Glacial oceanographic contrasts explain phylogeography of Australian bull kelp

Molecular Ecology (2009) 18, 2287–2296

doi: 10.1111/j.1365-294X.2009.04201.x

Glacial oceanographic contrasts explain phylogeography of Australian bull kelp Blackwell Publishing Ltd

C E R I D W E N I . F R A S E R , H A M I S H G . S P E N C E R and J O N AT H A N M . WAT E R S Allan Wilson Centre for Molecular Ecology and Evolution, Department of Zoology, University of Otago, 340 Great King St, Dunedin 9016, New Zealand

Abstract The evolutionary effects of Southern Hemisphere Pleistocene oceanographic conditions — marked by fluctuations in sea levels and water temperatures, and redirected currents — are poorly understood. The southeastern tip of Australia presents an intriguing model system for studying the biological impacts of palaeoceanography. In particular, contrasting oceanographic conditions that existed on eastern vs. western sides of the Bassian Isthmus during Pleistocene glacial periods allow for natural comparisons between putative refugial vs. re-invading populations. Whereas many western Tasmanian marine taxa were likely eliminated by cold subantarctic water during the last glacial period, eastern Tasmanian populations would have persisted in relatively warm temperatures mediated by the ongoing influence of the East Australian Current (EAC). Here we test for the effects of contrasting palaeoceanographic conditions on endemic bull kelp, Durvillaea potatorum, using DNA sequence analysis (COI; rbcL) of more than 100 individuals from 14 localities in southeastern Australia. Phylogenetic reconstructions reveal a deep (maximum divergence 4.7%) genetic split within D. potatorum, corresponding to the ‘eastern’ and ‘western’ geographical regions delimited by the Bassian Isthmus, a vicariant barrier during low Pleistocene sea levels. Concordant with the western region’s cold glacial conditions, samples from western Tasmania and western Victoria are genetically monomorphic, suggesting postglacial expansion from a mainland refugium. Eastern samples, in contrast, comprise distinct regional haplogroups, suggesting the species persisted in eastern Tasmania throughout recent glacial periods. The deep east–west divergence seems consistent with earlier reports of morphological differences between ‘western’ and ‘eastern’ D. potatorum, and it seems likely that these forms represent reproductively isolated species. Keywords: cytochrome c oxidase I, Durvillaea potatorum, palaeoceanography, phylogeography, RuBisCo Received 9 February 2009; revision received 18 March 2009; accepted 21 March 2009

Introduction The evolutionary and biogeographical significance of glacial periods is well recognized, particularly for terrestrial systems affected by ice sheets (reviewed by Hewitt 1996, 2000; Provan & Bennett 2008). In the marine realm, fluctuation of sea levels, water temperatures and sea ice driven by glacial cycles are also likely to have had a strong Correspondence: Ceridwen I. Fraser, Allan Wilson Centre for Molecular Ecology and Evolution, Department of Zoology, University of Otago, 340 Great King St, Dunedin 9016, New Zealand. Fax: +64-(0)3-4797584; E-mail: [email protected] © 2009 Blackwell Publishing Ltd

influence on species distributions and genetic connectivity (Maggs et al. 2008). For instance, lowered sea levels during glacial maxima can lead to the emergence of land bridges which, while facilitating dispersal of terrestrial biota (e.g. Tiffney 1985; reviews by Wen 1999; Sanmartin et al. 2001), effectively divide marine populations. Scientists have speculated on such biotic effects for centuries; Darwin (1872), wrote ‘a narrow isthmus now separates two marine faunas; submerge it, or let it formerly have been submerged, and the two faunas will now blend ... or may formerly have blended.’ The Sunda and Sahul shelves, for example, separated many Pacific and Indian Ocean marine taxa at Pleistocene sea level lowstands, leading to vicariant

