Looking across the gap: Understanding the evolution of eyes and vision among insects

The compound eyes of insects exhibit stunning variation in size, structure, and function, which has allowed these animals to use their vision to adapt to a huge range of different environments and lifestyles, and evolve complex behaviors. Much of our knowledge of eye development has been learned from Drosophila, while visual adaptations and behaviors are often more striking and better understood from studies of other insects. However, recent studies in Drosophila and other insects, including bees, beetles, and butterflies, have begun to address this gap by revealing the genetic and developmental bases of differences in eye morphology and key new aspects of compound eye structure and function. Furthermore, technical advances have facilitated the generation of high‐resolution connectomic data from different insect species that enhances our understanding of visual information processing, and the impact of changes in these processes on the evolution of vision and behavior. Here, we review these recent breakthroughs and propose that future integrated research from the development to function of visual systems within and among insect species represents a great opportunity to understand the remarkable diversification of insect eyes and vision.


Insect compound eye structure and diversity
Insect compound eyes are made up of optical units called ommatidia.
Photons are focused through the corneal lenses onto the underlying photoreceptor cells (PRCs), specifically their rhabdomeres that contain rhodopsins (Figure 1; Box 1).A major source of the great diversity in insect eyes is differences in ommatidia number, diameter, and length.The number of ommatidia can vary hugely between species, for example, from 29 in a parasitoid wasp [1] to tens of thousands in dragonflies. [2]There is also extensive natural variation within species; the eyes of D. melanogaster strains, for example, can differ by hundreds of ommatidia. [3][6][7] For example, male Chrysomya megacephala have wide ommatidia in the dorsal two thirds of the eye and much narrower ommatidia in the ventral third. [8]Differences in the relative number

Acuity
Measure of the ability of eyes to resolve different objects.

Apposition eyes
Compound eyes with fused rhabdoms where the photoreceptors collect light from a single ommatidial lens and project to one laminal cartridge.

Connectomics
Mapping of the axonal connections between neurons in the nervous system, for example, in the optic lobe neuropils of insects.

Contrast sensitivity
Measure of the ability of eyes to detect patterns with decreasing contrast.

Inter-ommatidial angle
The angle between ommatidia along their optic axes.

Laminal cartridge
Column of the lamina underlying a single ommatidium.

Lamina monopolar cells
Laminal neurons that collect information from photoreceptors and connect to other cartridges and the medulla.

Microsaccades
Rapid movements of the eyes that shift the fixed point of gaze, achieved by moving the whole head or retinas.Photoreceptor microsaccades are the contraction of individual photoreceptors in response to light.

Neural superposition eyes
Apposition eyes with open rhabdoms where photoreceptors with the same field of view project axons to the same laminal cartridge, and photoreceptors within an ommatidia with different fields of view project to different cartridges.

Ommatidia
Subunits that make up the compound eyes that are composed of a single lens, photoreceptor cells, cone cells and accessory cells including pigment cells.

Photoreceptor cell
Light sensitive cells in the ommatidia of compound eyes that express rhodopsins in specialized rhabdomeres.

Rhabdom
The group of rhabdomeres within a single ommatidia, which can be arranged in an open or fused configuration.

Rhabdomere
Specialized photoreceptor cell sub-structure with a large surface area made up of thousands of microvilli presenting rhodopsins.

Rhodopsins
Photosensitive molecules localized to rhabdomeres and responsive to different wavelengths of light.Rhodopsins are composed of a Gprotein coupled transmembrane receptor (opsin) covalently bound to the chromophore ligand (retinal).

