Durvillaea, a kelp genus occurring only in the Southern Hemisphere, presents an ideal system for studies of marine connectivity and postglacial recolonisation. Durvillaea contains five currently-recognised species, four of which are non-buoyant. Whereas all non-buoyant species are restricted to the south-western Pacific, the sole buoyant species (D. antarctica) has a far wider, circumpolar distribution, strongly suggesting that long-distance dispersal in D. antarctica is achieved by rafting. This contrast in predicted dispersal ability among Durvillaea species provides an opportunity for natural phylogeographic comparisons, thereby assessing the effectiveness of rafting as a long-distance dispersal mechanism. Additionally, the inability of D. antarctica to survive in ice-affected areas, combined with its broad distribution, make it an ideal candidate for studies of postglacial recolonisation. Phylogenetic and biogeographic relationships within Durvillaea were here assessed using sequence data from mitochondrial (COI), chloroplast (rbcL) and nuclear (18S) DNA. Genetic data were obtained from more than 500 specimens, including representatives from across the geographic range of each recognised species of Durvillaea. Mitochondrial data for Durvillaea were found to be highly phylogenetically informative, with 117 variable sites observed over a 629 bp fragment of COI. Chloroplast and nuclear markers, on the other hand, showed less variation than COI, but nonetheless contributed useful phylogenetic information. Phylogenetic analyses were performed using both Maximum Likelihood and Bayesian approaches. Contrasting patterns of genetic diversity were observed across the range of D. potatorum in Australia, with genetic homogeneity throughout western sites versus relatively high levels of diversity in eastern populations. Based on these results, I hypothesise that D. potatorum recolonised much of the western part of its range postglacially, perhaps being entirely eliminated from western Tasmania during the last glacial period by altered oceanographic systems. Additionally, 'western' and 'eastern' D. potatorum haplotypes formed deeply-divergent clades, likely reflecting geographic isolation on either side of the Bassian Isthmus during Pleistocene marine regressions. Substantial genetic diversity was observed across the range of the circumpolar species D. antarctica. Within New Zealand, phylogenetic and morphological analyses of D. antarctica indicate that two morphotypes ('cape' and 'thonged' forms) likely represent reproductively isolated species, with the 'cape' lineage apparently restricted to southern New Zealand. Whereas the 'cape' lineage showed little genetic variation throughout its range, the 'thonged' lineage exhibited marked phylogeographic structure, with high genetic diversity and a clear north - south genetic disjunction delineated by the Canterbury Bight. On a broader, circumpolar scale, D. antarctica showed contrasting patterns of genetic diversity, with high levels of variation in low-latitude regions (e.g., continental coasts of New Zealand and Chile), versus near-homogeneity at high, subantarctic latitudes. These phylogeographic contrasts strongly suggest that D. antarctica recolonised much of the subantarctic region only recently, most plausibly following extirpation by ice scour at the Last Glacial Maximum (LGM). Locations of putative recolonised islands relative to 'refugial' areas indicate that Antarctic sea ice was likely more extensive at the LGM than previous studies have suggested. Latitudinal contrasts in genetic diversity were also observed among Chilean populations of D. antarctica, with a single mitochondrial haplotype detected throughout Chilean Patagonia versus high diversity in central Chile (32�-42�S). The Patagonian populations appear to have recolonised the region postglacially, following recession of the Patagonian Ice Sheet after the LGM. These populations show transoceanic ancestry, with a closer relationship to populations in the subantarctic and southern New Zealand regions than to those in central Chile. Substantial phylogeographic structure was evident across small spatial scales in central Chile, and the correspondence of major genetic disjunctions among central Chilean sites with the presence of long stretches of unsuitable substrate (beaches) strongly suggests that habitat discontinuity drives genetic isolation in this dispersive species. Broad-scale molecular phylogenetic analyses indicate that the current taxonomy of Durvillaea species requires substantial revision. Previously-recognised 'morphotypes' of Durvillaea (of D. potatorum in Australia, and D. antarctica in New Zealand), for example, were here found to be genetically distinct, likely representing reproductively-isolated species. This phylogeographic research on Durvillaea kelp sheds new light on the historical impacts of climate change on Southern Hemisphere marine environments, and on the processes driving evolution in a marine macroalga.

