the high levels of percent polarization, ideal polarization-activity of the target, and clear, well-

16 illuminated water are rarely found in nature. If the percent polarization and illumination were less than ideal, it is uncertain whether polarization sensitivity would provide any significant enhancement to detection of a moving object.

Furthermore, the strong illumination in this experiment may have caused light adaptation within the eye, which would extend and isolate each ommatidium from its neighbor, creating a mosaic image that is highly sensitive to movement. Any slight shift in the visual field would then stimulate neighboring ommatidium. At lower light levels that the animal might normally encounter, the distal pigments move out of the crystalline cone areas and light from multiple ommatidia may converge on a single rhabdom to form a superposition image. Enhanced sensitivity of a light-adapted eye could exaggerate the response of crayfish to an advancing stimulus. The uniform intensity of spectral distribution between treatments (Fig. 3) ensures that this phenomenon cannot account for behavioral differences between the polarized and unpolarized conditions, however it may have resulted in an exaggerated response compared to that experienced by crayfish during the crepuscular period.

Understanding the function of polarization sensitivity in P. clarkii is complicated by the

fact that the polarization signal may interact directly with color vision. An ommatidium of P.

clarkii has seven photoreceptors containing a visual pigment that peaks at 530 or 567 nm

(depending on chromophore) (Crandall and Cronin, 1997). With the exception of those in the

dorsal retina, the ommatidia also contain an eighth cell expressing a visual pigment that peaks at

440 nm (Waterman and Fernandez, 1970). Although sample sizes are low, it appears that, in the

anterior ommatidia, most of the 440 nm receptors are sensitive to horizontal polarization, and

most of the 530/567 nm receptors are sensitive to vertical polarization. In the dorsal retina,

which contains only the 530/567 nm receptors, most of the cells are sensitive to vertical

17 polarization. These results suggest both regional and chromatic specialization of polarization sensitivity and complicate any attempt to understand its function.

Based on the high polarization sensitivity of optic neurons, in conjunction with the robust behavioral responses in this experiment, it is likely that crayfish use polarization sensitivity to enhance contrast in environments where intensity differences are minimal. This function may provide a selective advantage for crayfish in detecting prey and avoiding predators. What remains to be determined is the importance of the polarization signal relative to normal contrast vision, considering the overlap between polarization sensitivity, contrast, and color vision. Also, it will be important to understand the limits of the crayfish polarization system in order to understand under what optical conditions polarization sensitivity will provide a significant advantage in nature.

Escape responses as indicators of sensory abilities

One of the difficulties in research on polarization sensitivity is that the behavioral responses have generally been marginal (reviewed by Waterman, 1981). Thus, many significant results have been based on small differences and large samples sizes. This makes experiments difficult to replicate and generates concerns about potential biasing artifacts. In contrast, this study achieved a p-value less than 0.00005 with a sample size of only 40 animals. Given the same ratios of retreat to non-retreat, a p<0.05 result could have been obtained with approximately 12 individuals. We attribute the strength of these results to the stereotypical and critical nature of escape responses. While feeding, orientation, and social behaviors are important to the survival and fitness of an organism, they tend to be less reliably induced than escape responses, particularly after a period of captivity. For this reason, we believe that sensory

18 assays based on escape responses may be quite useful for other species and other sensory abilities.

Conclusion

This experiment demonstrates one effect of sensitivity to polarized light, namely enhanced motion detection of a polarization-active object against a linearly polarized background. True polarization vision is more complex, and must include the ability to discriminate between polarized and depolarized light, recognize particular angles of polarization, and integrate this information to identify differences between two stimuli based on polarization alone. The interesting problem now is to devise complex behavioral experiments that demonstrate the extent of polarization sensitivity, and attempt to understand how this capacity is neurally encoded within the crayfish visual system. Investigations into the relationship between color, polarization, and contrast vision will also provide insight into the visual perception of the crayfish.

Acknowledgments

We thank Dr. Rachel Merz for many critical readings of this paper. The Meinkoth Fund at Swarthmore College provided funding for JT. SJ was supported in part by a grant from the

National Science Foundation (IOB-0444674).

19 Literature cited

Bernard, G. D. and Wehner, R. (1977). Functional similarities between polarization vision and color vision. Vision Res 17 , 1019-28.