Superposition eyes
Compound eyes with fused rhabdoms where photoreceptor cells with the same field of view collect light from multiple ommatidial lenses to enhance light detection.and width of ommatidia often indicate regions with specialist functions, for example, larger frontal ommatidia of killer flies that hunt in flight. [7]en within the eye of D. mauritiana we have observed facet size differences up to 35%. [5]Ommatidia length also differs considerably within and between species, for example, rhabdomere length can vary from 0.46 to 1.1 mm within the eye of a dragonfly. [9,10]Differences in ommatidia number, diameter, and length affect the inter-ommatidial angles and determine how much light enters each ommatidium thus directly affecting acuity and contrast sensitivity [11][12][13] (Figure 1; Box 1).Addi-tionally, these parameters and the overall field of view vary with the curvature of the eye [11,12] (Figure 1).
The PRCs in each ommatidium express photo-sensitive rhodopsins (Rh) (Box 1) that can respond to different wavelengths of light.In Drosophila, the six outer PRCs (R1-R6) in each ommatidium express Rh1 and are involved in motion detection, while the two inner PRCs (R7 and R8) detect UV (Rh3 and RH4), blue (Rh5), or green (Rh6) light depending on which rhodopsins they express (Figure 2) and thus define different ommatidial subtypes. [14,15]The evolution of rhodopsin F I G U R E 2 Neuropils and connections in the optic lobes of Drosophila melanogaster.(A) The four neuropils of D. melanogaster showing selected major connections between the retina, lamina, and medulla.(B) Connections between lamina monopolar cells (L1-L5) in adjacent lamina cartridges in D. melanogaster and the parasitoid wasp Megaphragma viggianii. [1]pertoires and their expression underlies differences in spectral sensitivities, color vision, and motion detection within and among insect species [15][16][17][18] (Box 2).For example, it was recently shown that sexlinkage of UVRh1 underlies differential UV sensitivity between male and female Heliconius butterflies. [19]matidial subtypes can be found randomly distributed across the eyes of insects, for example, the "pale" (R7 expressing RH3, R8 expressing RH5) and "yellow" (R7 expressing RH4, R8 expressing RH6) ommatidia subtypes of Drosophila, or more regionalized and even in stripes or bands (reviewed in ref. [15]).Similar organizational principles are evident across insects, however, including the specialized dorsal rim area (DRA) and there are potentially conserved aspects of the regulation of rhodopsin expression and the specification and distribution of ommatidial subtypes. [15,20,21]gnals from the PRCs are transmitted to dedicated inter-connected neuropils of the optic lobe (lamina, medulla, lobula, and lobula plate) Shinomiya et al., 2022. [83] reconstruction of the connectivity of neural circuits for computation of visual motion in the lobula plate of Drosophila, which exemplifies how connectomics can help understand the processing of visual information in the optic lobes.
Song and Lee, 2018. [61]erview of insect color vision from spectral sensitivity of photoreceptors and processing of chromatic information to ecology and behavior.
Warrant, 2017. [23]view of the visual adaptations of nocturnal insects.
Warrant and Somanathan, 2022. [60]mprehensive recent overview of an example of a visual adaptation in insects -the color vision of nocturnal insects.
Warren and Kumar, 2023. [28]cent review discussing the genetic regulation and cellular processes underlying Drosophila eye development focusing on the morphogenetic furrow and how this differentiation wave triggers formation of ommatidia.
Wernet et al., 2015. [15]mprehensively reviews the regulation of photoreceptor specification and rhodopsin expression in the Drosophila retinal mosaic and in the eyes of other insects.
(Figure 2; Box 1) that harbor complex networks of tens of thousands of neurons that process and transmit this information to the central brain to regulate behavior.In Drosophila, the outer PRCs from each ommatidium send axons to the lamina whereas inner PRCs project to the medulla.The signals are then passed on to the higher-level processing centers-the lobula and lobula plate-for color and polarized light vision and/or motion detection respectively (reviewed in ref. [14]) (Figure 2).Most insects have apposition eyes (Box 1) where ommatidia are insulated from each other by pigment and the rhabdomeres of the PRCs are fused (fused rhabdom, Box 1).Information from a single ommatidium is transmitted via an axon "bundle" to the same laminal cartridge in the optic lobe [22] (Figures 1 and 2).This arrangement maximizes the spatial resolution that can be achieved by large numbers of ommatidia in diurnal insects. [11,22]In contrast, many nocturnal insects have superposition eyes (Box 1), where a "clear zone" between the lenses and the rhabdoms allows a single PRC to detect photons collected by multiple neighboring lenses. [23]This arrangement maximizes sensitivity in dim light at the cost of resolution. [23]pterans have evolved neural superposition eyes (Box 1), which are apposition eyes with separated ommatidia but open rhabdoms (Figure 1; Box 1), where axons from PRCs of different ommatidia but overlapping fields of view project to the same laminal cartridge (Box 1), while individual PRCs within an ommatidium with different fields of view project to different cartridges (Figure 2).This arrangement is thought to enhance sensitivity while also maximizing resolution. [22]cent modeling and analysis of morphological data of insect neural superposition and fused rhabdom apposition eye have shown that the competing costs of photoreceptors and optics greatly influence eye design and evolution, and visual ability, with investment in the former often exceeding the latter [9] (Box 2).However, detailed comparative analysis of eye structure and visual performance are needed across different regions of the eyes of a wider range of species to better understand their evolution and how this is influenced by costs and other potential constraints. [9]

