Response of the larvae of the sea urchin Centrostephanus rodgersii under light and polarization conditions

Abstract

In this paper, we investigated the response of the larvae of the sea urchin Centrostephanus rodgersii to unpolarized halogen light, darkness as well as polarized (linear, circular and elliptical) halogen light presented in side view to observation chambers. The larvae exhibited positive phototaxis when exposed to unpolarized, horizontally polarized, elliptically polarized, and circularly polarized light. However, they did not respond to vertically polarized light. In fact, larval swimming behavior after exposure to vertically polarized light was the same as keeping them in the dark. These findings indicate that not only may the larvae of C. rodgersii have photoreceptors capable of detecting light in the visible spectrum but they may also possess the capacity to differentiate between horizontally and vertically polarized light, which might help them navigate. These findings suggest the possibility of aligned photoreceptors in these larvae. In addition, we found that the larvae respond to circularly polarized light. This result is notable as there are few documented cases of circular polarization sensitivity in animals. As they did not appear to have a preference between the left‐ and right‐handed polarizations, one would need to study their photoreceptor cells to determine the mechanism by which they can detect circular polarization. This could pave the way to the development of new polarization detectors based on biological mechanisms.

We found that the larvae of the sea urchin Centrostephanus rodgersii exhibited positive phototaxis and also responded positively to horizontally and circularly polarized light. However, we did not observe any response to vertically polarized light. This gives an indication that not only do these larvae have photoreceptors, but they also possess a mechanism for detecting linear and circular polarization. Furthermore, they must also have the ability to differentiate between different linear polarizations. Further investigation into this mechanism, which could be related to the ordering of their photoreceptors, could allow for a better understanding of polarized vision and behaviors in animals.

Abbreviations

CP

circular polarization

DoP

degree of polarization

DVM

diel vertical migration

FSW

filtered sea water

HP

horizontal polarized

IR

infra‐red

ROI

region of interest

UP

unpolarized

UVA

ultraviolet A

UVB

ultraviolet B

UVC

ultraviolet C

VP

vertically polarized

INTRODUCTION

Sea urchins are ecologically important algal grazers and, in large numbers, can have detrimental effects on kelp forest ecosystems. ¹ , ² Sea urchins have larvae that develop ciliated arms that they use for swimming and feeding. Swimming behavior in response to environmental stimuli may affect larval distribution and dispersal and thus influence the location of adult sea urchin populations. ³ As such, it is important to understand the environmental factors that influence their behavior during the larval phase.

One of such factors is visible light. The larvae of Hemicentrotus pulcherrimus and Temnopleurus reevesii reversed their swimming direction under photoirradiation, ⁴ indicating that light may influence their displacement and that they therefore possess photoreceptors. Indeed, sea urchin larvae appear to have photoreceptor‐expressing cells as part of their digestive system. ⁵ Furthermore, opsin2 (a protein found in photoreceptors) was present in their arms, ⁶ which is an indication of the presence of photoreceptors. The arm skeletal rods of sea urchin larvae are composed of birefringent calcite. ⁷ As it is hypothesized that calcite may act as a polarizing filter that can give rise to polarization sensitivity in other echinoderms such as in brittlestars, ⁸ it seems natural to expect that sea urchin larvae may also be sensitive to polarization, in addition to being sensitive to light.

Animals use polarization for a range of purposes. One of such is navigation. For example, the African dung beetle Scarabaeus zambesianus uses the polarization of moonlight to roll a ball of dung in a straight line radially away from the sources is to escape aggressive competition for food foraging inside the dung ball. ⁹ , ¹⁰ Other insect species, such as migrating monarch butterflies, ¹¹ , ¹² crickets, ¹³ , ¹⁴ and desert locusts, ¹⁵ use a skylight compass ¹² , ¹⁵ , ¹⁶ , ¹⁷ to navigate under a variety of atmospheric conditions. Polarization may also be used for communication purposes (such as in the case of stromatopods who may be using circularly polarized light to potentially seek out partners) ¹⁸ , ¹⁹ , ²⁰ or to enhance contrast and visibility through the polarization angle of light reflected from shiny surfaces ²¹ or from water surfaces. ²² Underwater polarization vision is also quite widespread in nature (see, e.g., cephalopods, ²³ anchovies, ²⁴ damselfishes, ²⁵ rainbow trout ²⁶ ) and is often used for contrast enhancement or predator/prey responses.

Polarization arises when sunlight passes through the atmosphere and scatters off air molecules. This phenomenon is called Rayleigh scattering. The intensity of the scattered light is proportional to I ∝ λ ⁻⁴ (1 + cos θ ²), with λ the wavelength of the incident light and θ the angle between the incident and scattered light.^* The intensity of the scattered light is the strongest in the direction (and opposite direction) of the incident light, and the degree of horizontality and polarization can be predicted. The same is true as the light passes through the water and interacts with water molecules. In both cases, the polarization pattern can be affected by Mie scattering.^† Mie scattering occurs in the presence of cloudy atmospheric conditions or turbid waters. ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ However, Cheng et al. ³² showed that, even under these conditions, the polarization pattern can still be recognizable enough to be used as a navigational tool. Given the above and since the Mie scattering coefficient from ions found in pure sea water is extremely small (≪m⁻¹), Rayleigh scattering is more likely to influence sea larvae urchins than Mie scattering, if at all.

