Vision plays a major role in the life of most teleosts, and is assumingly well adapted to each species ecology and behaviour. Using a multidisciplinary approach, we scrutinised several aspects of the visual system and ecology of the Great Barrier Reef anemonefish, Amphiprion akindynos, including its orange with white patterning, retinal anatomy and molecular biology, its symbiosis with anemones and sequential hermaphroditism. Amphiprion akindynos possesses spectrally distinct visual pigments and opsins: one rod opsin, RH1 (498 nm), and five cone opsins, SWS1 (370 nm), SWS2B (408 nm), RH2B (498 nm), RH2A (520 nm), and LWS (554 nm). Cones were arranged in a regular mosaic with each single cone surrounded by four double cones. Double cones mainly expressed RH2B (53%) in one member and RH2A (46%) in the other, matching the prevailing light. Single cones expressed SWS1 (89%), which may serve to detect zooplankton, conspecifics and the host anemone. Moreover, a segregated small fraction of single cones coexpressed SWS1 with SWS2B (11%). This novel visual specialisation falls within the region of highest acuity and is suggested to increase the chromatic contrast of Amphiprion akindynos colour patterns, which might improve detection of conspecifics.
Vision plays a major role in the life of most teleost fishes to enable foraging, avoidance of predators, navigation, and mate choice (e.g., reviewed in1,2,3). Visual tasks may differ between species and individuals, and also depend on the light habitat or environment.
Visual adaptations may be found at the optical level. For example, filters present in the lens or cornea may block specific wavelengths of light, and thus indirectly shape an organism’s visual sensitivity before light reaches the retina4,5,6.
At the retinal level, the main adaptations are found in the contents of photoreceptors and the interconnections of ganglion cells. Photoreceptors are composed of rod and cone cells7 that house light-sensitive visual pigments8. Rods possess the highly sensitive rod opsin-based photopigment (RH1) and mediate vision in scotopic conditions. Cones contain up to four cone opsin-based photopigments being short- (SWS1 and SWS2), medium- (RH2) and long-wavelength-sensitive (LWS)8, are active in photopic conditions, and mediate colour vision. Teleost cones can further be divided based on their morphology into single cones and double or twin cones (i.e., two single cones fused together)9. Ganglion cells directly send the visual information to the central nervous system and their receptive field ultimately set the upper limit of spatial resolving power or visual acuity10.
The type, composition, arrangement and density of both photoreceptor and ganglion cells varies between species (interspecific variability) as well as across the retina of a single species (intraretinal variability), and has been shown to be directly linked to a species’ ecology and behaviour3,11,12,13,14. Interspecific variability in the number and type of photoreceptor cells correlates well with the light habitat and lifestyle of a species. Nocturnal species, for example, have more rods than diurnal ones and most deep-sea species have completely lost their cones in favour of a rod-only retina13. Conversely, diurnal coral reef fishes that live in a very bright and colourful environment, often have acute colour vision provided by complex photoreceptor mosaics of several cone types being sensitive to a wide spectrum of light15. Also, interspecific variability in the number and connectivity of ganglion cells, and therefore the visual acuity, often reflects the feeding behaviour of a species10.
Intraretinal variability is often assessed by constructing topographic maps of photoreceptor and ganglion cell distribution. This allows the identification of regions of high-cell densities or specialisations, which provide a higher sensitivity or acuity in a specific part of the visual field of the animal. The retinal topography usually reflects the structure of the habitat (terrain theory14,16), the behavioural ecology (e.g., feeding strategy17, predator avoidance18), the light environment10, and the life stage of an animal19.
At the molecular level, opsin genes have evolved by gene duplication20, deletion and conversion (e.g.,21), by mutation of the opsin sequence itself (e.g.,22), and by variability in opsin gene expression23. Changes in opsin gene repertoire and expression have been found to facilitate visual system adaptation to the prevailing light habitat (e.g.,24,25,26,27,28,29), predation density30 and/or feeding strategies25,31. This might also include intraretinal differences in opsin gene expression to optimise vision in different directions11. Recent research has shown that the coexpression of spectrally distinct opsin genes in double32 and single33 cones, together with intraretinal differences in distribution, may enable concise spectral tuning of the visual field to increase the detection of prey, predators and mates.
Damselfishes (Pomacentridae) are particularly well-suited to understand the mechanisms underlying visual system adaptation. With more than 390 species, they are an abundant, colourful, and diverse coral reef fish family34. They also possess some of the widest ranges of spectral sensitivities in the marine realm; species having three to four distinct visual pigments ranging in sensitivity from ultraviolet (UV) to long-wavelengths15,31. The anemonefishes (sub-family, Amphiprioninae) form a unique group within the damselfishes. They are famous not only for their highly contrasting colour patterns15,35,36, but also for their specialised ecology and life cycle. Besides living in close symbiosis with tropical sea anemones, anemonefishes are also sequential hermaphrodites37. Typically, anemonefishes form family groups with a size-dependent hierarchy consisting of several smaller sexually immature individuals and a dominant sexually mature pair out of which the largest individual is the female. Due to their striking appearance and lifestyle, we were interested to know how their visual system might be adapted to their visual ecology and behaviours.
We focused our study on the Barrier Reef Anemonefish, Amphiprion akindynos, a species mainly found on the Great Barrier Reef in Australia. Using high-throughput RNA sequencing (RNAseq), we first resolved opsin gene expression and tested whether gene expression varies across sex and/or the size of fishes. We then developed a fluorescent in situ hybridization method (FISH) to ascertain opsin specificity to a photoreceptor type, either single or double cones, identify potential opsin coexpression, and assess intraretinal variability in expression patterns. Furthermore, we determined the spectral absorbance of visual pigments by a combination of direct photoreceptor absorbance measurements and estimates. We then combined intraretinal patterns in opsin gene expression with the visual pigments‘ spectral absorbance, and with photoreceptor and ganglion cell topographic maps. This allowed us to identify the different areas of their visual field and their potential functional differences and relate these to their ecology and behaviour. Finally, using the newly constructed visual system, we modelled how the visual capabilities of A. akindynos are linked to primary visual tasks such as feeding, and the detection of their host anemone, conspecifics and predators.