Durvillaea, a kelp genus occurring only in the Southern Hemisphere, presents an ideal system for studies of marine connectivity and postglacial recolonisation. Durvillaea contains five currently-recognised species, four of which are non-buoyant. Whereas all non-buoyant species are restricted to the south-western Pacific, the sole buoyant species (D. antarctica) has a far wider, circumpolar distribution, strongly suggesting that long-distance dispersal in D. antarctica is achieved by rafting. This contrast in predicted dispersal ability among Durvillaea species provides an opportunity for natural phylogeographic comparisons, thereby assessing the effectiveness of rafting as a long-distance dispersal mechanism. Additionally, the inability of D. antarctica to survive in ice-affected areas, combined with its broad distribution, make it an ideal candidate for studies of postglacial recolonisation. Phylogenetic and biogeographic relationships within Durvillaea were here assessed using sequence data from mitochondrial (COI), chloroplast (rbcL) and nuclear (18S) DNA. Genetic data were obtained from more than 500 specimens, including representatives from across the geographic range of each recognised species of Durvillaea. Mitochondrial data for Durvillaea were found to be highly phylogenetically informative, with 117 variable sites observed over a 629 bp fragment of COI. Chloroplast and nuclear markers, on the other hand, showed less variation than COI, but nonetheless contributed useful phylogenetic information. Phylogenetic analyses were performed using both Maximum Likelihood and Bayesian approaches. Contrasting patterns of genetic diversity were observed across the range of D. potatorum in Australia, with genetic homogeneity throughout western sites versus relatively high levels of diversity in eastern populations. Based on these results, I hypothesise that D. potatorum recolonised much of the western part of its range postglacially, perhaps being entirely eliminated from western Tasmania during the last glacial period by altered oceanographic systems. Additionally, 'western' and 'eastern' D. potatorum haplotypes formed deeply-divergent clades, likely reflecting geographic isolation on either side of the Bassian Isthmus during Pleistocene marine regressions. Substantial genetic diversity was observed across the range of the circumpolar species D. antarctica. Within New Zealand, phylogenetic and morphological analyses of D. antarctica indicate that two morphotypes ('cape' and 'thonged' forms) likely represent reproductively isolated species, with the 'cape' lineage apparently restricted to southern New Zealand. Whereas the 'cape' lineage showed little genetic variation throughout its range, the 'thonged' lineage exhibited marked phylogeographic structure, with high genetic diversity and a clear north - south genetic disjunction delineated by the Canterbury Bight. On a broader, circumpolar scale, D. antarctica showed contrasting patterns of genetic diversity, with high levels of variation in low-latitude regions (e.g., continental coasts of New Zealand and Chile), versus near-homogeneity at high, subantarctic latitudes. These phylogeographic contrasts strongly suggest that D. antarctica recolonised much of the subantarctic region only recently, most plausibly following extirpation by ice scour at the Last Glacial Maximum (LGM). Locations of putative recolonised islands relative to 'refugial' areas indicate that Antarctic sea ice was likely more extensive at the LGM than previous studies have suggested. Latitudinal contrasts in genetic diversity were also observed among Chilean populations of D. antarctica, with a single mitochondrial haplotype detected throughout Chilean Patagonia versus high diversity in central Chile (32�-42�S). The Patagonian populations appear to have recolonised the region postglacially, following recession of the Patagonian Ice Sheet after the LGM. These populations show transoceanic ancestry, with a closer relationship to populations in the subantarctic and southern New Zealand regions than to those in central Chile. Substantial phylogeographic structure was evident across small spatial scales in central Chile, and the correspondence of major genetic disjunctions among central Chilean sites with the presence of long stretches of unsuitable substrate (beaches) strongly suggests that habitat discontinuity drives genetic isolation in this dispersive species. Broad-scale molecular phylogenetic analyses indicate that the current taxonomy of Durvillaea species requires substantial revision. Previously-recognised 'morphotypes' of Durvillaea (of D. potatorum in Australia, and D. antarctica in New Zealand), for example, were here found to be genetically distinct, likely representing reproductively-isolated species. This phylogeographic research on Durvillaea kelp sheds new light on the historical impacts of climate change on Southern Hemisphere marine environments, and on the processes driving evolution in a marine macroalga.