Development of insect compound eyes
A major challenge in understanding the evolution of many visual system adaptations is studying the mechanisms underlying their development.While field-collected adult insects can be used directly in experiments to study vision, characterizing the development of their eyes requires analysis of earlier developmental stages that may be inaccessible in the wild and/or currently difficult, if not impossible, to culture in the laboratory.However, great progress has been made in establishing exciting new models for the evolution of eye development and vision, including mayflies, where the males have an extra pair of turbinate eyes used to identify females during flight [24,25] (Box 2), and the diving beetle Thermonectus marmoratus, which has evolved superposition eyes and thus increased light sensitivity. [26]ile all insect compound eyes likely have the same evolutionary origin, [27] their basic development differs among lineages.Drosophila eyes develop from a pair of internalized eye-antennal imaginal discs that grow through the first two larval instars before differentiating during the third instar and then fusing during metamorphosis in the pupal stage [28][29][30] (Box 2).Although this process was already well-understood, [28,31] Navarro et al. (2024) [32] recently provided important new insights into how cell division, differentiation, and cell death are coordinated in the developing eye disc of Drosophila to specify cell number and eye size.
In contrast to Drosophila, most other insects do not have internalized eye imaginal discs and instead their eyes develop more directly from eye primordia in the larval head.In the beetle Tribolium castaneum, for example, the eyes develop from epithelia that are part of the lateral head of the larva. [33,34]Similarly in the honeybee, Apis mellifera, the eyes develop from placode-like epithelia visible from the third instar on either side of the head, flanking the optic lobe primordia, and ommatidia differentiate from pre-ommatidial clusters during the fifth instar. [35,36]Similar to Drosophila, the differentiation of the ommatidia in both Tribolium and Apis proceeds from posterior to anterior, suggesting a shared ancestral mechanism for retinal differentiation [36,37] (Box 2).However, further systematic analysis of the cellular processes that build the eyes of holometabolous insects like Tribolium and Apis as well as a much wider range of insects, including direct developers, is sorely needed to understand the development and evolution of insect eyes more generally.
Establishing the genetic regulation of compound eye development and how this is conserved or divergent across insect species is crucial to understanding the evolution of insect eye diversity.The eye gene regulatory network (GRN) is very well understood in Drosophila: eyes are specified by the Pax6 genes eyeless (ey) and twin of eyeless, which act upstream of other retinal determination genes including sine oculus and eyes absent [28,30,31,38,39] (Box 2).][42] (Box 2).Therefore there is likely very much to be learned from studying the genetics underpining the regulation of eye development in a wider range of insects.

DECIPHERING THE GENETIC BASIS OF INSECT EYE MORPHOLOGY EVOLUTION
The great diversity of eye size and structural composition among insects begs the question of what are the genetic and developmental bases of these evolutionary changes?45][46] Ramaekers et al. (2019) [47] showed that a single nucleotide change in a binding site for the transcription factor Cut in an enhancer of the Drosophila Pax6 gene eyeless (ey) underlies ommatidia number and consequently eye size differences between strains of D. melanogaster and this mechanism likely contributes to the eye size difference between D. melanogaster and D. pseudoobscura. [47]Furthermore, differences in ommatidia number between D. melanogaster and D. simulans are also associated with sequence changes in this Cut binding site in the enhancer of ey. [3]This suggests that changes in ey regulation could be an evolutionary hotspot, at least among Drosophila species, and it would be interesting to test whether this regulatory mechanism is involved in eye size differences in other insects.However, differences in Drosophila ommatidia number have been found to be polygenic [3,48,49].Consistent with this, the nucleotide change in ey alone does not explain the full ommatidia number difference in the focal strains studied by Ramaekers et al. ( 2019), [47] and this nucleotide is the same in other strains of D. melanogaster and D. simulans with intra-specific variation in ommatidium number. [3]Interestingly these intra-specific differences in eye size also appear to be caused by different developmental mechanisms including relative timing and rates of cell division versus differentiation. [3]In addition, Buchberger et al. (2021) showed that the gene pannier acts after Cut regulation of ey during sub-division of the eye-antennal disc to generate differences in ommatidia number and overall head shape between D. melanogaster and D. mauritiana. [50]ch less is understood about the development of ommatidia diameter differences within eyes, even in Drosophila (but see [51] ), never mind how this critical aspect of eye morphology evolves between species.It was recently shown that the wider ommatidia of D. mauritiana compared to its sibling species D. simulans is associated with changes in the expression of the transcription factor Orthodenticle (Otd) as the ommatidia mature. [4,44,52]Otd is a highly conserved transcription factor in insect eye development [41] and uncovering the downstream mechanism in these Drosophila species could help understand the regulation and evolution of ommatidia size among insects more broadly.
The studies above demonstrate the enormous potential to decipher the genetic basis of differences in eye size and structure among other Drosophila species [43,45,48] and this could help to pinpoint the nodes of evolution to better understand how GRNs evolve more generally.While much less is known in other insects, models generated for eye development such as beetles, bees, and butterflies are already demonstrating their great potential to identify and investigate candidate genes to test mechanisms from Drosophila and more fully explain diversification of eye morphology among insects. [26]compared the function of the transcription factor Cut between D. melanogaster with neural superposition eyes and the diving beetle T. marmoratus with optical superposition eyes using RNAi knockdown.They found that cut likely plays similar roles in ommatidium formation in these two insects including regulating development of the cone (Semper) cells, important support cells that secrete the lens. [26]This evidences the importance of cone cells and the conserved role of Cut in regulating ommatidia formation among insects with different optics.