Just below the surface of the water, light is usually found to be horizontally polarized, with a relatively high degree of polarization. ³³ Scattering events underwater also tend to create more horizontally polarized light. ³⁴ Internal reflections are also more horizontally polarized. ³⁵ Although the prominence of horizontal and vertical components is expected to change with the depth of the water column due to changes in density and in turbid conditions, previous studies have shown that polarization can be used by marine animals at depths up to 15 m ³⁶ and could exist up to 200 m. ³⁷ , ³⁸ Since sea urchin larvae reside in the plankton, they may be particularly sensitive to polarized light and, one may be tempted to speculate that horizontally polarized light may generate a stronger behavioral response in sea urchin larvae when compared to vertically polarized light exposure.

We note, in addition, that some species may be sensitive to circular polarization (CP). For example, the aforementioned mantis shrimp and the larvae of fireflies Photuris lucicrescens and Photuris versicolor emit weak left and right circularly polarized bioluminescent light ³⁹ while the light passing through the body of certain dinoflagellates is also circularly polarized. ⁴⁰ One may thus ask whether sea urchin larvae are also sensitive (or respond to) circularly polarized light.

We addressed the question as to whether the larvae of the sea urchin C. rodgersii change the course of their swimming direction under both unpolarized and polarized light. A positive response, exhibited by swimming toward the light source (positive phototaxis), would indicate their sensitivity to these environmental factors and could bring a wealth of information about the sea urchin larvae dispersal. This might be useful for the detection of potential photoreceptors and the design of new bio‐detectors.

MATERIALS AND METHODS

Methods

Our goal is to determine whether sea urchin larvae exhibit phototaxis. While it is known that, in their natural habitat, sea urchins exhibit Diel Vertical Migration (DVM), ⁴¹ , ⁴² one still needs to demonstrate that such a motion is the result of their response to the sunlight rather than other factors (e.g., temperature, pressure). For this reason, we do not test for vertical motion with our setup, that is, we do not illuminate the sea urchin Centrostephanus rodgersii from above. Instead, we study whether they exhibit phototaxis behavior when light is directed from the side of the water chamber. The other reason to not shine the light from above is that this helps to test their sensitivity to the different polarization modes. Indeed, when the light is shone from above, the larvae, which are free to rotate but not flip in the water column, can always orient themselves to perceive both linear polarization modes in the same way. Sea urchin larvae in general, however, do not rotate vertically, that is, “flip” as they prefer to remain upright. ⁴³ Thus, when the light is shone from the side, this orientation symmetry is broken, allowing us to test whether they exhibit a preference for a particular polarization mode as they now perceive the two polarization modes as different (see Figure 1).

FIGURE 1.

In images (A, B) larvae are subjected to different polarizations of halogen light incident from above. As the larvae are free to rotate along the horizontal plane, they can perceive both polarization modes as the same. In images (C, D), larvae are subjected to different polarization modes of halogen light incident from the side. The sea urchins are not free to rotate along the vertical plane (i.e., flip), thus they cannot perceive the two polarizations as the same.

We deliberately introduced a lateral light‐field so that the photic gradient lay orthogonal to gravity, an approach previously used to reveal subtle horizontal orientation cues in planktonic larvae. Although downwelling light predominates in nature, side‐lighting also occurs in turbid coastal waters where multiple scattering broadens the angular distribution of photons. Hence, the behavioral trends we report are likely to represent one ecologically relevant subset of the larval visual repertoire rather than an artifact.

We performed two sets of experiments: in the first set of experiments, the larvae were exposed to halogen light to mimic sunlight, whereas in the second set of experiments, the larvae were kept in the dark. Their swimming response was recorded using a 12‐megapixel Iphone 13 pro camera using the wide f/1.5 aperture. We note that the larvae were never directly exposed to sunlight during the experiments as there are various environmental factors associated with sunlight (such as flux, temperature, and angle of incidence) that could not be controlled to take meaningful measurements.

The spectrum of our halogen lamp ranges from 400 to 1000 nm with a wide peak from 480 to 750 nm, meaning that it contains visible light and some infrared components, while the sunlight spectrum ranges from 250 to 2500 nm. Hence, both the halogen and sunlight spectra are similar in terms of shape and location of the peak in the visible spectrum. Even though the halogen lamp does not include as much as infrared light as the solar radiation, we do not expect this to be a problem as sea urchins are not expected to be sensitive to those wavelengths. ⁴³ Also, while ultraviolet light exists in the sunlight spectrum and is mostly absent in the halogen spectrum, it only makes up about 9% of the total energy from the sun. Hence, its absence from the halogen lamp spectrum should not impact the larval swimming pattern. We also note that IR, UVB, and UVC lights are absorbed at a greater rate underwater than the visible spectrum, so their influence on larval behavior should be negligible. There would normally be UVA light present in the natural habitat of sea urchins. However, the proportion of UVA light compared to visible light is very low. ⁴⁴ We can therefore safely use the halogen light to simulate the light that sea urchin larvae would normally be exposed to in their natural habitat.