Opsin gene repertoire and gene expression
In total, retinal transcriptomes from ten A. akindynos individuals (four females, two males, four immature) were sequenced. We confirmed the repertoire of opsin genes known for A. akindynos38, and were able to obtain the complete coding regions of SWS1, SWS2B, RH2B, RH2A, LWS, and RH1 (for opsin accession numbers, see Table S1). Proportional expression as a fraction of the total of all opsin genes expressed was 50.3 ± 12.0 (mean and standard deviation) for rods and 49.7 ± 12.0 for cones. Proportional expression as a fraction of total cone opsin expression was for: SWS1 = 6.8 ± 3.6%, SWS2B = 0.7 ± 0.5%, RH2B = 49.2 ± 3.0%, RH2A = 42.0 ± 5.0%, and LWS = 1.3 ± 1.5% (Fig. 1ai). Proportional expression as a fraction of single and double cone gene expression calculated separately, resulted in single cone expression of SWS1 = 89.3 ± 10.7% and SWS2B = 10.7± 10.7%, and double cone expression of RH2B = 52.8 ± 4.1%, RH2A = 45.9 ± 4.6%, and LWS = 1.4 ± 1.7% (Fig. 1aii). No significant correlations between expression patterns and sex or size of individuals were found (Table S2).
We used two different methods to gather visual pigment spectral sensitivities; measurements of opsin protein absorbance (MSP), and estimates based on amino acid sequences (Fig. 1b, Table S3). Mean λmax of different photoreceptor types found in A. akindynos (MSP-based; n fish = 3, being kept in aquaria for 4 months (n = 1), or 6 months (n = 2)) were matched with opsin genes within the damselfish family28,31, and are listed in Table S3. Exemplary absorption curves for each visual pigment are shown in Fig. S1. A summary of estimated λmax values is given in Table S3. The aminoacids that were considered for this approach and estimated substitution effects thereof are listed in Table S4.
In theory, λmax values should match between approaches. This was the case for the rods, where MSP and estimates revealed a λmax of 498 ± 4 nm (n = 23) and 496 nm, respectively. Also, for medium-wavelength-sensitive double cone members, MSP revealed a ‘green’ visual pigment with 520 ± 5 nm (n = 13) λmax which matched the estimations for RH2A (516/518/523 nm λmax), and a ‘blue’ visual pigment with 498 ± 4 nm (n = 12) λmax which matched one estimate for RH2B (498 nm λmax). It is worth noting that RH2B estimates varied depending on the reference species used (Table S4). Since those differences are not fully explained by substitutions in known tuning sites, additional undescribed tuning sites or more general differences in the protein may cause λmax shifts in damselfishes. MSP and estimates differed for the long-wavelength-sensitive double cone member and for the single cones. While MSP λmax for the ‘red’ visual pigment was 541 nm (n = 1), the LWS λmax estimate was 554 nm. However, MSP might not be reliable as it was only based on one cell due to the low LWS expression (Fig. 1a), and showed a lot of deviation from the visual pigment nomogram. Furthermore, single cone λmax by MSP was 400 ± 3 nm (n = 6), but λmax estimates were 370 nm for SWS1 (‘UV’) and 408 nm for SWS2B (‘violet’).
As lenses have been shown to be the primary light-filter of the damselfish eye4, we measured lens transmittance [defined by the wavelength of 50% of maximal transmittance (T50)] in 8 individuals (3 females, 2 males, 3 immatures), and found that the lenses were UV-transmitting independent of ontogenetic stage. T50 was 358 ± 3 nm for females (Fig. 1b), 352 ± 9 nm for males, and 328 ± 8 nm for immature specimens.
We gathered spectral reflectance of one species of host anemone, and three living individuals of A. akindynos of different life/sex stages. No differences in reflectance spectra were observed between life/sex stages. Figure 1c illustrates spectral reflectance data for A. akindynos (i), and one of its host anemone (ii). White stripes reflect from <400 nm (UV) to beyond 700 nm (far-red, not shown in figure) with a peak around 520 nm, while oranges show a small peak in the UV (360–380 nm) and another larger one at longer wavelengths (500–700 nm). The anemone reflects from the UV to the far-red (not shown in figure) with peaks around 600 nm.
We performed visual modelling to explore how the different visual pigments might contribute to a fish’s ecology including conspecific, prey, and predator detection. Damselfishes are potentially trichromatic in colour sense but we first considered the optimal monochromatic visual pigment for detecting contrasts as this helps delineate which pigments are best for each visual task. As results were insensitive to which lens was used (that for a male, female or immature), we used the male lens for all subsequent results. As this study proves that single cone expression is very specific for A. akindynos, i.e., coexpression is only found in a localised area, we then modelled how coexpression in the single cones could affect a typical trichromatic A akindynos visual system.
The best monochromatic pigment varied with target combinations (Fig. S2). For the majority of A akindynos colours versus each other (Fig. S2a), versus the host anemone (Fig. S2b), or versus the horizontal spacelight (Fig. S2c), the best visual pigment is in the 444–450 nm range. This is the location where the colours of the white stripe and the orange patches were most different from each other and from the horizontal radiance and anemone backgrounds. The one exception for fish colours was the contrast for the white stripe against the side welling radiance, that was better for a very short pigment (370 nm). Non-fish colours also showed specific optima (Fig. S2d). The anemone was best distinguished from the horizontal radiance at 370 nm, and so was zooplankton, although the wavelength dependence for the latter was minimal. Similarly, the wavelength dependence of detecting a dark looming predator was relatively flat, though slightly improved at a longer wavelength of 520 nm. This would make it a matching pigment to the average horizontal radiance.
For the trichromatic A. akindynos, we compared the colour contrast of different target combinations while varying how much SWS2B was coexpressed with SWS1 in the single cones (Fig. 1d). We found that the majority of fish colours (A. akindynos) against each other (Fig. 1di), or against the anemone (Fig. 1dii) or background spacelight (Fig. 1diii), improved for higher SWS2B expression. This was particularly true for the body orange colour against white stripe. In addition, the head orange against the anemone or white stripe was best distinguished when having either very low or high SWS2B coexpression.
For detecting the white stripe (Fig. 1diii) or anemone (Fig. 1div) versus horizontal radiance, higher SWS1 expression was found to be best. For zooplankton and predator detection against the spacelight, those were relatively insensitive to single cone coexpression (Fig. 1div).
Cone photoreceptor arrangement
Cones in A. akindynos were arranged in a regular fashion forming a mosaic composed of single cones each surrounded by four double cones (Fig. 2a), with a double to single cone ratio of 2:1.
Fluorescent in situ hybridization (FISH) of opsin genes
FISH on retinas of A. akindynos illustrated that the longer wavelength-sensitive genes, RH2A, RH2B and LWS, were expressed in double cones only. RH2A was expressed in one member and RH2B in the other member in >99% of double cones (Figs 2b and S3a). LWS was only detected in very few cells (<1%; Figs S4 and S5).
The short-wavelength-sensitive genes, SWS1 and SWS2B, were expressed in single cones only (Figs 2c and S3b). While all single cones expressed SWS1 (Fig. 2ci), SWS2B was found to be coexpressed with SWS1 in very few cells (Fig. 2cii). Moreover, the coexpression of SWS genes was restricted to a specific small region in the temporal part of the retina as shown by whole retina scans (Fig. S3bii).