2288 C . I . F R A S E R , H . G . S P E N C E R and J . M . WAT E R S

Fig. 1 Approximate coastlines in southeastern Australia at three points in recent geological history: the present (with the highest sea levels in more than 100 000 years); 14 000 years ago (when Tasmania was last joined to the mainland); and 25 000 years ago (the lowstand of sea levels during the last glacial maximum, LGM) (shoreline reconstructions modified from Lambeck & Chappell 2001). Major contemporary surface currents — the East Australian (EAC) and Zeehan (ZC) (Baines et al. 1983; Cresswell 2000) — are shown by red arrows; altered oceanographic conditions during the LGM, with the subtropical convergence (STC) further north and the Zeehan current blocked, are indicated in the right panel (after Nürnberg et al. 2004). Approximate northern limits of the subtropical convergence at both modern and LGM time slices are indicated (after Nürnberg et al. 2004). The position of the enlarged region is shown on the map of Australia (inset: lower left). Distributional range of D. potatorum is indicated by blue arrows at the most western (Robe, South Australia) and eastern (Tathra, New South Wales) sites where this species can be found (Huisman 2000).

divergence (e.g. Chenoweth et al. 1998; Nelson et al. 2000; Reid et al. 2006), while similar marine biogeographical disjunctions have been noted on either side of the Baja California Peninsula (e.g. Bernardi et al. 2003). In Australia, the Bassian Isthmus — a land bridge that periodically connects Tasmania to the mainland as sea levels drop during glaciations (Lambeck & Chappell 2001) (Fig. 1) — has frequently been suggested as a potential vicariant barrier to gene flow among temperate marine taxa (Dartnall 1974). Indeed, recent phylogeographical analyses of several marine species (e.g. cirrhitoid fish: Burridge 2000; Patiriella sea stars: Waters et al. 2004; Catostylus jellyfish: Dawson 2005; Nerita gastropods: Waters et al. 2005; Spencer et al. 2007; Waters 2008; Catomerus barnacles: York et al. 2008) in southeastern Australia have revealed distinct geographical partitioning into eastern vs. western clades, generally consistent with the vicariant hypothesis of allopatric divergence during glacial periods. A less-explored aspect of climate change is the effect of altered oceanography (specifically currents) on the phylogeography of modern marine taxa. Pleistocene glacial periods were characterized by both substantially reduced water temperatures and redirected currents (Wells &

Okada 1996; Herbert et al. 2001), but although some authors have suggested that ancient changes in major Northern Hemisphere current systems may partially explain modern phylogeographical structuring (e.g. Muss et al. 2001; Wares 2001; Marko 2004), few studies, particularly in the Southern Hemisphere, have directly assessed the impacts of such palaeoceanographic features. Around southeastern Australia, the subtropical convergence (STC), a boundary zone between temperate and subantarctic waters, is believed to have abutted southern Tasmania and extended along its western coast during recent glacial maxima, blocking the eastward flow of the warm Leeuwin/ Zeehan current (Wells & Okada 1996; McGowran et al. 1997; Nürnberg et al. 2004) (Fig. 1). The absence of the Zeehan Current from southern Tasmania at glacial maxima would have strengthened east–west isolation by limiting dispersal, and the colder waters on the eastern coast may have driven temperate marine organisms to considerably higher latitudes; in contrast, the warm East Australian Current (EAC) is known to have extended its southward flow along eastern Tasmania during glacial periods (Nürnberg et al. 2004), potentially maintaining temperate marine conditions on the eastern, but not western, coasts of Tasmania. © 2009 Blackwell Publishing Ltd

P O S T G L A C I A L R E C O L O N I Z AT I O N I N A U S T R A L I A N B U L L K E L P 2289 Table 1 Morphological and ecological differences between ‘eastern’ and ‘western’ forms of Durvillaea potatorum in southeastern Australia, after Cheshire et al. (1995) Characteristic

Western

Eastern

Primary stipe length Primary stipe distally branched Secondary stipes on primary stipe Habitat

generally < 0.5 m never never always subtidal, < 10 m

generally < 0.5 m or < 1.0 m infrequently infrequently infrequently intertidal, or subtidal < 10 m