A subsequent set of experiments was undertaken to study the sensitivity of C. rodgersii to polarized light. The following experiments were done with polarized sunlight in mind as that is the type of light to which sea urchins would be most exposed, but it is worth noting that moonlight creates the same exact polarization pattern as sunlight, albeit being up to 1 million times dimmer. ³⁴ , ⁴⁵ In these new sets of experiments, the sea urchin larvae were in turn exposed to vertically and horizontally polarized halogen light as well as elliptically and circularly polarized halogen light. Their swimming response was recorded as described above.

Species

Centrostephanus rodgersii (test diameter 6–12 cm) were collected (1–2 m depth) from Camp Cove, Sydney in July 2022 when the sea temperature was 17°C. They were maintained in aquaria at 17–18°C. The urchins were used for spawning within 4 days of collection. Spawning was induced by injection of 1 mL of 0.5 M KCl through the peristome. Three fertilizations were done each obtaining gametes from two to three males and females, which were used to generate embryo populations. Sperm was collected dry directly from the gonopores using a glass pipette, pooled in equal portions from each male and stored at 4°C until use (<30 min). Egg quality was checked microscopically to select the best egg sources. Sperm from individuals was checked and quality assessed for motility. Eggs from selected females were pooled in 500–1000 mL beakers of filtered (1 μm) seawater (FSW; Millipore) at 18°C. The eggs were fertilized (10⁴ sperm mL⁻¹) and microscopic examination after 10 min revealed >90% fertilization as indicated by the presence of the fertilization envelope. The eggs were then rinsed to remove excess sperm and the FSW renewed. The density of the embryos was determined from an average of five counts of 10 μL aliquots using a Sedgwick Rafter Counting Chamber and placed in 5 L beakers at 10–20 embryos mL⁻¹. After hatching, the larvae were reared in 5 L beakers at a density of 10 mL⁻¹. The cultures were gently bubbled to keep the larvae in suspension and the water was changed by reverse aspiration every 2–3 days. After the water changed, the larvae were fed Proteomonas sulcata (20,000 cells mL⁻¹).

A sample of 250 mL of the larval culture (4–20 larvae mL⁻¹) was taken from the culture and 250 mL of 1 μL FSW was added. The larvae used for testing in the different lighting regimes were between the ages of 3 and 6 weeks after spawning (see Figure 2). The larvae were then placed in the observation chambers. These chambers were made by using a high purity borosilicate glass to reduce light scattering as well increased durability. The chambers were shaped as a square prism with dimensions of 5 cm × 5 cm × 30 cm (see Figure 3).

FIGURE 2.

Image of a larvae of the sea urchin species Centrostephanus rodgersii. Note their ciliated appendages (arms) which are used for feeding and swimming.

FIGURE 3.

Schematic of the setup. In this configuration, the halogen light source was level with the midpoint of the water chamber and a linear polarizing filter was placed in front of the halogen lamp. This same setup was also used for left and right circular polarizations, using circular polarizers. For elliptical polarization, the halogen lamp was moved further away and vertically upwards so that it made a ≈45° to the horizontal plane and the linear polarizer was placed in front of the lamp.

Setup

The chambers with the larvae were placed on a table in a dark room and in front of a halogen light source whose distance to the chamber could be adjusted. The larvae were exposed to different light conditions from various directions and their overall swimming behavior within the water chamber in the camera's FOV was recorded at 0, 5, 10, 20, 30, 45, and 60 min of light exposure. The lighting conditions included:

• Unpolarized light. The halogen lamp was aligned with the center of the water column, at an arbitrary distance of 85 cm. The lamp was then moved horizontally so that the light could shine incident either from the left or right sides of the water chamber in separate runs. This allowed us to test if the response was dependent on the physical direction of the incident light.^‡ The flux (which was measured with a light meter) was approximately 1750 lux at the center of the water column.

• Darkness. The halogen light source was switched off and the larvae were free to move around without any external influences. At each time interval, the light was quickly switched on and a photo taken before switching the light off again.