Topographic distribution of ganglion cells, cone photoreceptors, and opsin genes
For the retinal topography analyses, the eyes of several individuals were analysed: seven for photoreceptor analyses (3 females, 2 males and 2 immature), three for ganglion cell analyses (1 female, 1 male and 1 immature), and eleven for in situ analyses (4 females, 3 males, 4 immature). The topographic distribution of ganglion cells, cone photoreceptors, and opsin genes did not differ between female, male and subadult individuals. As such, only results for one female are presented here (Fig. 3), with the rest of the results available in the Supplementary Figs S5–S10.
Ganglion cell densities
There were only few differences in the ganglion cell topographies between individuals at different ontogenetic stages (Fig. S6). All three individuals analysed possessed a horizontal streak (i.e., elongated increase in cell density along the retinal meridian) with a peak density of cells in the temporo-central part of the retina close to the optic nerve (Fig. 3a). The horizontal streak was more pronounced in the temporal part of the retina than elsewhere. The total number of ganglion cells varied greatly with the size of the fish, from 432,756 cells in the immature individual (4.1 cm total length (TL)) to 969,964 cells for the female (8.8 cm TL; Table 1). The peak ganglion cell densities were similar between individuals ranging from 40,400 cell/mm2 in the female to 45,000 cell/mm2 in the immature individual. Consequently, the spatial resolving power (SRP) varied greatly between ontogenetic stages, becoming higher as the fish gets bigger and ranging from 3.63 to 5.80 cycles/deg in immature and female fish, respectively (Table 1). Using these SRP values, a female or an immature fish should be able to distinguish another individual that measures 8 cm TL from up to 26 m and 16 m away, respectively. They should also be able to distinguish a single white stripe, 5 mm in width, that corresponds to the size of a white stripe on a 8 cm TL fish, from 1.6 m and 1 m away, respectively.
The regular cone mosaic pattern of one single cone surrounded by four double cones (Fig. 2a) was consistent over the entire retina, resulting in similar topographic distributions for each cell type. As a result, we only provide and describe in detail the distribution pattern for total cones. The topographic distribution of total cones varied slightly between individuals irrespective of their size and sex (Fig. S7). However, the overall topography did not differ between fishes. Similar to the ganglion cell distribution, cone densities formed a horizontal streak with a peak density in the temporal or temporo-central part of the retina. However, this streak was generally wider for cones than ganglion cells, extending vertically into the temporal part of the retina (Fig. S7). This resulted, for some individuals, in a very wide area of increased cone cells densities in the temporal part of the retina (Figs 3bi, S7b,d). Other individuals possessed a less pronounced but more defined vertical component in the temporal part of the retina as well as two areas (i.e., concentric increases in cell densities) in the nasal and temporal part of the retina within the horizontal streak (Fig. S7e,f).
Total cone numbers varied from 667,466 for the smallest individual (3.2 cm TL) to 1,884,932 for one of the larger individuals (7.3 cm TL; Table 2). The peak total cone densities varied from 61,200–89,200 cell/mm2, with an average of 76,286 cells/mm2 (Table 2). However, since no subsampling was performed for photoreceptor counts, these values may not necessarily represent the highest cone density for each individual. For one individual (immature #2; Tables 1 and 2), both ganglion cells and photoreceptors were mapped (Figs S6c and S7f, respectively). For this individual the ratio of cone to ganglion cells varied from 1.5:1 in the peak ganglion cell area to 3:1 in the ventral part of the retina, with an average of 2:1 across the retina.
Opsin gene densities
The topographic distribution of cone photoreceptors expressing specific opsin genes was analysed using FISH on whole mounted retinas. Since SWS1 was found to be expressed in all the single cones and RH2A/RH2B in the vast majority of double cones, we expected to find similar topographies and densities as for single cone and double cone topographic maps. While this was more or less the case for some individuals (Fig. 3), topographic maps from FISH data were in general less reliable, showing greater variation in topography between individuals and underestimating densities (Figs S6 and S7). This can be explained by the more intrusive methods used during FISH resulting in the loss of several photoreceptor cells. An example of such an under-labelled retinal region is shown in Fig. S8c.
Nevertheless, FISH-based topographic maps were very useful to identify the peculiar distribution of the cones expressing SWS2B, which were found to be coexpressed with SWS1 in a small area of the temporal part of the retina. SWS2B cone densities formed an area centralis with a peak density of cells toward the retinal margin of the temporal area (Fig. 3ciii). This specific distribution was consistent between the three individuals irrespective of sex and size (Fig. S10). This small area centralis at the temporal margin of the retina was therefore the only area in the retina that coexpressed two opsin genes. This area also showed a high-density of ganglion cells, although lower than the area of highest ganglion cell density located toward the centre of the retina (Fig. 3a,ciii).
Topographic maps of LWS cone densities could not be created due to the extremely low number of cells expressing this gene (<0.1% of all double cones). However, we were able to highlight labelled LWS cones across the retina (Fig. S5), which revealed that they are randomly distributed throughout the retina.
The A. akindynos cone opsin expression (SWS1, SWS2B, RH2B, RH2A, and LWS, Fig. 1a) was similar to that known in other damselfishes31 and appears generally well-suited to its ecology. As such, the prominently expressed mid-wavelength-sensitive RH2B and RH2A genes, are matched to the environmental spacelight on coral reefs39. As for other damselfishes4,31, A. akindynos had UV-transmitting lenses and expressed the UV-sensitive SWS1 in its single cones. In these fishes, UV vision (Fig. 1a,b) together with UV-reflective body parts (Fig. 1c) indicate the use of a close-range ‘private’ communication channel, likely hidden from ‘UV-blind’ predators35,40. Although rare, SWS2B expression had previously been reported in some damselfishes31. However, to our knowledge, this is the first time SWS2B has been shown to be coexpressed with SWS1 in a coral reef fish. As discussed in detail below, the localised coexpression of these two genes (Fig. 3ciii) may improve the detection of host anemones and conspecifics. Finally, both retinal transcriptomes (i.e., low expression; Fig. 1a) and topographic maps (i.e., few cells, no pattern; Fig. S5) suggest a negligent or even non-functional role of the long-wavelength-sensitive LWS opsin in A. akindynos.