Rathore et al. (2023)
Recently, Netschitailo et al. (2023) [53] revealed how the sex determination pathway regulates eye size dimorphism in honeybees.[56] They identified glubschauge (glu) as a sex-specifically spliced transcription factor gene with only the female protein isoform (gluF) containing a zinc finger DNA-binding domain.Knockout of glu in females using CRISPR/Cas9 increased eye size, while overexpression of gluF in males reduced ommatidia size and overall eye size.Identifying the regulatory mechanisms of glu as well as the target genes of GluF offers an exciting opportunity to understand the role of the sex determination pathway and alternative splicing in generating different eye morphologies in male and female bees.Intriguingly, glu orthologs are found in other insects but are not alternatively spliced and lack the zinc finger domain. [53]Therefore, this could represent a novel mechanism specific to bees.However, sexual dimorphism in eye size is pervasive in insects, including Drosophila.Which components of the sex determination pathway are involved in differences in eye morphology in other insects and if they act through similar or distinctive developmental mechanisms and downstream targets in eye development GRNs remains an open and interesting question to pursue.

THE FUNCTIONAL IMPACT OF VARIATION IN EYE MORPHOLOGY ON VISION
The astonishing behavior of many insects is only possible because of their underlying visual capabilities.While this has been reviewed thoroughly elsewhere [11,22,[57][58][59][60][61][62] (Box 2) the following recent examples serve to illustrate this point.
The remarkable flight acrobatics of dragonflies rely on the amazing visual systems these aerial predators have evolved (reviewed in ref. [63]).The eyes of dragonflies are fused dorsally while those of the related damselflies are separated with binocular overlap.This difference in the eye arrangements as well as the position of targetselective descending neurons reflects the different hunting strategies of these two predators: dragonflies target their prey dorsally while damselflies approach from the front using their binocular vision. [64]e visual systems of many insects have also evolved to facilitate nocturnal behaviors including superposition eyes (see above).Indeed, some insects such as giant hawkmoths (reviewed in ref. [60]) and the Asian giant honeybee Apis dorsata [65] even have color vision at night.The broad patterns of the evolution of the major insect eye types and how they impact vision and may underlie adaptations are reasonably well understood.However, it is often difficult to identify the underlying mechanisms and evolutionary drivers across deep macroevolutionary timescales.Therefore, a better understanding of the impact of more recent evolutionary differences in eye morphology on vision within species and between closely related species is needed to both decipher proximal events and inform understanding of more ancient differences. [66]Several recent studies have shown that this can be a powerful approach to understand eye structure-function links and in some cases even why differences may have been selected during the course of evolution.
We recently modeled and tested differences in vision between D. mauritiana and D. simulans resulting from the wider ommatidia of the former species. [5]As predicted, the behavioral experiments showed that D. mauritiana eyes have greater contrast sensitivity while those of D. simulans have enhanced acuity.While D. simulans is a cosmopolitan generalist, D. mauritiana is restricted to Mauritius and this finding may help to better understand their ecology and behavior, and why these differences may have been selected.
Analysis of eye size differences between other Drosophila species has provided insights into their behavioral differences.D. subobscura has 25%-30% more ommatidia and larger optic lobes than the sympatric D. pseudoobscura, which has larger antennae and antennal lobes. [45]This is consistent with a trade-off between visual and olfactory sensory systems described more widely in other fly species. [43,47]e mating of D. subobscura is more light-dependent and therefore more visual than that of D. pseudoobscura and the differences in eye size between these two species may be related to niche partitioning, with D. subobscura inhabiting lighter environments than D. pseudoobscura. [45]vestigating the genetic and developmental bases of this difference in eye and optic lobe size between D. subobscura and D. pseudoobscura would likely provide valuable new insights into the evolutionary diversification of Drosophila eyes.
In the future, it would also be interesting to more broadly compare the changes in eye development, structure, and vision among Drosophila species to those among other closely related insects to study whether similar or different mechanisms are employed across lineages.
Heliconius butterflies represent a very promising system to do this.
Heliconius cydno inhabits dense forests whereas the closely related H. melpomene is found around the forest edges in more open habitats. [67] cydno has larger eyes, composed of more ommatidia, and bigger visual neuropil volume compared to H. melpomene and these morphological differences are under directional selection.[68][69][70] The males of both species also have larger eyes than the females.[68,70] These differences are presumably adaptations to the different habitats they occupy and for males to find mates.[68] Interestingly the eyes of the hybrid F1s of these two species are similar to H. melpomene while their neuropils are similar to H. cydno or intermediate between the species.[69] This exciting finding suggests there are differences in the coordination of eye and optic lobe development between these species, which disrupts hybrid visual systems and may affect their fitness.[68,69] This system has great potential to identify the genes underlying these differences in eye morphology and coordination with optic lobe development because these butterflies can be cultured and genetic mapping is possible (e.g.[71] ). Furtherore, the visual abilities of Heliconius species and their F1 hybrids could be investigated to determine the consequences of differences in optical parameters and visual information processing.This would provide a unique connection between genetics, eye development, morphology, and visual behavior and ecology in butterflies.
In Drosophila and other insects, eye size is often linked to variation in body size, which subsequently has implications for the number and size of ommatidia and therefore optic parameters.Recent papers have described striking natural variation in eye size within and between bumblebee species that are predicted to cause differences in their vision based on morphological measurements. [72,73]This plasticity in eye development within bumblebees has been suggested to produce workers with different visual abilities that allows optimal foraging in alternative niches with varying light conditions, [72] therefore facilitating microhabitat niche partitioning within species, as suggested above for partitioning between fly and butterfly species. [73]new study of the hummingbird hawkmoth Macroglossum stellatarum has shown that the superposition eyes of this insect scale hypoallometrically with body size and that smaller hawkmoths have fewer but wider ommatidia. [74]This allows these insects to maintain the balance between acuity and contrast sensitivity despite differences in body size. [74]Future studies focusing on the underlying developmental mechanisms have great potential to better understand eye size plasticity and evolution across insects.