• Horizontally and vertically polarized light. The halo‐ gen lamp was again aligned with the center of the water column at a distance of 85 cm. A Hoya 77 mm linear polarizing filter was placed in front of the halogen lamp in the vertical and horizontal orientations to reproduce the two polarization modes. The larval swimming response recorded accordingly. The flux of photons at the center of the water column and with the linear polarizing filter in place was approximately 400 lux. Absolute irradiance differed between polarization states because neutral‐density compensation would have reduced the degree of polarization (DoP). We therefore prioritized maximizing the degree of polarization over matching absolute illuminance, accepting some variation in lux to ensure the highest polarization quality. Critically, the intensity span among treatments (290–1750 lx) is well within the dynamic range over which C. rodgersii larvae show linear phototactic scaling, making it unlikely that luminance alone explains the differential responses. We note that Hoya claims that their polarizing filters have a 40%–29% transmission rate, whereas we measured a transmission rate of 23%. However, the difference may be accounted for via the reduced reflections of the walls of the experimental room with the polarizing filter in place. The filter is rated to have a polarizing efficiency of >99.9%. While there may be some scattering off various objects and off the walls of the room that may change the polarization of the incident light onto the larvae, the flux would be much smaller comparatively to the flux of the direct incident light with the desired polarization. Thus the effects of random scattering events should be negligible and we can expect a high degree of polarization.

• Elliptical polarized light. The halogen lamp was moved to a point 125 cm away from center of the water column to create a ≈45° above the horizontal. This results in an uneven mix of horizontal and vertical polarizations of light, which is mathematically equivalent to elliptically polarized light with a 0‐phase difference between the components. The polarizing filter was once again used in the “horizontal” and “vertical” orientations. The photon flux in this configuration was approximately 290 lux since the source is further away.

• Circularly polarized light. To test the effects of circular polarization, a circular polarizer (CP42HE and CP42HER from Edmund Optics with diameter of 50 mm) was placed in front of the halogen lamp, which was aligned with the center of the water column at a distance of 85 cm. The efficiency of these polarizers is supposed to be >99.9% so there should not be a large component of linearly polarized or unpolarized light going through. The effects of both left‐handed and right‐handed polarizations on the larval swimming responses were tested. The flux of photons at the center of the column in this configuration was approximately 600 lux. While random scattering events may effect the polarization of the incident light, we expect it to be negligible. Edmund optics claim a transmission rate of 42% for wavelengths of 400–700 nm, whereas we measured a transmission rate of 34% but this could also be due to the reduced reflections from the walls of the room.

The experiment with each light regime was repeated three times. We observed that the response of the larvae was evident after 10 min of exposure and less apparent at 60 min as the larvae tend to sink to the bottom after long exposure times. Hence, further images were taken at the 10 min mark only.

RESULTS

Polarization of the halogen light

First of all we tested for the potential depolarization of light going through the water chamber. We found >97% of linearly polarized light and >98% of circularly polarized light retained their polarization. This means that the light that the sea urchins are experiencing are still polarized and that and depolarization effects due to scattering between the water, glass or particles in the water can be effectively ignored. See Appendix A for more details.

Image analysis

The larvae and their motion were detected using the software ImageJ (version 1.54i). ⁴⁶ To minimize the counting of objects that were not larvae, we assumed a size between 20 and 300 pixels that covers most larval sizes while excluding most contaminants. While each larva is meant to be counted individually by the software, the glare stemming from the light reflection off the glass made this difficult. Indeed, ImageJ counts all visible objects in the water tank with a size and luminosity consistent with that of the larvae. As the glare saturates the luminosity in that part of the tank, the identification of the larvae becomes less accurate.

To make the larvae more visible, the image contrast was increased (see Figure 4). However, since this also made the glare more apparent, the ROI had to be cut again thus leading to a smaller ROI. The total number of counts found by ImageJ is therefore a compromise between larvae visibility and the size of the ROI needed to get a decent number of counts. The percentage distributions of larvae were calculated for each light condition after the number of larvae on both sides of the ROI was returned by ImageJ.

FIGURE 4.

In this example, the larvae (see arrows) were exposed to horizontally polarized light from the left and we use ImageJ for the counts after increasing the contrast.

To minimize the number of false counts, we explicitly cut the part of the image where the glare was (see Figure 5) and defined a region of interest (RoI) accordingly. As the glare was present at the edges of the water chamber (due to the reflections occurring at the edge), this cut does not overly impact the results as fewer larvae are present on the glass walls. The cut was performed manually for each image since the extent of the glare varied from image to image. As a result, the RoI's sizes and positions vary across the set of images. We mitigate this effect in our data analysis by considering two approaches: one was based on the number of counts and the other one based on the count density of larvae per ROI.

FIGURE 5.

Example of a region of interest for larvae analysis while minimizing the effects of unwanted counts. By imposing specific objects size and luminosity restrictions, we could eliminate unwanted counts. Left: Area that was used for the larvae distribution analysis. Middle: Outlines of the selected larvae (indicated by dots) in the ROI. Right: Remaining larvae after cut and ROI selection to avoid glare which were used for the analysis.

