Visual Evolution of Scincoidea

Abstract

Several clades are now known to have undergone nocturnal bottlenecks, changing their eye morphology, photoreceptor morphology and opsin complement. Within Squamata, snakes and geckos are the most widely studied clades that are known to have undergone nocturnal bottlenecks. The visual evolution of representatives of most major clades within Squamata have been examined, yet Scincoidea are a particularly understudied clade, considering their species richness, cosmopolitan nature and diversity in body form and habitat. With this diverse clade practically unexplored in the field of visual evolution, I investigated the visual and non-visual opsins of Scincoidea. Within Scincimorpha there are several clades that have transitioned to a fossorial lifestyle, and exhibit morphological adaptations consequently. Of particular interest are Lerista, a large and diverse clade with a spectrum of adaptation across its species. There have been several independent digit and limb loss events in Lerista, and the full range of limb configurations is seen within, from pentadactyl to limbless. This contrasts with Ctenotus, its sister clade, whose species are strictly pentadactyl. The transitions to low light in this clade are an excellent system in which to examine the effect of recent scotopic transitions on the visual opsins. I sequenced the visual opsins of 86 Lerista and Ctenotus using gene capture, finding all five ancestral visual opsins in every species. Selection tests were then conducted on each opsin to see if selection correlated with any morphological indicator of low light adaptation. We found several selected sites, though the relevance of these in terms of spectral tuning is as yet unknown. We also found that LWS is under relaxed selection in Lerista that have experienced limb and digit loss, and that there are site changes in RH1 that may have implications for the structural stability of the opsin. Thus we found that colour vision retains importance for fossorial skinks, and that eye size reduction in fossorial species may only reduce acuity, and reduce the maximum distance at which they can focus, which we hypothesise is of lower importance to skinks that spend their time manoeuvering through sand, leaf litter and interstitial spaces. I then expanded our focus to all of Scincoidea, with a view to comparing their nonvisual opsin complement to other clades of squamate. Snakes, geckos and mammals have all undergone nocturnal bottlenecks, and in addition to visual opsin loss, they are now known to have lost several non-visual opsins. The implications of these losses are not yet fully understood, but many of the lost genes have roles in circadian rhythm regulation, photoentrainment, sensitivity to polarised light and skin pigmentation changes. There have recently been studies that suggest skinks have had a non-diurnal ancestor, but the retinal morphology of the few skinks studied does not provide clear evidence of a dim-light ancestry in snakes and geckos. I took non-visual opsins from publicly available genomes and transcriptomes, and adding to those my gene capture data and skink transcriptomes we synthesised, we compiled a database of presence and absence of 15 non-visual opsins across 70 species of squamate. Supplementing our skink dataset with sequences from the Tiliqua rugosa genome, we were able to infer that Scincoidea as a clade has lost at least three, and possibly as many as six non-visual opsins, many of which are convergently lost in snakes and geckos. These results provide the first evidence of non-visual opsin loss in Scincoidea, support previous hypothesis of a mesopic bottleneck in Scincoidea, and have also allowed us to infer that the pineal gland of skinks, as in snakes and mammals, may not be photosensitive. I also uncovered previously unknown opsin losses in diverse squamate clades Gymnopthalmoidea and Acrodonta, related to parietal eye and pineal gland function, respectively. Our final investigation built on our previous one, as we then tested selection on nonvisual opsins across Squamata. Taking the same dataset as in Chapter 3, sampling across Squamata, various selection tests uncovered that snakes and geckos showed signals of relaxed selection in most non-visual opsins, compared to other squamate clades such as lacertids and anguimorphs. Skinks however showed enhanced constraint in five opsins, OPN5, OPN3, OPN4m, OPN4x and RRH. This is in stark contrast to snakes and geckos, and differs even from findings in lacertids and anguimorphs. From this I inferred that skinks, having experienced a mesopic bottleneck and lost several non-visual opsins, as I found previously, had then emerged and readopted diurnality, increasing selective pressures on the remaining non-visual opsins. Parapinopsin and parietopsin, parietal eye genes, have been maintained in skinks, while pinopsin, the ancestral pineal photopigment, has been lost. Based on this I suggest that the parietal eye opsins have assumed the role of melatonin regulation from this inactivated opsin. This thesis provides deep insight into the adaptation of visual systems to ecological transitions, from the lateral visual pigments within a clade of Scincoidea, to the extra retinal photopigments across Squamata. It also represents the first molecular evidence of non-visual opsin loss due to a mesopic bottleneck in Scincoidea, and purifying selection pressures resulting from their hypothesised re-adoption of diurnality. It provides evidence of a loss of function in an important circadian rhythm mediator, and co-option of existing opsins to compensate for the loss of others. Finally it demonstrates the diversity of photic adaptation in all Squamata, and particularly in the previously overlooked Scincoidea, which I show to be a rich system for understanding the evolution of visual and non-visual photoreception.