Beyond the ecological significance, the multi-facetted approach of this study highlights how using opsin gene expression alone, without knowing in which cells and where across the retina genes are expressed, may lead to misinterpretations about the visual system of an animal. For example, if looking at expression data only, SWS2B may appear of minor importance for anemonefish vision as it was only found to be expressed at very low levels (1% of total cones [Fig. 1ai] and 10% of single cones [Fig. 1aii]). However, by labelling SWS2B expressing cells and imaging them across the retina, we were able to show that it is coexpressed with SWS1 (Fig. 2c) in a restricted region (Figs 3ciii, S3b). In this temporal region of coexpression, the percentage of single cones expressing SWS2B goes up to 30–70%, which highlights that a low overall expression level may not necessarily equate to no functionality.
In accordance with findings from other fishes26,29,41, our data further suggests that changes in the light habitat may drive plasticity in opsin gene expression in A. akindynos. MSP revealed a consistent λmax of ~ 400 nm for single cones in fishes that were kept under laboratory conditions for an extended period of time. This spectral sensitivity is longer-shifted compared to the estimated 370 nm λmax for the UV-sensitive SWS1 (measured at 347–376 nm λmax in other damselfishes15), but is shorter than the estimated 408 nm λmax for the violet-sensitive SWS2B (Table S3). This suggests that single cones in fishes kept under artificial lighting coexpressed SWS2B and SWS1 across the retina. Since under natural lighting conditions coexpression was restricted to a small area (Figs 2ciii, S3b), a substantial change in gene expression likely occurred when transferring animals to the laboratory. Luehrmann et al.29 found that in two related damselfish species, SWS1 expression decreased while SWS2B expression increased within a few weeks when moving fish to an artificial, UV-deprived light environment. Our findings, along with others42, emphasise the need to take the environment and especially the light habitat into account when drawing conclusions about the visual ecology of an animal. They also show that opsin gene expression and by extension spectral sensitivities only represent the state of a specimen at the time of sampling, a state that may change over time or spatial distribution of the fish (e.g., ontogeny43, time of day44, season45, or depth28,46).
The retinal topography of teleost fishes, and especially their ganglion cell topography, closely reflects the ecological characteristics of species16,17,18,47,48. In reef fishes, a clear link between the type of retinal specialization and the symmetry of the visual environment or habitat has been established. While species living in an open environment with an uninterrupted view of a horizon, like the sand to water interface, usually have streaks17, species living in a more enclosed, 3D environment with an interrupted horizon, have one or multiple areae of high cell densities47. Anemonefishes live in close association with their anemone host and as such, have relatively enclosed and restricted habitat ranges where area specialisations might be expected. Contrary to this expectation, A. akindynos possess a well-defined horizontal streak with a peak density in the centro-temporal part of the retina (Fig. 3a). A horizontal streak allows fishes to concentrate on a broad range of the environment without distinctive eye and head movements, and may provide a better threshold for movement detection17. While this type of specialisation may not fit perfectly with the symmetry of the anemonefish’s habitat, it may be very useful in light of their behaviour and peculiar social hierarchy.
Anemonefishes are highly territorial fishes and will show aggressive behaviour toward any inter- or intraspecific intruders49,50. A horizontal streak may therefore allow them to visualise their environment at a long-range and look for possible intruders while staying within the safety of the anemone. The peak-cell density in the central part of the retina provides higher acuity in the monocular field of view screening toward either side of the fish. This may assist during aggressive encounters at a close-range with conspecifics. Anemonefishes use body size to gauge the social status of conspecifics51 and this may be assessed through side-by-side swimming at the beginning of an encounter49.
Less is known about the retinal topography of photoreceptors in teleost fishes, but it generally matches the pattern seen in their ganglion cells33,52. Discrepancies between photoreceptor and ganglion cell topographies may indicate different visual requirements in different parts of the visual system. In addition to the horizontal streak, the photoreceptor pattern of A. akindynos shows a vertical component in the temporal part of the retina (Fig. 3b), therefore providing increased sensitivity in the field of view situated above and below the fish. This vertical component could be associated with the visualization of vertical body stripes used for species recognition53, or help with the detection of predators/intruders situated above or below them.
Adult A. akindynos have an acuity (SRP) of 5.8 cycles per degree which is relatively low compared to other coral reef fishes (3–27 SRP) and is the lowest recorded so far for any damselfish species10. Anemonefishes live in relatively small territories close to their host anemone and interact over a close-range with conspecifics and intruders. As a result, none of their behavioural tasks appear to require high acuity. Spatial resolving power estimation and behavioural observation show that their acuity is high enough to recognise or at least detect conspecifics, even at the level of a single white stripe, from few meters away. Interestingly, immature specimens that occupy smaller territories compared to adults50, also have a lower acuity, which restricts the distance at which they can spot a conspecific.
Opsin gene distribution in double cones across the retina was found to show a pattern of RH2B expression in one member and RH2A in the other. Consequently, their distribution matches the photoreceptor topography i.e., a horizontal streak with a peak density in the centro-temporal part and a vertical component in the temporal part of the retina (Fig. 3ci). The double cone photopigments are well matched to the environmental spacelight on coral reefs (i.e., from blue to green39), which may help detect intruders against the background. Indeed, visual models show that the detection of looming dark objects (e.g., predators or competitors) is best achieved at longer wavelengths, similar to the λmax of the RH2 photopigments. Interestingly, cichlids also tune their double cone sensitivities to the prevalent light environment presumably to increase the detection of prey, predators, or mates32.
Topographic analysis of single cone opsin densities across the retina highlighted: (1) SWS1 is present in every single cone and therefore has the same topography as the photoreceptors and the double cone opsins (Fig. 3cii); and (2) a novel specialisation in fishes, in the form of an area temporalis formed by the coexpression of SWS2B with SWS1 (Fig. 3ciii). Our monochromatic modelling showed that UV-sensitivity, as provided by the pure expression of an SWS1-based visual pigment, helps, albeit not in an outstanding manner, in luminance detection of zooplankton (Fig. S2d). This agrees with previous findings looking at the function of UV-sensitivity in fish foraging54,54. Furthermore, the mono- and trichromatic models also revealed that the use of SWS1-based single cones increases the colour contrast of the host anemone against spacelight (Figs 1div and S2d). The colour contrast of conspecifics, however, varied depending on the degree of SWS2B coexpression and the patterns that were compared (Fig. 1di–iii).