In the first approach (where we used the total number of counts), each image's ROI has a different volume and is cut into two equal parts from the center of the ROI to define a number of counts in the left and right sides. In the second approach (where we used densities), the center of the water column was kept as a reference point for defining the left and right sides, meaning that the volumes enclosed in each side of the ROI were no longer equal. As volume asymmetry could affect our conclusions (especially when it comes to performing statistical tests of the impact of polarization), we used the density of counts in each side, instead of the number of counts, to perform our analysis. The results displayed in Tables 1 and 2 are only for the first approach but they hold true for the second as well.

TABLE 1.

Proportion of larvae for each light setup in the half of the ROI closer to the light source compared to the half further away for the horizontally, vertically polarized, unpolarized (halogen) light and darkness. There were $\mathcal{n}$ = 8 repeats for horizontally, vertically polarized, and unpolarized light and $\mathcal{n}$ = 4 for darkness.

Light regime	Mean	Standard deviation	n
Horizontal polarization	0.72373	0.05228	8
Vertical polarization	0.62502	0.06294	8
Halogen light	0.73811	0.06472	8
Dark	0.51330	0.08842	4

TABLE 2.

Paired comparison of the different light regimes after performing a post‐hoc Tukey's multiple comparison test.

Comparison	p Value	Statistical difference
Horizontal vs. vertical	0.0216	Yes
Horizontal vs. halogen	0.9947	No
Horizontal vs. dark	0.0366	Yes
Vertical vs. halogen	0.0182	Yes
Vertical vs. dark	0.4969	No
Halogen vs. dark	0.0234	Yes

The data sets corresponding to the incident light from the left and right sides for the different lighting conditions were analyzed separately. Not only did this allow us to assess whether the larvae responded to the light but also to determine whether they swam in a preferred direction, independent of the location of the light source. To perform the statistical analysis, an arcsine transformation was first applied to the data to stabilize variances ⁴⁷ and then a one‐way analysis of variance (ANOVA) test was performed, after verifying the assumptions of normality (Shapiro–Wilk test) and homogeneity of variances (Bartlett's test). The ANOVA test comparing unpolarized light, horizontally and vertically polarized light with light incident from the left and from the right gave F(5, 18) = 3.098, p = 0.0343 showing a statistically significant difference between the means of some of the data sets, see Table 3. The post‐hoc Tukey's multiple comparison tests of all the different setups were then performed to determine if any two lighting conditions were statistically similar and whether the larvae had a preferred swimming response to the different light conditions. We compared the larvae swimming response to the light source incident from the left and right sides of the water chamber to determine whether the source position had an influence on their response.

TABLE 3.

ANOVA test for comparison of sea urchin distribution for horizontally, vertically polarized, and unpolarized light for the light incident from the left separately from the light incident from the right. DFn is the number of data sets minus one and DFd is the number of data points minus the number of data sets.

Source of variations	Sum of squares	DOF	Mean squares	F(DFn, DFd)	p Value
Between data sets	0.07564	5	0.01513	F(5, 18) = 3.098	0.0343
Within data sets	0.08789	18	0.004883
Total	0.1635	23

Any differences in swimming response to exposure to horizontal versus vertical polarization and unpolarized light versus darkness were also tested. It was found that, under the same lighting regimes, the larvae distribution was always consistent relative to the location of the incident light (p‐values of 0.9614, 0.9988, and >0.9999, respectively for horizontally and vertically polarized light and unpolarized light). Therefore, the two data sets were combined and after combination, we have eight sets of 10 min runs for larvae under exposure of unpolarized light, horizontally and vertically polarized light, and four sets of 10 min runs for larvae in darkness (since in the absence of a light source, there is no left‐side or right‐side positions). Thus, a new ANOVA with Welch's correction was done to account for the number of repetitions (once again, the data sets met the assumptions of normality and homogeneity of variances). The ANOVA test between unpolarized light, horizontally and vertically polarized light, and darkness gave F(3, 24) = 17.05, p < 0.0001, once again showing a difference in the means between data sets (see Table 4).

TABLE 4.

ANOVA test for comparison of sea urchin distribution after combining left and right incident light datasets for each light environment plus darkness.

Source of variations	Sum of squares	DOF	Mean squares	F(DFn, DFd)	p Value
Between data sets	0.2218	3	0.07393	F(3, 24) = 17.05	<10−4
Within data sets	0.1041	24	0.004335
Total	0.3258	27

Response of sea urchin larvae under lighting conditions

Linear and unpolarized light

In Figure 6, we see that after 10 min of unpolarized halogen light, the larval distribution was more skewed toward the light source compared to 10 min of darkness, thus displaying positive phototaxis. It is also evident that the distribution of larvae is more toward the light after 10 min of horizontally polarized halogen light exposure than after 10 min of vertically polarized light exposure.

FIGURE 6.

Swimming responses of larvae after 10 min of exposure to different light regimes. The contrast has been increased in these images to make the larvae (white dots) more visible. In all of them, unpolarized halogen, horizontally polarized and vertically polarized halogen light was shining from the left. We can see that for unpolarized and horizontally polarized light, the larvae swam toward the light source (left) and for vertically polarized light and darkness the larvae were more spread out.