The ecological importance of drifting will depend on the abundance of drifting algae and whether it is reproductively viable. However, the ability of adult plants to successfully disperse long-distances by drifting is largely unknown, particularly for fucoids. The abundance, species composition, and reproductive status of drifting algae was examined by transect surveys around Kaikoura and Banks Peninsula. Abundance and species composition varied between sites, but all drifting algae that were in reproductive season, and had reproductive structures intact, were reproductively active. The reproductive longevity and viability of drifting and beach-cast Hormosira banksii, Durvillaea antarctica and Cytophora torulosa was compared with attached populations. Drifting algae remained reproductively viable, and fecundity did not differ from that of attached algae. Viable propagules were released from drifting algae for the duration of the experiments (H. banksii 57 days, D. antarctica 62 days, and C. torulosa 43 days). In contrast, beach-cast algae ceased to release propagules after 14 days. Dispersal by drifting relies on offshore transport after detachment. To determine the influence of wind and tidal currents on the nearshore transport of drifting algae, tagged H. banksii, D. antarctica, C. torulosa and GPS-tracked drifters were released from shore. Drifters generally moved in the direction of the prevailing wind, but some influence of tidal direction and bathymetry was detected. Offshore winds and outgoing tides were favourable for the offshore transport of surface drifting algae. Following dispersal and arrival at new locations, the distance between gametes may be important in determining the fertilisation success of dioecious species. Experiments testing the fertilisation success of H. banksii and D. antarctica, over increasing distances, showed that fertilisation success decreased with increasing distance between male and female gametes. Despite this, eggs were fertilised when male and female gametes were up to 2m apart, and sperm remained viable for 2 hours.

The ecological importance of drifting will depend on the abundance of drifting algae and whether it is reproductively viable. However, the ability of adult plants to successfully disperse long-distances by drifting is largely unknown, particularly for fucoids. The abundance, species composition, and reproductive status of drifting algae was examined by transect surveys around Kaikoura and Banks Peninsula. Abundance and species composition varied between sites, but all drifting algae that were in reproductive season, and had reproductive structures intact, were reproductively active. The reproductive longevity and viability of drifting and beach-cast Hormosira banksii, Durvillaea antarctica and Cytophora torulosa was compared with attached populations. Drifting algae remained reproductively viable, and fecundity did not differ from that of attached algae. Viable propagules were released from drifting algae for the duration of the experiments (H. banksii 57 days, D. antarctica 62 days, and C. torulosa 43 days). In contrast, beach-cast algae ceased to release propagules after 14 days. Dispersal by drifting relies on offshore transport after detachment. To determine the influence of wind and tidal currents on the nearshore transport of drifting algae, tagged H. banksii, D. antarctica, C. torulosa and GPS-tracked drifters were released from shore. Drifters generally moved in the direction of the prevailing wind, but some influence of tidal direction and bathymetry was detected. Offshore winds and outgoing tides were favourable for the offshore transport of surface drifting algae. Following dispersal and arrival at new locations, the distance between gametes may be important in determining the fertilisation success of dioecious species. Experiments testing the fertilisation success of H. banksii and D. antarctica, over increasing distances, showed that fertilisation success decreased with increasing distance between male and female gametes. Despite this, eggs were fertilised when male and female gametes were up to 2m apart, and sperm remained viable for 2 hours.