Here we present the first direct assessment of the expected phylogeographical effects of these contrasting glacial oceanographic conditions. Australian bull kelp, Durvillaea potatorum (Labillardière), forms a dense band in the upper subtidal and intertidal of exposed rocky reefs from Robe in South Australia (Huisman 2000) to Tathra on the east coast of New South Wales (Millar 2007) (Fig. 1), and along the western and eastern shores of Tasmania. It grows abundantly on King Island, on the western sill of Bass Strait, but is rare on the Furneaux Island Group of the eastern sill (Hay 1994). Durvillaea potatorum is considered endemic to Australia (Cheshire et al. 1995); additional historical records reporting D. potatorum from the subantarctic Auckland Islands (cited in Papenfuss 1964) and from Constitución in Chile (cited in Ramirez & Santelices 1991) are extremely dubious and require confirmation. In western Bass Strait, detached, beach-cast D. potatorum wrack is so abundant that a flourishing kelp-wrack harvesting industry for alginates operates on King Island (Kirkman & Kendrick 1997). Durvillaea potatorum is an unusually large macro-alga, with individuals commonly reaching several metres in length (Cheshire & Hallam 1988) and weighing over 40 kg (Hay 1994); an early botanist described it as ‘a plant which, when fully grown, is measured, not by inches, but by fathoms’ (Harvey 1863). Cheshire & Hallam (1989) noted morphological differences in individuals from eastern vs. western populations: those from the east (eastern coasts of Tasmania and New South Wales) were found to have some — albeit variable — characteristics (such as stipe length, habitat and position of secondary stipes) that differed from western specimens (southern and western coasts of Victoria, and King Island in Bass Strait) (Table 1). Like other members of its genus, D. potatorum is dioecious, with both female and male individuals, producing eggs and sperm, respectively. Although these gametes are released into the water column, and could theoretically disperse with currents, unfertilized eggs are unlikely to survive in a viable condition for more than a few days at sea (M. Clayton, Monash University, Melbourne, personal communication) and fertilized eggs are rapidly — within minutes — surrounded by adhesive secretions that effectively attach zygotes to the substrate (Clayton & Ashburner 1994). Dispersal via gametes or zygotes seems therefore © 2009 Blackwell Publishing Ltd

unlikely to be successful over long distances, particularly where populations are patchy or separated by stretches of sandy coast or open ocean (Hidas et al. 2007). As D. potatorum is a solid-bladed, nonbuoyant member of its genus (Hay 1994), adults are also unlikely to disperse far once detached. A kelp with such limited potential for dispersal would be expected to show strong phylogeographical differentiation and this feature, combined with the distribution of D. potatorum in Australia around Bass Strait, make it well suited for studies of the effects of Pleistocence vicariance and oceanography. Here we examine phylogeographical structure in Australian bull kelp to test the hypotheses that: (i) Pleistocene emergence of the Bassian Isthmus separated this species into distinct eastern and western lineages, and (ii) altered oceanography of Tasmania during recent glacial periods, with subantarctic water impacting the western (but not eastern) coasts, will be reflected in contrasting phylogeographical patterns consistent with elimination of the western and persistence of the eastern populations.

Materials and methods Site selection and sample collection Tissue samples were collected by hand, at low tide, from the distal ends of fronds of attached Durvillaea potatorum growing in the intertidal or shallow subtidal at 14 sites across the species’ range (Table 2). In some cases, where attached individuals were not obtainable due to extreme conditions, samples were taken from fresh beach-cast wrack (listed in Table 2). At one site — King George III Reef — all samples were taken from attached plants in the subtidal, using SCUBA, at between 10 and 12 m depths. Sites were categorized on the basis of geography as either eastern or western, and either mainland or Tasmania (Table 2).

DNA extraction, sequencing and phylogenetic analyses Tissue preservation, DNA extractions and polymerase chain reactions were carried out as described in Fraser et al. (2009a), with a 629-bp portion of the mitochondrial gene cytochrome c oxidase I (COI) amplified using primers

2290 C . I . F R A S E R , H . G . S P E N C E R and J . M . WAT E R S Table 2 Sample sites and localities, showing sample type (attached kelp: ‘a’; fresh beach-cast wrack: ‘w’), number of samples sequenced and haplotypes found. For precise location of sites, refer to Fig. 2 No. of samples sequenced

Haplotype/s

Site name

Site locality

Sample

COI

rbcL

COI

rbcL

Port Fairy Warrnambool Peterborough Phillip Island Green Point Couta Rocks KGIII Reef Bruny Island Pirates Bay Bicheno Binalong Bay The Gardens Gabo Island Tathra TOTAL sequenced