While this is the first report of SWS2B and SWS1 being coexpressed in a coral reef fish, their orthologs have previously been found to be coexpressed in the single cones of the cichlid Maylandia zebra33. However, both the distribution and function of the coexpression seem to differ between species. In M. zebra, coexpression showed a very high intraspecific variability both in topography and level of gene expression. On the other hand, in A. akindynos, all individuals showed a similar topographic pattern irrespective of sex/life-stage, indicating that the coexpression has a defined purpose in this species. Furthermore, in the cichlid opsin gene coexpression was prevalent throughout the retina except for the central part of the retina representing an area of high ganglion cell density. In the anemonefish, by contrast, the coexpression was restricted to the temporal part of the retina to an area of high cell density (Fig. 3ciii). Finally, visual modelling showed that this coexpression is likely to increase achromatic contrast detection in the peripheral visual system of M. zebra33. In A. akindynos, on the other hand, coexpression may improve the chromatic contrast of orange body colour versus white stripes of a conspecific situated in front of the fish as well as orange body colour versus the anemone host and the spacelight (Fig. 1di–iii). In this case the trichromatic models are similar to the monochromatic ones in that having a visual pigment close to 450 nm improves contrast between fish colours and other targets. Since the single cone visual sensitivity is closest to this wavelength, it is the one best able to improve contrast. Higher SWS2B expression enables the single cone to get closer to the 440–450 nm optimum. This specialised area of coexpression might then have evolved to better recognize conspecifics to maintain hierarchy and avoid territorial bouts. It might also have evolved to readily recognize other anemonefish species when sharing a host, thus generally reducing aggression between species. Detailed investigations of the visual systems/ecology of related damselfish species and of anemonefishes in particular, as well as behavioural experiments will help answer these questions in the future.
All A. akindynos specimens were either collected on reefs surrounding Lizard Island, Australia, or obtained from an aquarium supplier (Cairns Marine) collecting fishes from the Northern Great Barrier Reef. Fish were anaesthetized with Tricaine methanesulfonate (Sigma-Aldrich) or an overdose of clove oil, and killed by decapitation. Specimens were classified as subdominant males, dominant males (as defined by the presence of testes and/or being the largest male of the group), and females (as defined by the presence of ovaries and/or being largest individual of the group), and size (total length) as well as affiliation to the same family (as defined as specimens living in the same anemone) were noted. A summary of the number of individuals used with their sex, size and family, where they were sourced from, and for which type of analysis they were used is provided in Table S5. A detailed description of the methods used in this study is provided in the Supplementary Information and move the whole sentence and place it directly under Methods (before Specimen Collection).
Opsin gene study
Transcriptomic sequencing and processing
For each individual, the retina from one eye was dissected out and preserved in RNAlater (Ambion) until further processing. Transcriptomes were sequenced on an Illumina HiSeq 2000 and processed following previously published methods21,55. Trinity was used for de-novo assembly of transcripts that were mapped to known and publicly available opsin genes of A. akindynos (HQ286499, HQ286509, HQ286519, HQ286529, HQ286539, HQ286549). Further bioinformatics analyses were performed using Geneious software (Version 9.0.4).
Opsin gene expression
We analysed proportional cone and rod opsin expression as a fraction of the total of all opsin genes expressed. To test for cone expression and as this study shows evidence that damselfish single cones only express the SWS opsin genes whereas double cones express RH2 and LWS opsin genes, we analysed the relative expression of opsin genes as a fraction of either total cone expression or as single and double cone opsin expression separately.
In order to test whether sexual status or size had an effect on cone opsin gene expression, we used a beta regression as implemented in the R package BETAREG56, which allows handling of non-transformed data to model percentages and proportions. Statistical analyses were performed in R57 using the interface RSTUDIO v.0.98.1062.
Visual pigments maximal absorbance (λmax)
Due to the limitations of MSP in gathering information on visual pigments that are sparse across the retina or expressed at low-levels, and to additionally confirm λmax of opsins measured by MSP, we further estimated the λmax of each opsin protein based on differences of their amino acid sequences to reference species. We focused on variable amino acid residues that either occurred in areas corresponding to the retinal binding pocket and were substitutions resulted in a change in polarity, or at known tuning sites.
We gathered lens transmission curves (300–800 nm) following previously published protocols4. Transmission curves were normalized at 700 nm, and the wavelength at which 50% of the maximal transmittance (T50) was reached was determined using a linear regression5,60. We then assessed whether the A. akindynos eyes would be UV blocking (T50 > 400 nm) or UV transmitting (T50 < 400 nm).
Spectral reflectance of different areas of the fish with focus on white stripes and orange body as well as different areas of the anemone were measured following the methods described in Marshall et al.61.
Because very little is known about damselfish photoreceptor opponency, modelling was performed using the receptor noise limited model which does not require any explicit understanding of processing post-photoreceptor62,63. This enabled us to explore the relative function of various visual pigment combinations in performing different visual tasks. For the monochromatic modelling, we tried all the possible lens transmissions (male, female, and juvenile) to see if they had any impact on the results.
Visual tasks included conspecific, prey, and predator detection. Conspecific detection included comparing A. akindynos colours against each other or against the spacelight or an anemone. The foraging task was to detect zooplankton against the background spacelight. Since zooplankton scatters downwelling light, and has a somewhat flat reflectance into the UV64,65, we modelled zooplankton as having a flat reflectance of 10% and assumed it scattered the higher intensity downwelling irradiance. Detecting a predator was modelled as discriminating a dark grey object against the background spacelight as would occur for a looming stimulus. This was modelled as a dark grey with a flat reflectance of 10%, which was assumed to scatter the horizontal irradiance.
We calculated the quantum catch of different photoreceptors as they viewed the light coming from different targets. Fishes or anemone were illuminated by horizontal irradiance, while zooplankton was illuminated by downwelling irradiance. One additional target was the horizontal radiance from a small patch of the background space light. Downwelling irradiance, and horizontal radiance and irradiance were obtained from previous measurements in reefs around Lizard Island28.
We use the receptor noise limited model to quantify the relative discrimination of two targets based either on luminance or colour contrast62,63,66,67,68. Luminance contrast is calculated using the receptor noise-based contrast to be consistent with colour discrimination, providing the results in terms of just noticeable differences (JNDs) i.e., the threshold at which two objects should be distinguishable from one another.
As shown in Fig. 2, A. akindynos has one single cone for each pair of double cones so that nS: nM: nL = 1: 2: 2. We further set the Weber fraction, νL to be 0.1, based on colour experiments in other fishes69,70. Luminance noise has not been quantified, so we assume that the Weber fraction is similar for both colour and luminance68. In these calculations, we then vary either the peak λmax for a given photoreceptor or the amount of coexpression of the SWS1 and SWS2B opsin genes in the single cones. Since we are making relative comparisons between visual pigments with different peak sensitivities or different coexpression values, the actual Weber fraction does not impact our conclusions.
Retinal wholemount preparation
For photoreceptor and ganglion cell topographies
The lenses and corneas were dissected out and the eye cups fixed in 4% paraformaldehyde (PFA) in 0.1 M phosphate buffer saline (PBS, pH = 7.4) overnight at 4 °C. Retinal wholemounts were then prepared according to standard protocols71,72,73.
For in situ analyses on wholemount retinas
Eyes were enucleated and prepared following published methods74.