Elliptically polarized light

Figure 7 shows the response of the larvae when the halogen lamp was moved to a higher position allowing for elliptical polarization. We found that under both configurations of the linear polarizing filter, the distribution of larvae after 10 min were skewed toward the light source, thus displaying no visible difference with the horizontally polarized light case, despite a smaller flux. As elliptical polarization features both vertical and horizontal polarization, it is likely that the larvae are actually responding to the horizontal component.

FIGURE 7.

Swimming response of larvae after 10 min of exposure to elliptically polarized light incident from the left. The contrast of the images was adjusted to better visibility of the larvae. There was a positive phototaxis response to both elliptical polarizations.

Circularly polarized light

Figure 8 shows the larvae response to circularly polarized light, both left‐handed and right‐handed. In both instances, the larvae showed positive phototaxis. This sensitivity to circularly polarized light could indicate that they possess optical filters (e.g., quarter‐waveplates) to transform the circularly polarized light into a linearly polarized signal (assuming that photoreceptors can only detect linear polarization). They did not show a preference for left‐ versus right‐handed circular polarizations.

FIGURE 8.

Swimming response of larvae after 10 min of exposure to circularly polarized light incident from the left. The contrast of the images was adjusted to better visibility of the larvae. There was a positive phototaxis response to both left‐handed and right‐handed circularly polarized light.

Data analysis

Table 1 shows the mean number of larvae in the different lighting treatments, and the arcsine transformed data is shown visually in Figure 9. When comparing horizontal polarization versus vertical polarization as well as unpolarized light versus darkness, we found that there was a significant difference in the proportion of larvae in the left and right sides of the chamber (p values of 0.0216 and 0.0234, respectively). There was no difference between unpolarized light and horizontal polarization setups nor darkness versus vertical polarization setups.

FIGURE 9.

Proportions of larvae responding to each lighting regime, comparing the counts of larave closer to the light to the counts of larvae further away (positive phototaxis), after an arcsine transformation. The numbers indicated on each bar are obtained from Table 1 for the horizontally polarized (HP), vertically polarized (VP), unpolarized light (UP), and darkness (Dark).

For the elliptical polarizations, there was no difference in the response of the larvae to the “horizontal” and “vertical” orientations of the polarizing filter. The larvae showed positive phototaxis to both configurations (F(2, 17) = 0.1290 and p = 0.8798, see Appendix B). For the circular polarizations, larvae also showed positive phototaxis for both the left‐ and right‐handed polarizations (F(2, 17) = 0.8498 and p = 0.4449, see Appendix B).

In both Tables 1 and 2, we used larval counting (splitting the ROI into two equal halves) rather than densities. However, the comparison of the results for the different lighting conditions also holds true for the second approach.

DISCUSSION

Limitations

There are a number of caveats to our study. For example, we tried to account for the counting of nonlarval particles by setting criteria on particle sizes. However, this meant that some larvae toward the further side of the water chamber might not have been counted due to their smaller apparent size. In addition, larvae in the background were less lit up so they also might not have been incorporated in the counts. We assumed that any of the larvae that were not counted in the back due to the above reasons were in the same proportion as the larvae that were counted in the front due to the symmetrical nature of the water chamber (i.e., we have assumed that the proportional distribution of the larvae within the ROI should be relatively unchanged) but this might not be the case. Furthermore, we note that the images that we took were not from the same perspective every time. This may impact the estimates of the larvae counts due to the projection of a 3‐dimensional image onto a 2‐dimensional plane. This should not affect the comparison of larvae counts in the left and right sides of the ROI for each individual image but would be an issue if we wanted to compare the counts of larvae across multiple images. It appears also that changing the comparison regions did not impact greatly the results as long as the analysis was done consistently. Furthermore, we note that changes in perspective were mainly due to changes in the vertical location of the camera (as the water chamber was centered in the frame with each image taken). As our study focused on the horizontal movement of the sea urchin larvae, vertical changes in perspective should not have a great impact on the comparison of horizontal larvae distributions. Thus, overall, we do not expect the changes in perspective to greatly affect our final results. We do acknowledge that future experiments should have a fixed camera. Finally, we note that the circular polarizers used in our setups have a polarizing efficiency of >99.9%. The fact that the efficiency is not perfect could mean that some larvae were exposed to unpolarized light. However, the corresponding flux would be extremely small so the larvae would need to have extremely sensitive photoreceptors to detect it. In the future, experimental improvements could involve reducing the glare and contaminants, keeping a constant perspective for taking images, increasing the image quality for more accurate counting of the larvae as well as increasing the number of runs. However, since the glare is more of a visual artifact on the results and is only present at the edges of the water chamber such that it represents only a small region of the testable area, we do not expect the results to change significantly We also note that the larvae were exposed to different intensities under the different lighting regimes of unpolarized halogen, polarized halogen and darkness, which could have led to different behaviors. However, the intensity of horizontally and vertically polarized light was equal, as the same filter was used for both. Therefore, any differences in larval behavior for the linear polarization modes were not caused by changes in intensity. Furthermore, given the statistical similarity of larvae swimming direction between horizontally polarized light and unpolarized light as well as for vertically polarized light and darkness, it is unlikely that the different intensities of light played a major role in the behaviors of the larvae (see Appendix C).