Flamarique, I. N., Hawryshyn, C. W. and Harosi, F. I. (1998). Double-cone internal reflection as a basis for polarization detection in fish. Journal of the Optical Society of America a-Optics Image Science and Vision 15 , 349-358.

Frisch, K. v. (1967). The dance language and orientation of bees. Cambridge, Mass.: Belknap Press of Harvard University Press.

Gherardi, F. (2002). Behaviour. In Biology of freshwater crayfish , (ed. D. M. Holdich), pp. 258-90. Oxford, [England] Ames, Iowa: Blackwell Science; Distributor USA and Canada, Iowa State University Press.

Glantz, R. M. (1996a). Polarization sensitivity in crayfish lamina monopolar neurons. Journal of Comparative Physiology a-Sensory Neural and Behavioral Physiology 178 , 413-425.

Glantz, R. M. (1996b). Polarization sensitivity in the crayfish optic lobe: Peripheral contributions to opponency and directionally selective motion detection. Journal of Neurophysiology 76 , 3404-3414.

Glantz, R. M. (2001). Polarization analysis in the crayfish visual system. Journal of Experimental Biology 204 , 2383-2390.

Glantz, R. M. and Bartels, A. (1994). The spatiotemporal transfer function of crayfish lamina monopolar neurons. J Neurophysiol 71 , 2168-82.

Glantz, R. M. and McIsaac, A. (1998). Two-channel polarization analyzer in the sustaining fiber dimming fiber ensemble of crayfish visual system. Journal of Neurophysiology 80 , 2571-2583.

20 Horváth, G. and Varjú, D. (2004). Polarized light in animal vision: polarization patterns in nature. Berlin; New York: Springer.

Labhart, T. and Meyer, E. P. (2002). Neural mechanisms in insect navigation: polarization compass and odometer. Curr Opin Neurobiol 12 , 707-14.

Lythgoe, J. N. and Hemmings, C. C. (1967). Polarized light and underwater vision. Nature 213 , 893-4.

Marshall, J., Cronin, T. W., Shashar, N. and Land, M. (1999). Behavioural evidence for polarisation vision in stomatopods reveals a potential channel for communication. Curr Biol 9, 755-8.

Marshall, N. J. (1988). A unique colour and polarization vision system in mantis shrimps. Nature 333 , 557-60.

Novales Flamarique, I. and Hawryshyn, C. W. (1997). Is the use of underwater polarized light by fish restricted to crepuscular time periods? Vision Res 37 , 975-89.

Nyström, P. (2002). Ecology. In Biology of freshwater crayfish , (ed. D. M. Holdich), pp. 192-235. Oxford, [England] Ames, Iowa: Blackwell Science.

Sabbah, S. and Shashar, N. (2006). Polarization contrast of zooplankton: A model for polarization-based sighting distance. Vision Res 46 , 444-56.

Schwind, R. (1999). Daphnia pulex swims towards the most strongly polarized light - a response that leads to 'shore flight'. J Exp Biol 202 Pt 24 , 3631-5.

Shashar, N. and Cronin, T. W. (1996). Polarization contrast vision in Octopus. J Exp Biol 199 , 999-1004.

Shashar, N., Hagan, R., Boal, J. G. and Hanlon, R. T. (2000). Cuttlefish use polarization sensitivity in predation on silvery fish. Vision Res 40 , 71-5.

21 Shashar, N., Hanlon, R. T. and Petz, A. D. (1998). Polarization vision helps detect transparent prey. Nature 393 , 222-223.

Shashar, N., Rutledge, P. and Cronin, T. (1996). Polarization vision in cuttlefish in a concealed communication channel? J Exp Biol 199 , 2077-84.

Shashar, N., Sabbah, S. and Cronin, T. W. (2004). Transmission of linearly polarized light in seawater: implications for polarization signaling. J Exp Biol 207 , 3619-28.

Waterman, T. H. (1981). Polarisation sensitivity. In Handbook of Sensory Physiology , vol. 7/6B (ed. H. Autrum), pp. 281-469. New York: Springer-Verlag.

Waterman, T. H. (2006). Reviving a neglected celestial underwater polarization compass for aquatic animals. Biol Rev Camb Philos Soc 81 , 111-5.

Wiersma, C. A. and Yamaguchi, T. (1966). The neuronal components of the optic nerve of the crayfish as studied by single unit analysis. J Comp Neurol 128 , 333-58.

22