Arid and semiarid river systems in Western Queensland, Australia, are characterized by the unpredictable and highly variable nature of their hydrological regimes as a result of the episodic nature of rain events in the region. These dryland rivers typically experience episodic floods and extremely low or no flow periods. During low or no flow periods, water persists only in relatively wide and deep sections of the river channels, which are called 'waterholes'. These isolated waterholes serve as refugia for aquatic species during protracted intervals between floods. In such discontinuous riverine habitat, dispersal of freshwater species may be achieved only during wet seasons, when water is flowing in rivers and the nearby floodplains. Obligate aquatic species occur in habitats that represent discrete sites surrounded by inhospitable terrestrial landscapes. Thus, movements are very much limited by the physical nature and arrangement of the riverine system. In addition, the distribution of a species may be also largely dependent on historical events. Landscape and river courses continually change over geological time, often leaving distinct phylogenetic 'signatures', useful in reconciling species' biology with population connectivity and earth history. The main aim of this study was to resolve the relative importance of contemporary and historical processes in structuring populations of two freshwater species in Western Queensland river systems. To address this aim, a comparative approach was taken in analysing patterns of genetic variation of two freshwater invertebrates: a snail (Notopala sublineata) and a prawn (Macrobrachium australiense). Mitochondrial sequences were used for both the species. In addition, allozyme and microsatellites markers were employed for N. sublineata. These species have similar distributions in Western Queensland region, although N. sublineata appears to be extinct in some catchments. M. australiense is thought to have good dispersal abilities due to a planktonic larval phase in its life cycle and good swimming capabilities, whereas N. sublineata is thought to have limited dispersal abilities, because of its benthic behaviour and because this species is viviparous. It was hypothesised that these freshwater invertebrates, would display high levels of genetic structure in populations, because physical barriers represented by terrestrial inhospitable habitat, are likely to impede gene flow between populations inhabiting isolated river pools. Genetic data for the two species targeted in this study supported this hypothesis, indicating strong population subdivision at all spatial scales investigated (i.e. between and within catchments). This suggests that contemporary dispersal between isolated waterholes is relatively restricted, despite the potential good dispersal abilities of one of the species. It was hypothesised that levels of gene flow between populations of aquatic species were higher during the Quaternary (likely movements of individuals across catchment boundaries) and that they have been isolated relatively recently. There is evidence that historically gene flow was occurring between populations, suggesting that episodic dispersal across catchment boundaries was likelier in the past. Episodic historical movements of aquatic fauna were facilitated by higher patterns of river connectivity as a result of the climate changes of the Pleistocene. Because the two species targeted in this study exhibit analogous spatial patterns of evolutionary subdivision it is likely that they have a shared biogeographic history. The unpredictable flow regime of rivers in Western Queensland is likely to have considerable effects on the genetic diversity of aquatic populations. First, if populations of obligate freshwater organisms inhabiting less persistent waterholes are more likely to experience periodic bottlenecks than those inhabiting more persistent ones, they would be expected to have lower levels of genetic diversity. Second, if populations inhabiting less persistent waterholes periodically undergo local extinction with subsequent recolonisation, there should be higher levels of genetic differentiation among them, due to the founder effects, than among those populations inhabiting more persistent waterholes. Contrary to the first prediction, the observed levels of genetic diversity in both N. sublineata and M. australiense were high in both more persistent and less persistent waterholes. There was no tendency for genetic diversity to be lower in less persistent than in more persistent waterholes. However, when Cooper waterholes were ranked in order of persistence, positive correlation between water persistence time in waterholes and genetic diversity was detected in N. sublineata but not in M. australiense. Contrary to the second prediction, highly significant genetic differentiation was found among populations from both less persistent and more persistent waterholes. This indicates that not only populations from less persistent but also those from more persistent waterholes were very dissimilar genetically. This study demonstrated the importance of both contemporary and historical processes in shaping the population structure of obligate freshwater species in Western Queensland river systems. It has indicated that contemporary movements of freshwater species generally are extremely limited across the region, whereas episodic dispersal across catchment boundaries was possible during the Pleistocene, due to different patterns of river connectivity.

Macrobrachium australiense is a common freshwater prawn found throughout most of eastern and inland Australia. Debate has been ongoing on the systematics of this species due to high morphological variation: past studies have relied on external morphology assessments to describe genetic relationships between populations. An individual's morphological phenotype results however, from an interplay of genetic factors, environmental, and interactions between genes and the environment. The current study examined the strength of genetic constraint on morphological traits in this species. Examination of over 1000 M. australiense museum specimens collected from across the species' extensive natural range, documented high phenotypic variation with no regional pattern of variation. Within regions, 88% of variation in morphological traits in mature males and juveniles was present between local rivers. Therefore, morphological variation is not structured at the regional level. If there is a strong genetic base to morphological variation then populations in a single river system must be evolving essentially in isolation. More intensive sampling within a single river system demonstrated high morphological variation in 600 M. australiense individuals from 18 populations within a geographically connected system. Populations separated by as little as 1km showed significant morphological differences in 50% of mature male traits. If morphological variation is primarily genetically based, then populations within a river system were evolving independently at a very fine spatial scale. This hypothesis was tested by breeding morphologically divergent populations of M. australiense in a controlled environment to isolate genetic influences on morphological variation. Low heritability for morphological traits in five divergent populations raised under identical environments established that there is no strict genetic control on morphological variation in these M. australiense populations. Morphologically homogenous offspring resulted from wild parents that had exhibited significant differences at 73% of traits examined. Therefore, the fundamental assumption that morphological variation in M. australiense is dictated by strict genetic control is not supported in these representative populations. Moreover, significant variation in 41% of morphological traits was produced by raising a single population at different environmental temperatures (28oC and 22 oC). A single homogenous stock of M. australiense should not produce morphologically divergent offspring if genetic factors are the major influence of phenotypic expression. Crossing of pure line divergent stocks resulted in hybrid offspring with significant differences in 50% of female morphological traits, whereas male offspring varied for only 31% of morphological traits. This result suggests that female morphological expression is affected more strongly by genetic factors than male offspring in this trial. The growth and maturation of external morphological traits during development in M. australiense is under limited genetic constraint, especially in the later phases of growth. Only 17% of traits varied between juvenile stocks in the last three months of development when individuals were exposed to identical environmental conditions. Maturation size was homogenous, except for females in the absence of maturing males in divergent stocks exposed to identical environmental conditions. Females were much larger in size and shape of morphological traits in the absence of mature males in the population. Thus environmental factors strongly influence phenotypic expression of external morphology in M. australiense. Past problems with the taxonomy of this species are therefore understandable as many important traits used in systematics appear to be under limited genetic control. Past evolutionary studies based on morphological diversity in this species therefore may be unreliable as the traits used to identify divergent forms may not provide a true reflection of genetic divergence.