<p><strong>Population connectivity drives processes both at ecological and evolutionary scales, including source-sink dynamics that impact population dynamics, and population structuring that results in spatially explicit genetic variation. Because of this, an understanding of connectivity is vital to successful species management. This is especially critical for kaimoana (seafood) species that support important seafood industries in New Zealand, such as the green-lipped mussel, Perna canaliculus, the blackfoot paua, Haliotis iris, and the red rock lobster, Jasus edwardsii. Previous studies of population connectivity in these species have applied population genetic analyses to identify spatially explicit genetic structure. However, important questions about connectivity patterns in these species still remain. The possibility of finer scale population structure than has been reported by previous genetic analyses may require a revisiting of the stock boundaries currently used by New Zealand’s quota management system, and the discovery of potential source-sink dynamics could help identify sites for conservation priority. The present research addresses these questions by applying a multidisciplinary genetic and oceanographic approach to describing population connectivity. For all three species, publicly available nationwide microsatellite marker data sets were analysed using a new panel of statistical tools designed to both reveal higher resolution population structure and identify asymmetries in gene flow between populations in order to describe source-sink dynamics (Chapter 2). For P. canaliculus (but not H. iris or J. edwardsii), this data was complemented with a biophysical Lagrangian particle tracking model that was used to simulate larval dispersal events in a hydrodynamic model of ocean circulation around New Zealand. This modelling approach allowed for further examination of connectivity at the national scale, particularly into the existences of asymmetries in particle transport among populations, and allowed for connectivity estimates for all populations around the country, including sites where genetic samples were not available. Nearly 900 millions particle were released over 23 years of the hydrodynamic model and were given settlement parameters to mimic the pelagic larval dispersal of P. canaliculus. Oceanographic connectivity as simulated by the model was calculated between all sites from which genetic data were available (Chapter 3) and then from all possible sites containing suitable mussel habitat, including connectivity over multiple generations through stepping stones (Chapter 4). Additional genetic analyses of previously published microsatellite data sets provided very little new information on population connectivity for any of the subject species. While these analyses did not reveal either higher resolution spatial genetic structuring or evidence for asymmetries in gene flow among populations as had been hypothesized, several potential explanations for this are identified. Limitations of genetic analyses for studying population connectivity in high dispersal broadcast spawners are discussed, along with avenues for future research. Modelled oceanographic connectivity between populations showed a significant correlation with pairwise measure of genetic distance. Additionally, clusters identified in the oceanographic matrix reflect those reported by genetic analyses, namely a northern and southern cluster with a discontinuity near Cook Strait. Importantly however, this discontinuity did not coincide exactly with the break reported by genetic analyses, suggesting that factors other than oceanographic circulation also play a role in maintaining this genetic structure. The biophysical model data was able to detect important asymmetries in larval transport between populations, driven by oceanographic currents at the national and regional scale, identifying potentially important larval source populations in the southeastern South Island and north of Banks Peninsula Other notable sites include Kahurangi and the Marlborough Sounds at the northern end of the South Island, which play a role as larval sources as well as acting as stepping stones between the northern and southern clusters of mussel populations, and mussel populations along the Waikato coast and at Ninety Mile Beach in Northland, where populations are especially reliant on self recruitment, making them susceptible to local environmental disturbances. This is especially significant in the case of Ninety Mile Beach, where wild caught mussel spat supplies 80% of the individuals grown by New Zealand’s Greenshell™ aquaculture industry.</strong></p><p>This research represents the first application of this multidisciplinary genetic-oceanographic approach to connectivity at the national scale in New Zealand. By simulating larval dispersal across the entire range of the endemic green-lipped mussel, P. canaliculus, several regions of interest for future study demographic and genetic studies have been identified. This work highlights the value and efficiency of combining high power computational models with empirical population genetic analyses to achieve a better understanding of population connectivity and is recommended as a tool be used by managers of marine invertebrate species around the world to ensure adequate protection of all clusters of genetic diversity as well as important source and stepping stone populations.</p>