Fluorescent in situ hybridization (FISH)
We performed dual-labelling FISH following previously described methods32,75,76 with the difference that we reversed transcribed RNA using the High Capacity RNA-to-cDNA kit (Applied Biosystems), and subsequently used cDNA to generate probe templates by standard PCR using the MyTaqTM HSRED DNA Polymerase (Bioline) with primers (listed in Table S6) designed to target the 3′ untranslated region (3′UTR) (RH2B and RH2A) or the coding sequence (SWS1, SWS2B, and LWS).
Image acquisition of labelled opsins
For visualization of the distribution of labelled opsin genes throughout the retina, one exemplary whole retina scan for each dual-labelled opsin pair was performed using a spinning-disk confocal microscope (Spectral Applied Research).
Stereological analysis and topographic map construction
For each analysis, the outline of the retina was digitized using a x5 objective (numerical aperture 0.16) mounted on a compound microscope (Zeiss Imager.Z2) equipped with a motorized stage (MAC 6000 System, Microbrightfield, USA), a digital colour camera (Microbrightfield, USA) and a computer running Stereo Investigator software (Microbrightfield, USA).
Photoreceptor and ganglion cell densities
Following the protocols described in de Busserolles et al.17,77, topographic distribution of single cones, double cones, total cones and ganglion cells were assessed using the optical fractionator technique78 modified by Coimbra et al.79,80 for use in retinal whole mounts.
Topographic maps were constructed using the statistical program R v3.5.0 (R Foundation for Statistical Computing, 2018) with the results exported from Stereo Investigator software according to Garza-Gisholt et al.81. For each map we used the Gaussian Kernel Smoother from the Spatstat package82 and adjusted the sigma value to the grid size.
Opsin gene densities
For each wholemount, around 200 image stacks were randomly and systematically acquired using the SRS image stack flow function from the Stereo Investigator software. After additional processing using custom ImageJ scripts to remove background and enhance labelled photoreceptor cells, cells were counted for each colour channel in each counting frame.
Finally, topographic maps of opsin densities were constructed using R and a custom script adapted from Garza-Gisholt et al.81.
Spatial revolving power estimation
The upper limit of the spatial resolving power (SRP) in cycles degrees i.e., visual acuity, was estimated for each individual using the ganglion cell peak density as described by Collin & Pettigrew10.
Ethics approval and consent to participate
A. akindynos specimens were collected under the Great Barrier Reef Marine Park Permit G12/35005.1 and the Queensland General Fisheries Permit 140763. All experimental procedures were approved by The University of Queensland Animal Ethics Committee (QBI/223/10/ARC/US AIRFORCE (NF) and QBI/192/13/ARC), and all experiments were performed in accordance with relevant guidelines and regulations.
Marshall, J., Carleton, K. L. & Cronin, T. Colour vision in marine organisms. Curr. Opin. Neurobiol. 34, 86–94 (2015).
Carleton, K. L., Dalton, B. E., Escobar-Camacho, D. & Nandamuri, S. P. Proximate and ultimate causes of variable visual sensitivities: insights from cichlid fish radiations. Genesis 45, 299–325 (2016).
Busserolles, F. D & Marshall, N. J. Seeing in the deep-sea: visual adaptations in lanternfishes. Proc. R. Soc. B Biol. Sci. 372, pii: 20160070 (2017).
Siebeck, U. E. & Marshall, N. J. Potential ultraviolet vision in pre-settlement larvae and settled reef fish–a comparison across 23 families. Vision Res. 47, 2337–2352 (2007).
Siebeck, U. E. & Marshall, N. J. Ocular media transmission of coral reef fish - can coral reef fish see ultraviolet light? Vision Res. 41, 133–149 (2001).
Thorpe, A., Douglas, R. H. & Truscott, R. J. W. Spectral transmission and short-wave absorbing pigments in the fish lens—I. Phylogenetic distribution and identity. Vision Res. 33, 289–300 (1993).
Hunt, D. M., Hankins, M. W., Collin, S. P. & Marshall, N. J. Evolution of visual and non-visual pigments. Evolution of Visual and Non-Visual Pigments, (eds Hunt, D. M., Hankins, M. W., Collin, S. P. & Marshall, N. J.). Boston, MA: Springer US. (2014).
Yokoyama, S. Evolution of dim-light and color vision pigments. Annu. Rev. Genomics Hum. Genet. 9, 259–282 (2008).
Lyall, A. H. Cone arrangements in teleost retinae. Q. J. Microsc. Sci. 98, 189–201 (1957).
Collin, S. P. & Pettigrew, J. D. Quantitative comparison of the limits on visual spatial resolution set by the ganglion cell layer in twelve species of reef teleosts. Brain. Behav. Evol. 34, 184–192 (1989).
Temple, S. E. Why different regions of the retina have different spectral sensitivities: a review of mechanisms and functional significance of intraretinal variability in spectral sensitivity in vertebrates. Vis. Neurosci. 28, 281–293 (2011).
Shand, J. Ontogenetic changes in retinal structure and visual acuity: a comparative study of coral-reef teleosts with differing post-settlement lifestyles. Environ. Biol. Fishes 49, 307–322 (1997).
Walls, G. L. The vertebrate eye and its adaptive radiation. (Bloomfield Hills, Mich., Cranbrook Institute of Science, 1942).
Hughes, A. The topography of vision in mammals of contrasting life style: comparative optics and retinal organisation. In Handbook of sensory physiology (ed. Crescitelli, F.) 613–756 (Springer, Berlin, 1977).
Marshall, J. N., Vorobyev, M. & Siebeck, U. E. What does a reef fish see when it sees a reef fish. In Communication in fishes (eds Ladich, F., Collin, S. P., Moller, P. & Kapoor, B. G.) 393–422 (Science Publisher Inc, 2006).
Collin, S. P. A web-based archive for topographic maps of retinal cell distribution in vertebrates: Invited Paper. Clin. Exp. Optom. 91, 85–95 (2008).
de Busserolles, F., Marshall, N. J. & Collin, S. P. Retinal ganglion cell distribution and spatial resolving power in deep-sea lanternfishes (myctophidae). Brain. Behav. Evol. 84, 262–276 (2014).
Collin, S. P. & Pettigrew, J. D. Retinal topography in reef teleosts. I. Some species with well-developed areae but poorly-developed streaks. Brain. Behav. Evol. 31, 269–282 (1988).
Shand, J., Chin, S. M., Harman, A. M., Moore, S. & Collin, S. P. Variability in the location of the retinal ganglion cell area centralis is correlated with ontogenetic changes in feeding behavior in the black bream, Acanthopagrus butcheri (Sparidae, Teleostei). Brain. Behav. Evol. 55, 176–190 (2000).