ACTIVE VISION
Compound eye structure alone does not fully explain visual ability and so we need to consider other aspects of eye function and information processing to understand the evolution of visual systems across insects.Most insects cannot move their eyes independently of their head to shift gaze or generate optic flow for spatial information.[77] Several recent studies have provided new insights into the structure and functionality of D. melanogaster eyes that affect their active vision with implications for understanding the quality of visual information that insects can perceive.
It was previously observed in the housefly Musca domestica that the retina was served by two muscles that could move it in response to external cues [78] and thus move PRCs relative to the lens while the head remained still. [78]Fenk et al. (2022) [79] recently described and investigated the function of these muscles, musculus orbito tentoralis and musculus orbito scapalis, in D. melanogaster.Optogenetic activation of the motor neurons attached to these muscles resulted in approximately 15 • of movement of the PRCs.Importantly, this movement did not affect the alignment of PRCs in adjacent ommatidia necessary for neural superposition.The results suggest that these muscles allow the fly to move its retina in two directions to stabilize gaze when tracking a moving environment (similar to human eyes) in concert with head and body movements. [79]found that these retinal movements were used during gap crossing and this ability was perturbed when signaling of retinal motor neurons was blocked.This suggests that the retinal dynamics are involved in a "binocular ruler mechanism" and the authors conclude that these results indicate that the retinal movements allow flies to realign their PRCs to better visualize fine-scale features and may be involved in spatial attention and object recognition. [79]This exciting discovery complements recent further characterization of PRC microsaccades (Box 1) in D. melanogaster. [80]showed that previously observed [81] PRC contractions in response to light (PRC microsaccades, Box 1) sweep across both eyes mirror symmetrically in D. melanogaster.Interestingly, stimulating one PRC in an ommatidium caused all its other PRCs to move but not the PRCs in neighboring ommatidia, which suggests these movements are not caused by the retinal muscles described above.In fact, the findings of Kemppainen et al. (2022) [80] suggest the PRC microsaccades are photomechanical and induced by phototransduction.Moreover, the symmetrical movement suggests that PRCs in the left and right eyes with overlapping frontal fields are both scanning in the stereo-range.Models and simulations predict that this results in better resolution and stereopsis than predicted by classic models. [80]Kemppainen et al. (2022) [80] experimentally demonstrated hyperacute stereopsis in Drosophila using salience and learning experiments and showed that flies can detect 3D objects and 2D shapes at higher resolution than predicted by static morphology.Interestingly, PRC microsaccades were also recently recorded in the fused rhabdom eyes of honeybees implying that these are a conserved feature of insect compound eyes. [82]gether these two recent key studies [79,80] show that the visual ability of Drosophila eyes is more complex than previously understood and by what can be inferred from their static morphology alone.This has profound implications for understanding insect compound eye function and vision.It will be important to identify and characterize retinal movements and PRC microsaccades within and between other insect species to complement studies of their static eye morphology.Variation in these factors is likely to impact visual ability and could help insects adapt to different environments and affect their behavior.

INFORMATION PROCESSING
[85][86][87] ) (Box 2).This has provided unprecedented new insights into color vision, motion detection, movement, and navigation in this fly (Box 2).For example, flies and other insects must be able to adjust to changes in light intensity over different time scales, from rapid differences during flight to gradual changes throughout the day. [88,89]It was recently shown in D. melanogaster that this gain control over different timescales happens downstream of the PRCs by the luminance-sensitive lamina monopolar cell (LMC) L3 [88,89] (Figure 2; Box 1).
