A taxonomic revision of the Gold Coral genus Chrysogorgia (Coelenterata: Octocorallia: Alcyonacea: Chrysogorgiidae) using a combined morphological and molecular approach, and alternative methods for assessing diversity of deep-sea octocorals in underwater imagery

My PhD research program has focussed on a prominent and ecologically important group of sessile marine invertebrates, the octocorals, colloquially referred to as soft corals or sea fans. Octocorals are among the most abundant and diverse components of deep-sea ecosystems where many are large ‚ÄövÑv=tree-forming‚ÄövÑv¥ species that form habitat and refugia for other marine fauna. Because octocorals typically grow slowly and are long-lived, they are vulnerable to impacts from human activities such as bottom fishing and seabed mining. Accordingly, octocorals are focal taxa for the conservation management of deep-sea marine environments globally and are a prominent indicator taxon of Vulnerable Marine Ecosystems (VMEs). VMEs are frequently used as the spatial units on which management of threatening activities in the deep-sea is based. My research comprised two primary topics: an integrative taxonomic investigation within a species-rich but poorly documented genus of octocorals, Chrysogorgia Duchassaing & Michelotti, 1864, using morphological taxonomy and molecular methods. An improved knowledge of the taxonomy of Chrysogorgia is important because the genus is distributed globally, in a wide range of depths and is among the most speciose of all soft corals ‚ÄövÑv¨ including species established as ecologically important and VME indicator taxa. The genus required comprehensive taxonomic revision because its species displayed such a widely varying range of morphological characters it was suspected to be polyphyletic. Molecular methods have become a valuable tool in taxonomy, providing additional characters which can guide distinction between morphologically similar taxa. However, while previous analyses of Chrysogorgiidae using mtMutS, 18S and mt COI markers showed that Chrysogorgia appears monophyletic, and also showed the congruence of genetic variation in mtMutS with discrete morphologies for four Atlantic species, in each case only a small number of taxa (<50), some with unknown morphology, were surveyed. The need for a comprehensive study of DNA sequencing over a much broader range of forms, along with detailed morphological analysis, has been identified. The second topic was an evaluation of the limitations of identifying octocorals in imagery and the implications of this for a range of ecological assessments. I analysed the assemblage composition of octocorals in a large number of underwater deep-sea images and compared them with specimens from collections of octocorals to assess what level of taxonomic identification can be consistently achieved in image-derived data. This knowledge informs scientific field sampling of benthic biodiversity that has an increasing reliance on nonextractive image-based methods to address ethical concerns about sampling impacts in sensitive habitats. Species of Chrysogorgia, commonly known as ‚ÄövÑv¿gold corals‚ÄövÑvp, have historically been viewed as some of the most beautiful and intricate of all gorgonians, with colonies typified by their delicate spiralling, zig-zagged golden axis and, commonly, a bottlebrush shape. This genus has been considered among the most speciose of all alcyonaceans (81 nominal species), with representatives found worldwide, occurring on both hard and soft bottoms, at depths of 31‚ÄövÑv¨ 4327 m. But, while virtually all nominal species of Chrysogorgia share many similarities in gross morphology, they also exhibit wide variations in microscopic features related to sclerite shape, ornamentation and arrangement, especially on the polyps. Consequently, the genus was suspected of being polyphyletic (Alderslade pers. comm.). There has been increased interest in the taxonomy of Chrysogorgia in the last two decades, partly due to an increase in deep-sea exploration, and several new species have been recently described. However, a comprehensive revision and critical examination of morphological characters in nominal species of Chrysogorgia has not been conducted since 1902. The lack of adequate and comprehensive taxonomic research to establish the reliability of morphological characters used to delineate species is a significant impediment to our understanding of diversity within this genus, and to consistently interpret the group in respect to VMEs. I evaluated and reviewed the taxonomy of Chrysogorgia with a two-pronged approach, employing a leading-edge molecular technique (Chapter 2) to inform and guide a morphological taxonomic revision and the establishment of new genera (Chapter 3). The material and data for both research components was acquired from first-hand examination of a large set (>600) of Chrysogorgia s.l. specimens housed in museums and institutes around the world (Smithsonian: National Museum of Natural History; Museum of Comparative Zoology at Harvard; Yale Peabody Museum; Iziko South African Museum; Natural