Halogen illumination can generate weak thermal gradients; however, in our small, closed cylindrical tank any convective flow would be negligible (Rayleigh number well below the threshold for organized circulation). Using a linear or circular polarizer would effectively reduce the photon flux in the aquarium by half compared to unpolarized halogen runs. As a result, the water temperatures between the halogen, darkness, and polarized experimental runs could have differed. However, if the movement of the sea urchin larvae was caused by changes in water temperature, then the reactions of the larvae should have been the same when exposed to horizontally and vertically polarized light, as these runs would have delivered a similar photon flux to the aquarium. From our results, we observed that this was not the case, indicating that the differences were caused solely by the different polarization modes.

Findings

Here we showed that the larvae of (C. rodgersii) were sensitive to white (halogen) light, exhibiting positive phototaxis. In contrast, larvae sank to the bottom of the chamber or showed no visible swimming response in the darkness (no light exposure). We also show that the polarization of light matters to larvae swimming patterns. Positive phototaxis of larvae was observed when they were exposed to horizontally polarized light but not to vertically polarized light. This trend does not appear to be the case when the light source was moved to a higher angle to mimic elliptical polarization (which includes light that is polarized both in the horizontal and vertical directions). This indicates that the larvae respond to polarized light parallel to the ground, that is, horizontally polarized light.

We also found no statistical difference in larvae distributions in response to horizontally polarized light and unpolarized light (p‐value = 0.9947), nor were there any differences between vertically polarized light and dark (p‐value = 0.4969). This suggests that the larvae have the same reaction to vertically polarized light as they have to no light exposure. This could mean that the photoreceptors of larvae are only responsive to horizontal polarization.

Additionally the sea urchin larvae had positive phototaxis under illumination of both right‐handed and left‐handed circularly polarized light, indicating sensitivity to circular polarization. To our knowledge, there is no documented case of cells being intrinsically sensitive to circularly polarized light. ⁴⁸ For example, stromatopods are sensitive to circular polarization, but they possess specialized cells that act as quarterwaveplates that convert circularly polarized light into linearly polarized light. ⁴⁹ We purposefully couch our conclusions in behavioral, not molecular, terms. The data are consistent with CP sensitivity but stop short of proving a unique photopigment; alternative explanations (e.g., differential scattering or microlensing across the ciliary band) cannot be excluded at present. We therefore view the study as generating a testable hypothesis for future electrophysiological or ultrastructural work, paralleling the discovery trajectory of CP vision in stromatopods.

CONCLUSION

In this paper, we compared the swimming responses of the larvae of the sea urchin species C. rodgersii after exposure to unpolarized halogen, horizontally polarized halogen, vertically polarized halogen and to darkness. We found that there was a positive phototaxis response after 10 min of exposure to unpolarized and horizontally polarized halogen light and no phototaxis after 10 min of exposure to vertically polarized halogen light and darkness. In fact, larval response to unpolarized halogen and horizontally polarized halogen light were statistically equivalent as well as larval response to vertically polarized light compared to no light exposure. Our results suggests that not only do these sea urchin larvae have photoreceptors sensitive to light in the visible spectrum, but they also may have photoreceptors that are sensitive to horizontal polarization (and not sensitive to vertical polarization). Photoreceptor cells usually contain the retinal molecule 11‐cis‐retinal and a G‐coupled protein receptor forming the opsin protein. Upon a conformal change of the 11‐cis‐retinal to the all‐trans‐retinal, the opsin protein activates and sends a phototransducton cascade down the neuron. ⁵⁰ , ⁵¹ Photons with polarization parallel to the double carbon–carbon bonds present in the retinal molecules are absorbed and converted to an electrical impulse. ⁵² Hence, polarization sensitivity can arise when groups of photoreceptors are aligned with each other in aligned microvilli. ⁵³ Given that sea urchin larvae seem to possess Opsin2 ⁴ and are sensitive to horizontally polarized light, one may speculate that sea urchin larvae have photoreceptors cells that are aligned and preferentially horizontal.

The larvae also demonstrated a positive phototaxis response to circularly polarized light, indicating that they may possess a mechanism for circular polarization detection. However, as these larvae did not show a preference between left‐handed and right‐handed circular polarizations, one should investigate whether sea urchin larvae possess cells similar to the quarter‐wave plate cells of stromatopods. The absence of such cells would indicate a novel circular polarization detection mechanism.

These results all give an interesting indication of the potential polarimetric photosensitivity of larvae of C. rodgersii. It motivates the need for further testing with an increased sample size, and across other species of sea urchin to further strengthen the conclusions. This could lead to a new understanding of underwater vision and possibly lead to the development of new technologies that use biomimicry for polarization detection.

