This thesis investigates aspects of the ecology, morphology, bioacoustics, genetic relatedness, biogeography and taxonomy of cicadas in the Pauropsalta annulata Goding and Froggatt species complex, across their entire eastern Australian geographical distribution. These cicadas seem to emerge at the same sites annually, although the exact duration of their life cycle is not known. They are wary and mobile insects, with individuals generally taking flight at the first sign of disturbance, which makes them challenging to record, capture and observe. The males produce distinctive calling songs that have a consistent rhythmic structure, which is critically important for attracting females. In most cases, the songs are biphasic, with a penetrating introductory or "buzzing" component that apparently functions in mate attraction and a strongly structured cueing or "lilting" component for mate localisation. Initially Pauropsalta annulata was thought to comprise a single species with some degree of geographically structured variation in its calling song. Consequently, various "song types" had been recognised, but their precise relationship to one another had never been investigated. Therefore the structure of their calling songs were compared statistically across individuals of three song types, and this revealed four discrete clusters that were demonstrated to be independent of one another and consistent in their calling song structure. Plotting the geographical distributions of these cicadas revealed that each of the P. annulata song types is independently distributed geographically, with areas of overlap that are relatively small. Calling song structure is consistent for each song type across extensive geographical space and this structure holds even into areas of sympatry. One song type showed consistent differences from the others in male genitalia structure, and female ovipositor length differs significantly among three of the other song types. Each song type was found to be strongly associated with a small number of tree species and these associations are maintained into areas of sympatry. The spatial ecological and morphological comparisons in song structure, plant associations and morphology made in this study demonstrate that the P. annulata song types actually represent a complex of cryptic species. Two of the song types appear to intergrade in areas of sympatric overlap in terms of calling songs, plant associations and morphology, and thus represent subspecies as defined in population genetics terms. The components of the male calling song were then investigated to determine their specific functions and thus how they could delimit species boundaries. Significant differences were found in dominant song frequency between three of the four species, and may contribute to differential mate attraction. Both components of the song of each species have the same dominant frequency. Consistent differences were also evident, among species and subspecies, in the rhythmic structure of the "lilting" component of the male calling song. This component contains repeated phrases and each one of these provides a cue to which the conspecific female may respond. Her response is timed for the brief silent interval between the phrases. At this point the calling song becomes a duet, which enables the male to locate the female, as he actively searches for her on the surrounding branches at this stage. The calling song is discussed in terms of random mating within gene pools of these cicadas and in terms of its role as part of their broader fertilisation mechanism. A molecular analysis of the P. annulata species complex was performed to examine the phylogenetic relationships across 12 species and four subspecies defined in this study, and estimate divergence times within the group. Individual specimens were sampled widely across the geographical distributions of the species and subspecies where possible to account for genetic variation across space. DNA sequences from two loci were amplified: mitochondrial CO1 ("barcoding region") and a large intron from the dynamin nuclear gene. Separate phylogenies were reconstructed for each locus using maximum parsimony procedures and Bayesian posterior sampling with implementation of a relaxed molecular clock. The phylogenies from both genes provided strong support for the monophyly of the P. annulata species complex, and nine of the species were monophyletic based on the CO1 gene. The remaining three emerged non-monophyletic. Based on a clock calibration of 0.0165s/s/myr, the monophyletic clades represented by extant P. annulata species diverged about 4.5-8.0 million years ago. Those species that emerged non-monophyletic had shallower divergences, with the exception of one species, which exhibited haplotype diversity that conferred up to 13.2% sequence divergence between allopatric populations in CO1. Dynamin produced a broadly similar phylogenetic pattern to that of CO1, but the relationships among individuals across the species and subspecies that emerged non-monophyletic differed substantially. This lack of congruence between the two genomes, in combination with the dominance of internal haplotypes in both loci, indicates an overall pattern of deep coalescence rather than interspecific hybridization. Therefore the molecular data do not provide an alternative definition of species limits in the P. annulata species complex, despite some emerging non-monophyletic in this analysis. Sound interpretation of the phylogenetic pattern discovered here would not have been possible without the acoustic, ecological and geographical investigations on species limits that preceded this work. To determine what biological and climatic factors influence the present day distributions of these cicadas, the distributions of two of the most closely related species in the P. annulata species complex were compared with the distributions of the tree species with which each is mainly associated. Because a large part of their life cycle is subterranean, soil texture, pH, electrical conductivity and force required for surface penetration were compared across sites where each of the cicadas occur in sympatry and allopatry. Finally, the