Musilova, Z. et al. Vision using multiple distinct rod opsins in deep-sea fishes. Science. 364, 588–592 (2019).
Cortesi, F. et al. Ancestral duplications and highly dynamic opsin gene evolution in percomorph fishes. Proc. Natl. Acad. Sci. 112, 1493–1498 (2015).
Sugawara, T. et al. Parallelism of amino acid changes at the RH1 affecting spectral sensitivity among deep-water cichlids from Lakes Tanganyika and Malawi. Proc. Natl. Acad. Sci. USA 102, 5448–5453 (2005).
Hofmann, C. M. & Carleton, K. L. Gene duplication and differential gene expression play an important role in the diversification of visual pigments in fish. Integr. Comp. Biol. 49, 630–643 (2009).
Shand, J. et al. The influence of ontogeny and light environment on the expression of visual pigment opsins in the retina of the black bream, Acanthopagrus butcheri. J. Exp. Biol. 211, 1495–1503 (2008).
Hofmann, C. M., O’Quin, K. E., Smith, A. R. & Carleton, K. L. Plasticity of opsin gene expression in cichlids from Lake Malawi. Mol. Ecol. 19, 2064–2074 (2010).
Fuller, R. C. & Claricoates, K. M. Rapid light-induced shifts in opsin expression: finding new opsins, discerning mechanisms of change, and implications for visual sensitivity. Mol. Ecol. 20, 3321–3335 (2011).
Sandkam, B. A., Young, C. M., Breden, F. M. W., Bourne, G. R. & Breden, F. Color vision varies more among populations than among species of live-bearing fish from South America. BMC Evol. Biol. 15, 1–11 (2015).
Stieb, S. M., Carleton, K. L., Cortesi, F., Marshall, N. J. & Salzburger, W. Depth-dependent plasticity in opsin gene expression varies between damselfish (Pomacentridae) species. Mol. Ecol. 25, 3645–3661 (2016).
Luehrmann, M. et al. Short-term colour vision plasticity on the reef: changes in opsin expression under varying light conditions differ between ecologically distinct fish species. J. Exp. Biol. 221, jeb175281 (2018).
Sandkam, B., Young, C. M. & Breden, F. Beauty in the eyes of the beholders: colour vision is tuned to mate preference in the Trinidadian guppy (Poecilia reticulata). Mol. Ecol. 24, 596–609 (2015).
Stieb, S. M. et al. Why UV- and red-vision are important for damselfish (Pomacentridae): structural and expression variation in opsin genes. Mol. Ecol. 26, 1323–1342 (2017).
Dalton, B. E., Loew, E. R., Cronin, T. W. & Carleton, K. L. Spectral tuning by opsin coexpression in retinal regions that view different parts of the visual field. Proc. R. Soc. B Biol. Sci. 281, 20141980 (2014).
Dalton, B. E., de Busserolles, F., Marshall, N. J. & Carleton, K. L. Retinal specialization through spatially varying cell densities and opsin coexpression in cichlid fish. J. Exp. Biol. 220, 266–277 (2017).
Allen, G. R. Damselfishes of the World. (Mergus, Melle, Germany, 1991).
Marshall, N. The visual ecology of reef fish colours. In Animal signals: signaling and signal designs in animal communication (eds Espmark, Y., Amundsen, T. & Rosenqvist, G.) 83–120 (Tapir, 2000).
Merilaita, S. & Kelley, J. L. Scary clowns: adaptive function of anemonefish coloration. J. Evol. Biol. 31, 1558–1571 (2018).
Allen, G. R. The anemonefishes: their classification and biology. (T.F.H. Publications, 1975).
Hofmann, C. M. et al. Opsin evolution in damselfish: convergence, reversal, and parallel evolution across tuning sites. J. Mol. Evol. 75, 79–91 (2012).
Losey, G. S. et al. Visual biology of Hawaiian coral reef fishes. I. Ocular transmission and visual pigments. Copeia 3, 433–454 (2003).
Marshall, N. J. & Cheney, K. VISION|Color vision and color communication in reef fish. In Encyclopedia of fish physiology 1, 150–158 (Elsevier, 2011).
Nandamuri, S. P., Yourick, M. R. & Carleton, K. L. Adult plasticity in African cichlids: rapid changes In opsin expression in response to environmental light differences. Mol. Ecol. 26, 6036–6052 (2017).
Lythgoe, J. N. The ecology of vision. (Clarendon Press, 1979).
Cortesi, F. et al. From crypsis to mimicry: changes in colour and the configuration of the visual system during ontogenetic habitat transitions in a coral reef fish. J. Exp. Biol. 219, 2545–2558 (2016).
Johnson, A. M., Stanis, S. & Fuller, R. C. Diurnal lighting patterns and habitat alter opsin expression and colour preferences in a killifish. Proc. Biol. Sci. 280, 20130796 (2013).
Shimmura, T. et al. Dynamic plasticity in phototransduction regulates seasonal changes in color perception. Nat. Commun. 8, 1–7 (2017).
Fuller, R. C., Carleton, K. L., Fadool, J. M., Spady, T. C. & Travis, J. Population variation in opsin expression in the bluefin killifish, Lucania goodei: a realtime PCR study. J. Comp. Physiol. A. Neuroethol. Sens. Neural. Behav. Physiol. 190, 147–154 (2004).
Shand, J. Changes in retinal structure during development and settlement of the goatfish Upeneus tragula. Brain. Behav. Evol. 43, 51–60 (1994).
Shand, J., Archer, M. S. & Collin, S. P. Ontogenetic changes in the retinal photoreceptor mosaic in a fish, the black bream, Acanthopagrus butcheri. J. Comp. Neurol. 412, 203–217 (1999).
Iwata, E. & Manbo, J. Territorial behaviour reflects sexual status in groups of false clown anemonefish (Amphiprion ocellaris) under laboratory conditions. Acta Ethol. 16, 97–103 (2013).
Ross, R. W. Territorial behavior and ecology of the anemonefish Amphiprion melanopus on Guam. Z. Tierpsychol 36, 71–83 (1978).
Buston, P. Size and growth modification in clownfish. Nature 424, 145–146 (2003).
Champ, C., Wallis, G., Vorobyev, M., Siebeck, U. & Marshall, J. Visual acuity in a species of coral reef fish: Rhinecanthus aculeatus. Brain. Behav. Evol 83, 31–42 (2014).
Salis, P. et al. Ontogenetic and phylogenetic simplification during white stripe evolution in clownfishes. BMC Biol. 16, 1–13 (2018).