History Museum, London; Natural History Museum of Denmark; Museum Nacional, Brazil; The National Institute of Water and Atmospheric Research, New Zealand; The Commonwealth Scientific and Industrial Research Organisation, Australia; Harvey Mudd College, USA). Initially, I identified nine groups, based on a set of broad, morphological characters that were hypothesized to represent distinct lineages. I applied a recently developed universal target enrichment bait method for octocoral exons and ultra-conserved elements (UCEs) on 96 specimens varying in morphology, collection ages and DNA quality and quantity to determine whether there was genetic support for morphologically defined groups within Chrysogorgia s.l. and to test the utility of this technique for analysis of museum octocoral specimens. Target-capture enrichment of UCEs and exons was pioneered for anthozoans by Quattrini et al., (2017), subsequently refined and used to investigate species-level relationships in the octocorals Alcyonium (Linnaeus, 1775) and Sinularia (May 1898), demonstrating that this technique has greater utility for resolving phylogenetic relationships and species boundaries in octocorals than existing single and multi-locus markers. I found strong concordance between assigned morphological groups and highly-supported clades recovered from the UCE+exon phylogeny, in all but one case. This result has important implications for Chrysogorgia taxonomy: the phylogenetic results forming the basis to propose that 12 candidate genera, with specific morphologies, should be recognized. In addition, 53 of the 96 specimens could not be assigned to an existing species, clearly demonstrating the diversity and morphological complexity within Chrysogorgia s.l. This highlights the value of integrative research using molecular phylogenetic analysis to guide taxonomy, and to understand the homoplasious nature of some morphological characters. My results also demonstrated, for the first time, that this target-enrichment approach can be successfully applied to degraded museum specimens of up to 60 years old. The morphological taxonomic review of Chrysogorgia (Chapter 3), based on the examination of established characters, identified a considerably more complex structure than had been previously envisaged, and one that I was able to satisfactorily reconcile with my molecular data. The morphological and molecular research together made possible a major review and reorganisation of Chrysogorgia s.l., splitting it into 12 genera (Chrysogorgia s.s., Dasygorgia, Verrillogorgia, Spiculogorgia, Kinoshogorgia, Squamogorgia, Nudogorgia, Rehbockogorgia, Robertogorgia, Scoleriogorgia, Aldersladeogorgia and Dentogorgia). Of these, 11 had been erroneously incorporated within Chrysogorgia s.l.: one was a synonym (Dasygorgia) and 10 are new and described here (Chapter 3). Importantly I found that the characters of the type species, Chrysogorgia desbonni, provided by Verrill (1883) do not conform to those that can be deduced from the original description (Duchassaing and Michelotti 1864). In addition, this species has been incorrectly synonymised with C. occidentalis, I have therefore re-established occidentalis as a species under a new genus, Verrillogorgia and re-described Chrysogorgia desbonni, based on what can be established from literature. The second major focus of my research was to determine the highest resolution at which octocorals could be confidently and consistently identified (annotated) in imagery ‚ÄövÑvÆ as ‚ÄövÑv¿phototaxa‚ÄövÑvp ‚ÄövÑvÆ and compare my classification with alternative schemes (Chapter 4). Many alternatives exist because individual studies typically use a set of region-specific phototaxa/morphospecies as proxies for species. It is difficult to compare or combine data across studies or regions because schemes are rarely standardised. My comparison evaluated the trade-offs (e.g. complexity and resource requirements) of different schema against the needs of different ecological applications. These included the standardised ‚ÄövÑv=Collaborative and Automated Tools for Analysis of Marine Imagery‚ÄövÑv¥ (CATAMI) scheme based on morphotypes to classify faunal units that is widely used in Australian studies, and increasingly used internationally. The implementation of this comparison required me to update the CATAMI scheme for octocorals, and in so doing I was able to build a new ‚ÄövÑv=combined‚ÄövÑv¥ annotation methodology based on both phototaxa from taxonomic schema and morphological entities from the CATAMI schema. I generated my project dataset by annotating the octocorals seen in 5,939 high-resolution still images and cross-referenced these data with physical specimens collected during surveys of seamounts off Tasmania, Australia ‚ÄövÑvÆ including a month-long survey in 2018 on which I was present. I found coral diversity was greatly underestimated when assessed in imagery: only 55 black and octocoral ‚ÄövÑv=phototaxa‚ÄövÑv¥ (best-possible identifications) were consistently distinguishable from the 210 species known in the region. This is because it is impossible ...