influence of temperature and rainfall variables were investigated by testing 'predicted distribution' models (formed using positive distribution records) against negative records for both sets of variables, individually and in combination. The results show that the extent of the distribution of the cicada species is substantially less than that of the associated tree species. The geographical distributions of one of the species may be influenced more by rainfall, or a combination of temperature and rainfall, whereas the other species appears to be more influenced by temperature alone. Both species tolerate soils with a wide range of pH levels, electrical conductivity and forces required for surface penetration. They both showed a strong association with soils that had a silt loam texture, with only few records from sandy soils. However, none of the soils sampled where the cicadas occurred were heavy clays, which suggests that the physical properties of such soils may provide an unsuitable environment for the nymphal stages of the life cycle of these particular cicadas. The resolution of species limits within the P. annulata species complex allowed the redescription of Pauropsalta annulata Goding and Froggatt sensu stricto and the description of 11 new species belonging to the P. annulata species group, all from eastern Australia. Two of these species comprise two subspecies each, also all new. New distribution records and calling song data are documented for the allied species, P. ayrensis Ewart, which is redescribed to include the characters newly recognised in the present study as significant taxonomically with respect to Pauropsalta cicadas. The treatment includes comprehensive descriptions of the morphology and calling songs of the species and subspecies, and separate descriptive keys are provided for both sets of characters. The new taxa comprise P. artatus sp. nov., P. corymbiae sp. nov., P. decorus sp. nov., P. graniticus sp. nov., P. inversus inversus subsp. nov., P. i. laboris subsp. nov., P. notialis notialis subsp. nov., P. notialis incitatus subsp. nov., P. simplex sp. nov., P. subtropicus sp. nov, P. torrensis sp. nov. and P. tremulus sp. nov. Areas of hybridization between P. n. notialis subsp. nov. and P. n. incitatus subsp. nov. are also documented, together with their calling songs and morphology, which justifies their subspecific status. The P. inversus subspecies are allopatric, but consistently differ in the duration between phrases of the calling song. Finally, the results and conclusions are amalgamated into a critical reassessment of what defines species limits and the most appropriate approaches to investigating species limits in sexual organisms. Some historical discussions are revisited, such as the question of the reality of species and how species are perceived under the premises of neoDarwinism. The realism of species demands that species limits are most realistically defined in terms of their fertilisation mechanism, for this delimits the gene pool and thus the distribution of adaptations (the calling song of cicadas, for example).

This thesis investigates aspects of the ecology, morphology, bioacoustics, genetic relatedness, biogeography and taxonomy of cicadas in the Pauropsalta annulata Goding and Froggatt species complex, across their entire eastern Australian geographical distribution. These cicadas seem to emerge at the same sites annually, although the exact duration of their life cycle is not known. They are wary and mobile insects, with individuals generally taking flight at the first sign of disturbance, which makes them challenging to record, capture and observe. The males produce distinctive calling songs that have a consistent rhythmic structure, which is critically important for attracting females. In most cases, the songs are biphasic, with a penetrating introductory or "buzzing" component that apparently functions in mate attraction and a strongly structured cueing or "lilting" component for mate localisation. Initially Pauropsalta annulata was thought to comprise a single species with some degree of geographically structured variation in its calling song. Consequently, various "song types" had been recognised, but their precise relationship to one another had never been investigated. Therefore the structure of their calling songs were compared statistically across individuals of three song types, and this revealed four discrete clusters that were demonstrated to be independent of one another and consistent in their calling song structure. Plotting the geographical distributions of these cicadas revealed that each of the P. annulata song types is independently distributed geographically, with areas of overlap that are relatively small. Calling song structure is consistent for each song type across extensive geographical space and this structure holds even into areas of sympatry. One song type showed consistent differences from the others in male genitalia structure, and female ovipositor length differs significantly among three of the other song types. Each song type was found to be strongly associated with a small number of tree species and these associations are maintained into areas of sympatry. The spatial ecological and morphological comparisons in song structure, plant associations and morphology made in this study demonstrate that the P. annulata song types actually represent a complex of cryptic species. Two of the song types appear to intergrade in areas of sympatric overlap in terms of calling songs, plant associations and morphology, and thus represent subspecies as defined in population genetics terms. The components of the male calling song were then investigated to determine their specific functions and thus how they could delimit species boundaries. Significant differences were found in dominant song frequency between three of the four species, and may contribute to differential mate attraction. Both components of the song of each species have the same dominant frequency. Consistent differences were also evident, among species and subspecies, in the rhythmic structure of the "lilting" component of the male calling song. This component contains repeated phrases