Taxonomy and phylogeny of Australian cubozoa

Jellyfishes in the class Cubozoa are species rich and often abundant in Australian waters. They are geographically widespread in tropical and temperate waters and they have global significance both economically and recreationally as dangerous marine stingers. They are interesting evolutionarily and with respect to ecology and life history. Despite this, the taxonomy of cubozoans is too coarse to allow discrimination of closely related species, hindering further advances in all aspects of cubozoan biology. The objectives of this thesis were to revise the taxonomy of the Cubozoa based on structural characters, and to elucidate the evolutionary relationships of cubozoan species based on qualitative comparison of morphological and molecular phylogenetic analyses. I present a detailed historical and contemporary review of 85 morphological characters, many of which have not been previously used. These include (where possible) nematocysts and statoliths (balance stones), that allow the identification of ethanol-preserved specimens, frozen-defrosted material, fragmented or badly damaged samples, and possibly even fossil species. Additional characters that give increased taxonomic resolution include apical decorations, pedalial keel ratios and armament, pedalial canal shape and branching, tentacle decorations and banding forms, phacellae branching and cirri length, rhopalial horns and windows, number of eyes, frenulae, perradial lappets, velarial armament, lips shape, and a new approach to interpreting mesenteries. Accurate identification of cubozoans is based on many morphological characters. There is no small set of characters that can be universally compared to identify taxa with high reliability, but rather, different sets of characters are reliable for different groups and at different levels. For example, the historical split of the chirodropids (with gastric saccules) from the carybdeids (without gastric saccules) is no longer accurate; the undescribed spotted chirodropid (Chirodropus sp. A) lacks gastric saccules. Similarly, rhopaliar niche ostium shape and direction of phacellae work well for separating many (but not all) families of carybdeids, but are uniform in the chirodropids. The synoptic identification tools presented in this thesis will allow for reasonably reliable identification for the species herein, being mindful of preservational distortions, ontogenetic character changes, biological variation, and unrealized species. I recommend the use of the full range of characters presented in this study for identification and recognition of new species and species outside Australian waters. Phylogenetic relationships within the Cubozoa were inferred by comparing parsimony analysis of 31 species scored for 85 morphological characters against Bayesian maximum likelihood analysis of partial 18SrDNA sequences from 42 individuals representing 13-16 species. Numerous patterns are congruent and well supported in both data sets as follows: separation of the "Carybdea alata" species complex from the other Carybdea spp., a grouping of Carybdea sivickisi with Tripedalia spp., and monophyly of the Chirodropida. Furthermore, there were three distinct groups of highly toxic jellyfish whose stings result in Irukandji syndrome; although differences exist between the morphological and molecular tree topologies, there was nonetheless strong support for a Glade herein referred to by the non-taxonomic common designation "Irukandjiidae". Based on the combination of morphological and molecular phylogenetic analyses, numerous changes to the existing taxonomic framework were indicated. A revised classification is proposed, along with synopses of the species and a dichotomous key to taxa collected in Australian waters. Furthermore, a new family is proposed, the Alatinidae, with detailed descriptions of a new genus, Alatina, and two new species, A. mordens and A. rainensis. Other new taxa are indicated throughout the text, but will be fully treated in a monographic revision of the Cubozoa generated from this work. Practical application of these results has already begun. The Irukandji Glade identified in this study contains at least two assemblages of medical interest, the Carukia spp. and the "Pseudo-Irukandji" group. Species from each of these sub-clades have been associated with Irukandji syndrome, the latter linked with a fatal sting event. These two groups are further sorted on numerous macro-morphological features, cnidomes, statoliths, behavioural patterns, and spatio-temporal distribution, and there is some indication that syndrome severity may sort along phylogenetic lines. The link between these species and symptoms remains to be conclusively shown, but the correlative evidence suggests it should be an active area of research. This study covers new ground in many respects, including detailed examination of a wide range of morphological characters and production of comparable robust phylogenies from molecular and morphological data sets. A sound taxonomy is required as the basis for communication and comparison in all other types of cubozoan studies, such as ecology, toxinology, and basic biology, all of which will, in turn, be necessary for the successful management of Australia's jellyfish problem