Browman, H. I., Novales-Flamarique, I. & Hawryshyn, C. W. Ultraviolet photoreception contributes to prey search behaviour in two species of zooplanktivorous fishes. J. Exp. Biol. 198, 187–198 (1994).
de Busserolles, F. et al. Pushing the limits of photoreception in twilight conditions: the rod-like cone retina of the deep-sea pearlsides. Sci. Adv. 3, eaao4709 (2017).
Cribari-Neto, F. & Zeileis, A. Beta regression in R. J. Stat. Softw. 34, 1–24 (2010).
Development Core Team, R. R: A language and environment for statistical computing. R foundation for statistical computing, Vienna, Austria. ISBN 3-900051-07-0 (2011).
Hart, N. S. Microspectrophotometry of visual pigments and oil droplets in a marine bird, the wedge-tailed shearwater Puffinus pacificus: topographic variations in photoreceptor spectral characteristics. J. Exp. Biol. 207, 1229–1240 (2004).
Hart, N. S., Theiss, S. M., Harahush, B. K. & Collin, S. P. Microspectrophotometric evidence for cone monochromacy in sharks. Naturwissenschaften 98, 193–201 (2011).
Douglas, R. H. & McGuigan, C. M. The spectral transmission of freshwater teleost ocular media - an interspecific comparison and a guide to potential ultraviolet sensitivity. Vision Res. 29, 871–879 (1989).
Marshall, N. J., Jennings, K., McFarland, W. N., Loew, E. R. & Losey, G. S. Visual biology of Hawaiian coral reef fishes. II. Colors of Hawaiian coral reef fish. Copeia 2003, 455–466 (2003).
Vorobyev, M. & Osorio, D. Receptor noise as a determinant of receptor thresholds. Proc. R. Soc. B Biol. Sci. 265, 351–358 (1998).
Vorobyev, M., Brandt, R., Peitsch, D., Laughlin, S. & Menzel, R. Colour thresholds and receptor noise: behaviour and physiology compared. Vision Res. 41, 639–653 (2001).
Johnsen, S. & Widder, E. A. Ultraviolet absorption in transparent zooplankton and its implications for depth distribution and visual predation. Mar. Biol. 138, 717–730 (2001).
Gagnon, Y. L., Shashar, N., Warrant, E. J. & Johnsen, S. J. Light scattering by selected zooplankton from the Gulf of Aqaba. J. Exp. Biol. 210, 3728–3735 (2007).
Siddiqi, A. Interspecific and intraspecific views of color signals in the strawberry poison frog Dendrobates pumilio. J. Exp. Biol. 207, 2471–2485 (2004).
Kinoshita, S., Yoshioka, S. & Miyazaki, J. Physics of structural colors. Reports Prog. Phys. 71 (2008).
Olsson, P., Lind, O. & Kelber, A. Chromatic and achromatic vision: parameter choice and limitations for reliable model predictions. Behav. Ecol. 29, 273–282 (2018).
Champ, C. M., Vorobyev, M. & Marshall, N. J. Colour thresholds in a coral reef fish. R. Soc. Open Sci. 3, 160399 (2016).
Escobar-Camacho, D., Marshall, J. & Carleton, K. L. Behavioral color vision in a cichlid fish: Metriaclima benetos. J. Exp. Biol. 220, 2887–2899 (2017).
Keeney, A. H. The wholemount handbook. A guide to the preparation and analysis of retinal wholemounts. Am. J. Ophthalmol. 92, 128 (1981).
Coimbra, J. P., Videira Marceliano, M. L., Da Silveira Andrade-Da-Costa, B. L. & Yamada, E. S. The retina of tyrant flycatchers: topographic organization of neuronal density and size in the ganglion cell layer of the great kiskadee Pitangus sulphuratus and the rusty margined flycatcher Myiozetetes cayanensis (Aves: Tyrannidae). Brain. Behav. Evol. 68, 15–25 (2006).
Ullmann, J. F. P., Moore, B. A., Temple, S. E., Fernández-Juricic, E. & Collin, S. P. The retinal wholemount technique: a window to understanding the brain and behaviour. Brain. Behav. Evol. 79, 26–44 (2012).
Barthel, B. L. K. & Raymond, P. A. In situ hybridization studies of retinal neurons. Methods Enzymol. 316, 579–590 (2000).
Allison, W. T. et al. Ontogeny of cone photoreceptor mosaics in zebrafish. J. Comp. Neurol. 518, 4182–4195 (2010).
Raymond, P. A. & Barthel, L. K. A moving wave patterns the cone photoreceptor mosaic array in the zebrafish retina. Int. J. Dev. Biol. 48, 935–945 (2004).
de Busserolles, F., Fitzpatrick, J. L., Marshall, N. J. & Collin, S. P. The influence of photoreceptor size and distribution on optical sensitivity in the eyes of lanternfishes (myctophidae). PLoS One 9, e99957 (2014).
West, R. A. Optical properties of aggregate particles whose outer diameter is comparable to the wavelength. Appl. Opt. 30, 5316 (1991).
Coimbra, J. P. et al. Number and distribution of neurons in the retinal ganglion cell layer in relation to foraging behaviors of tyrant flycatchers. J. Comp. Neurol. 514, 66–73 (2009).
Coimbra, J. P., Nolan, P. M., Collin, S. P. & Hart, N. S. Retinal ganglion cell topography and spatial resolving power in penguins. Brain. Behav. Evol. 80, 254–268 (2012).
Garza-Gisholt, E., Hemmi, J. M., Hart, N. S. & Collin, S. P. A comparison of spatial analysis methods for the construction of topographic maps of retinal cell density. PLoS One 9 (2014).
Baddeley, A. & Turner, R. spatstat: An R package for analyzing spatial point patterns. J. Stat. Softw. 12, 1–42 (2005).
We would like to thank Cairns Marine for supplying fish and the staff at Lizard Island Research Station, Lorenz Sueess and Genevieve Phillips for support during field work. We thank Janette Edson from the Queensland Brain Institute’s (QBI) Genomics Facility for library preparation and transcriptome sequencing, Rumelo Amor from the QBI Advanced Microscopy Facility for technical support, and Prof. Helen Cooper and Michael Langford (QBI) for providing lab facilities to conduct FISH experiments. This work was supported by the German Research Foundation (DFG) awarded to S.M.S., The Australian Research Council [Discovery Project (DP150102710) and the AFOSR/AOARD awarded to J.M., DECRA Project (DE180100949) awarded to F.d.B. and to J.M., the ARC DP180102363 awarded to J.M. and F.C., and a UQ Development Fellowship awarded to F.C.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Stieb, S.M., de Busserolles, F., Carleton, K.L. et al. A detailed investigation of the visual system and visual ecology of the Barrier Reef anemonefish, Amphiprion akindynos. Sci Rep 9, 16459 (2019). https://doi.org/10.1038/s41598-019-52297-0