CONFLICT OF INTEREST STATEMENT

The authors declare no competing interests.

APPENDIX A. Spectral analysis

FIGURE A1.

Spectra of the halogen lamp used.

Figure A1 shows the spectrum of the halogen lamp used. We have intensity in counts on the y‐axis and the wavelength in nanometers on the x‐axis. The peak wavelength in around 640 nm. The shape of the spectrum nicely resembles the spectrum of sunlight on the surface of Earth.

TABLE A1.

Measured intensities of light passing through the water chamber with a polarizer and after a cross polarizer.

Polarization	Intensity (a.u.)	Cross polarization intensity (a.u.)
Linear	565.07	6.99
Circular	846.62	8.10

Table A1 shows the measured intensities of the halogen lamp after passing through one polarizer (linear and circular) and also the measured intensities after their respective cross‐polarizer. A ratio of cross polarization intensity to polarization intensity of 0.5 would mean that it has become completely depolarized, while a ratio of 0 would mean that it has retained all of its polarization. We see that for linear polarization, the ratio is 0.0124, which means that over 97% if light passing through the water chamber retained its linear polarization. For circular polarization, this ratio is 0.0096, which means that over 98% of the light retained its circular polarization. This means that any depolarization effects from the interactions of light to the glass and the water can be safely ignored.

APPENDIX B. Elliptical and circular polarization data

In this appendix, we analyze the elliptical and circular polarization data and provide the results from the ANOVA tests for each.

In Table A2, we display the proportions of the larvae in the half of the ROI closer to the light source compared to the half further away after an arcsine transformation, for the elliptical polarizations as well as for unpolarized halogen light. Table A3 is the corresponding ANOVA table. We find $F$(2, 17) = 0.1290 and a p‐value = 0.8798, thus indicating that the response of the larvae to the two elliptical polarizations were statistically similar and also the same as their response to the unpolarized halogen light.

TABLE A2.

Proportion of larvae in the half of the ROI closer to the light source compared to the half further away for the two elliptical polarization setups and the unpolarized light setup. The numbers are obtained after an arcsine transformation.

Light regime	Mean	Standard deviation	n
Horizontal elliptical	1.029	0.1439	6
Vertical elliptical	1.058	0.09525	6
Unpolarized halogen light	1.041	0.05098	8

In Table A4, we display the proportions of the larvae in the half of the ROI closer to the light source compared to the half further away after an arcsine transformation, for left‐handed and right‐handed circular polarizations as well as for unpolarized halogen light. Table A5 is the ANOVA table for the data obtained for Table A4 and we can see again from $F$(2, 17) = 0.8498 with $p$ = 0.4449 that the larvae reacted the same to the left‐handed and right‐handed circular polarizations. They had the same reaction as for unpolarized halogen light.

TABLE A3.

ANOVA test based on the data from Table A2 for the sea urchin larvae response to elliptically polarized light.

Source of variations	Sum of squares	DOF	Mean squares	F(DFn, DFd)	p Value
Between data sets	0.002537	2	0.001268	F(2, 17) = 0.1290	0.8798
Within data sets	0.1672	17	0.009833
Total	0.1679	19

This shows that the larvae can detect elliptical and circular polarizations and do not show a preference for one of the elliptical polarizations nor for one of the circular polarizations.

TABLE A4.

Proportion of larvae responding with positive phototaxis for each light setup, after an arcsine transformation, comparing their counts in the half of the ROI closer to the light source to the half further away, for left‐handed and right‐handed circularly polarized halogen light, as well as for unpolarized halogen light. The numbers are obtained after an arcsine transformation.

Light regime	Mean	Standard deviation	n
Left‐handed	0.9977	0.1240	6
Right‐handed	1.084	0.1608	6
Unpolarized halogen light	1.041	0.05098	8

TABLE A5.

ANOVA test based on the data from Table A4 for the sea urchin larvae response to circularly polarized light.

Source of variations	Sum of squares	DOF	Mean squares	F(DFn, DFd)	p Value
Between data sets	0.02243	2	0.01122	F(2, 17) = 0.8498	0.4449
Within data sets	0.2244	17	0.01320
Total	0.2468	19

APPENDIX C. Intensity comparisons

In this section, we have collated the differences in intensities and sea urchin larvae responses under illumination of different polarizations of light as shown in Table A6. We can see that the intensity ranges of our experiments (290–1750) lux did not seem to be a large factor in the phototaxis of sea urchins and that the polarization state was more important.

TABLE A6.

Comparison of the responses of the sea urchin larvae to different intensities of the different lighting regimes.

Light regime	Intensity (Lux)	Phototaxis
Darkness	0	None
Unpolarised halogen	1750	Positive
Horizontally polarised	400	Positive
Vertically polarised	400	None
Horizontal elliptical	290	Positive
Vertical elliptical	290	Positive
left‐handed circular	600	Positive
right‐handed circular	600	Positive

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.