These studies in Drosophila have been complemented by the generation of connectomes of other insects, including the lamina of a butterfly Papilio xuthus [90] and the projectome of the central complex of the bumblebee Bombus terrestris. [91]While the acquisition and comparison of these connectomes is a complex task, it is beginning to reveal conserved and divergent aspects of visual processing. [1,90]ua et al. (2023)?? recently reported their reconstruction of the adult eye structure and lamina connectome of the parasitoid wasp Megaphragma viggianii.Females of this remarkable insect only have 29 ommatidia in each eye and their brain only contains about 8600 cells. [1]is study verified differences in ommatidia structure in Megaphragma between the DRA and the rest of the eye with specialization to detect polarized light in the DRA as has been observed in other insects. [15]is specialization was also reflected in the connectivity (synapses received) of PRC R7 versus R7′ (and R8) between non-DRA and DRA, which correlated with variance in the orientation of microvilli in these regions. [1]Cs are key conserved laminal cell types in arthropod visual systems, which receive synapses from the PRCs and transmit information to the medulla (Figure 2; Box 1).While other hymenopterans, including Apis, likely only have four distinctive LMCs, [1] Megaphragma has five (L1-L4, LN) like Drosophila.Interestingly, four of the Megaphragma LMCs have lost their nucleus presumably as part of miniaturization. [1]alysis of the Megaphragma connectome showed that R1-R6 inputs to the LMCs are mostly to L1-L3 consistent with the Drosophila and Papilio laminae (Figure 2), which suggests a similar function in detecting contrast changes in this insect. [1,90]LN in Megaphragma receives almost no R1-R6 synapses, which is again similar to L5 in Drosophila and Papilio. [90]Furthermore, in Drosophila and Megaphragma one LMC (L2 in Megaphragma) forms a large number of connections to other LMCs, for example, between L2 and L4, similar to Drosophila and Apis (Figure 2).Overall, LMC connections are generally similar between Megaphragma, Drosophila, and Papilio suggesting this is the ground plan for insects.However, in Drosophila L4 has both pre-and post-synaptic connections to the cartridge of the posterior neighbor aligned with optic flow during flight, although these connections do not appear to be involved in the detection of motion [92][93][94][95] (Figure 2B).In Megaphragma, the connections from L2 (in posterior cartridges) to L4 (in anterior cartridges) are only post-synaptic and opposite to optic flow (Figure 2). [1]e authors speculate that this may be associated with differences in the role of L4 between these insects. [1]e Megaphragma lamina shows additional differences to other insects.For example, Megaphragma has no neurons homologous to Lawf, C2 and C3 neurons in Drosophila, which connect with the medulla. [1,95]Megaphragma also has far fewer lamina inter-cartridge connections compared to the lamina of Papilio. [1,90]The similarities and differences between the laminal connectome of Megaphragma to other insects show both the essential and superfluous laminal cells and connections retained and discarded respectively during miniaturization of this micro-wasp.
Taken together, these comparisons of insect connectomes also have broader implications for understanding the evolution of the neural wiring that encodes and processes visual information in these animals.
First, they reveal the basic templates of cells and connections conserved across the visual processing systems of insects.Second, they highlight lineage-specific features of insect visual systems that are starting to help better understand their evolution and how this gives rise to different visual preferences and behaviors.
A complementary approach to connectomics is to genetically map natural differences in visual preferences and ask how these genes are integrated into visual processing.Rossi et al. (2024) [96] recently used such an approach to map the genetic basis for mate color preference among species of Heliconius butterflies.H. melpomene and H. cydno have red and white colored bands respectively on their forewings.[99] H. melpomene and H. timarata males also both show mating preference for females with red patterns.The new work by Rossi et al. (2024) [96] used mapping and RNA-seq to show that H. melpomene and H. timarata preference for red patterns is associated with lower expression of regucalcin1 compared to H. cydno, and that CRISPR/Cas9 mutation of this gene perturbs courtship.The identification of this role of regucalcin1 in mating preference among Heliconius butterfly species now offers an excellent opportunity to better understand how visual preferences are wired into the nervous system and how it evolves.Specifically, it will be interesting to understand the function and interactions of regucalcin1 in the brain and where and how differences in the expression of this gene alter visual processing and color preference.