and each one of these provides a cue to which the conspecific female may respond. Her response is timed for the brief silent interval between the phrases. At this point the calling song becomes a duet, which enables the male to locate the female, as he actively searches for her on the surrounding branches at this stage. The calling song is discussed in terms of random mating within gene pools of these cicadas and in terms of its role as part of their broader fertilisation mechanism. A molecular analysis of the P. annulata species complex was performed to examine the phylogenetic relationships across 12 species and four subspecies defined in this study, and estimate divergence times within the group. Individual specimens were sampled widely across the geographical distributions of the species and subspecies where possible to account for genetic variation across space. DNA sequences from two loci were amplified: mitochondrial CO1 ("barcoding region") and a large intron from the dynamin nuclear gene. Separate phylogenies were reconstructed for each locus using maximum parsimony procedures and Bayesian posterior sampling with implementation of a relaxed molecular clock. The phylogenies from both genes provided strong support for the monophyly of the P. annulata species complex, and nine of the species were monophyletic based on the CO1 gene. The remaining three emerged non-monophyletic. Based on a clock calibration of 0.0165s/s/myr, the monophyletic clades represented by extant P. annulata species diverged about 4.5-8.0 million years ago. Those species that emerged non-monophyletic had shallower divergences, with the exception of one species, which exhibited haplotype diversity that conferred up to 13.2% sequence divergence between allopatric populations in CO1. Dynamin produced a broadly similar phylogenetic pattern to that of CO1, but the relationships among individuals across the species and subspecies that emerged non-monophyletic differed substantially. This lack of congruence between the two genomes, in combination with the dominance of internal haplotypes in both loci, indicates an overall pattern of deep coalescence rather than interspecific hybridization. Therefore the molecular data do not provide an alternative definition of species limits in the P. annulata species complex, despite some emerging non-monophyletic in this analysis. Sound interpretation of the phylogenetic pattern discovered here would not have been possible without the acoustic, ecological and geographical investigations on species limits that preceded this work. To determine what biological and climatic factors influence the present day distributions of these cicadas, the distributions of two of the most closely related species in the P. annulata species complex were compared with the distributions of the tree species with which each is mainly associated. Because a large part of their life cycle is subterranean, soil texture, pH, electrical conductivity and force required for surface penetration were compared across sites where each of the cicadas occur in sympatry and allopatry. Finally, the influence of temperature and rainfall variables were investigated by testing 'predicted distribution' models (formed using positive distribution records) against negative records for both sets of variables, individually and in combination. The results show that the extent of the distribution of the cicada species is substantially less than that of the associated tree species. The geographical distributions of one of the species may be influenced more by rainfall, or a combination of temperature and rainfall, whereas the other species appears to be more influenced by temperature alone. Both species tolerate soils with a wide range of pH levels, electrical conductivity and forces required for surface penetration. They both showed a strong association with soils that had a silt loam texture, with only few records from sandy soils. However, none of the soils sampled where the cicadas occurred were heavy clays, which suggests that the physical properties of such soils may provide an unsuitable environment for the nymphal stages of the life cycle of these particular cicadas. The resolution of species limits within the P. annulata species complex allowed the redescription of Pauropsalta annulata Goding and Froggatt sensu stricto and the description of 11 new species belonging to the P. annulata species group, all from eastern Australia. Two of these species comprise two subspecies each, also all new. New distribution records and calling song data are documented for the allied species, P. ayrensis Ewart, which is redescribed to include the characters newly recognised in the present study as significant taxonomically with respect to Pauropsalta cicadas. The treatment includes comprehensive descriptions of the morphology and calling songs of the species and subspecies, and separate descriptive keys are provided for both sets of characters. The new taxa comprise P. artatus sp. nov., P. corymbiae sp. nov., P. decorus sp. nov., P. graniticus sp. nov., P. inversus inversus subsp. nov., P. i. laboris subsp. nov., P. notialis notialis subsp. nov., P. notialis incitatus subsp. nov., P. simplex sp. nov., P. subtropicus sp. nov, P. torrensis sp. nov. and P. tremulus sp. nov. Areas of hybridization between P. n. notialis subsp. nov. and P. n. incitatus subsp. nov. are also documented, together with their calling songs and morphology, which justifies their subspecific status. The P. inversus subspecies are allopatric, but consistently differ in the duration between phrases of the calling song. Finally, the results and conclusions are amalgamated into a critical reassessment of what defines species limits and the most appropriate approaches to investigating species limits in sexual organisms. Some historical discussions are revisited, such as the question of the reality of species and how species are perceived under the premises of neoDarwinism. The realism of species demands that species limits are most realistically defined in terms of their fertilisation mechanism, for this delimits the gene pool and thus the distribution of adaptations (the calling song of cicadas, for example).