CONCLUSIONS
Our review of recent research demonstrates that the gap is closing between our detailed genetic and developmental understanding of Drosophila eyes and insights of visual behavior and ecology in other insects.Comparative studies linking the evolution, development, and function of visual systems across a wide range of insect species are required to gain a full understanding of the diversification of insect eyes.We propose that this would be greatly aided in the future by research focusing on four main areas.
First, the application of single cell transcriptomics in Drosophila has verified many of the known regulators and cell types but also identified new genes and cell markers [100][101][102][103] as well as detailed new insights into the regulation and function of the Drosophila visual system more broadly [104,105] (Box 2).08] Applying these technologies to the developing eyes and visual systems of other insects, as recently carried out for honeybees, [109] will provide a framework to characterize how these genetic programmes evolve to produce visual systems with different morphologies, information processing, and functions.Second, overcoming challenges to culturing other insects in the lab is crucial to understand, compare, and manipulate eye development and has recently successfully been carried out for mayflies for the first time [24] (Box 2).This would facilitate RNA-seq approaches and CRISPR/Cas9, which are broadly applicable in insects to analyze gene expression and function.Third, the growing number of insect visual system connectomes has greatly advanced our understanding of the detection and processing of visual cues and how this underlies behaviors.However, higher throughput connectomics is needed to better understand the function and evolution of visual systems within and between species.Finally, even where eye development and structure are well understood, especially within and among Drosophila species, a greater focus on the consequences for vision and how this underpins behavioral and ecological differences as exemplified in recent studies [45,68] is needed in future research.

F I G U R E 1
Compound eyes in holometabolous insects.Examples of the head and compound eyes of representatives of four major orders of holometabolous insects (top).Schematic of an ommatidium with cross section views of open and closed rhabdom configurations illustrated (bottom left).Schematic of ommatidia indicating field of view (FOV), inter-ommatidial angle (Δϕ), and ommatidium diameter (D) (bottom right).Pictures were kindly provided by Markus Friedrich (T.castaneum), Kentaro Arikawa and Michiyo Kinoshita (P.xuthus), Vivek Nityananda (B.terrestris), and Javier Figueras Jimenez (D. melanogaster).