Western-mainland Western-mainland Western-mainland Western-mainland Western-Tasmania Western-Tasmania Eastern-Tasmania Eastern-Tasmania Eastern-Tasmania Eastern-Tasmania Eastern-Tasmania Eastern-Tasmania Eastern-Mainland Eastern-Mainland

a a a a w w a w, a a w a a a a

5 7 6 9 9 10 7 10 10 5 8 7 6 8 107

5 5 4 10 6 6 3 9 7 4 0 8 5 5 77

C-I C-I C-I C-I C-I C-I C-IV, C-V C-I, C-IV C-II, C-III C-II, C-III C-II, C-III C-II, C-III C-VI, C-VII, C-VIII C-VI, C-VIII

R-I R-I R-I R-I R-I R-I R-III R-I, R-III R-II, R-III R-I, R-II, R-III

GazF1 and GazRI (Saunders 2005), and a 886-bp region of the chloroplast gene RuBisCO (rbcL) amplified using primers KL2 and KL8 (Lane et al. 2006). Phylogenies for each data set were built by maximum likelihood (ML) and Bayesian analyses, and included outgroup sequences from Durvillaea willana (South Island, New Zealand; GenBank accessions: COI: EU918569; rbcL: EU918578) and Durvillaea antarctica (Falkland Islands; GenBank accessions: COI: FJ550107; rbcL: FJ550124). ML analyses were performed with a HKY + G model for both COI (base frequencies A = 0.2182, C = 0.1726, G = 0.1895, T = 0.4197; Tratio = 3.7086; gamma shape parameter = 0.0144) and rbcL (base frequencies A = 0.2922, C = 0.1667, G = 0.2168, T = 0.3243; Tratio = 2.9827; gamma shape parameter = 0.0081), that was selected using the hLRT of ModelTest 3.06 (Posada & Crandall 1998). The robustness of the ML topology was estimated by bootstrapping (Felsenstein 1985), with heuristic analysis of 1000 replicate data sets (Figs 2 and 3). Bayesian posterior probability (PP) values, also indicated on the ML phylogenies (Figs 2 and 3), were calculated using MrBayes 3.1.2 (Ronquist & Huelsenbeck 2003). Markov chain Monte Carlo (MCMC) searches, each with four chains of 5 000 000 generations and trees sampled each 100 generations, were carried out. The first 10 000 trees sampled were discarded as ‘burn-in’, based on the stationarity of ln L as assessed using Tracer version 1.4 (Rambaut & Drummond 2007); a consensus topology and posterior probability values were calculated with the remaining trees. An unrooted statistical parsimony network was reconstructed using tcs 1.21 (Clement et al. 2000) in order to examine relationships among ‘eastern’ haplotypes for COI — these analyses were

R-II, R-III R-IV R-IV

not performed on rbcL data due to limited diversity (only two haplotypes in each major clade). Analysis of the precise expansion time of the ‘western’ clade was not possible, as robust mutation rate estimates have not yet been developed for the Phaeophyceae (Hoarau et al. 2007). All unique DNA sequences obtained during this study were deposited with GenBank (Accession nos FJ872919–FJ873102).

Results DNA sequencing of 107 Durvillaea potatorum specimens yielded eight distinct haplotypes for COI, whereas 77 rbcL sequences yielded a total of four haplotypes. Thirty variable sites were found over the 629-bp fragment amplified for COI (4.8% of sites), with approximately one-third of the inferred changes being transversions. For rbcL, seven variable sites were found across the 886-bp fragment (0.8%), with all but two sites showing transitional mutations. In both markers, variable sites occurred at the third codon position, with the exception of one inferred transitional mutation in COI. ML and Bayesian phylogenetic analyses yielded trees with consistent topologies for each genetic data set. Specifically, both COI (Fig. 2), and rbcL (Fig. 3) analyses revealed a deep and strongly supported genetic split within D. potatorum (uncorrected distances up to 4.7% for COI and up to 0.9% for rbcL; Bayesian PP values of 1.00), largely corresponding to ‘eastern’ and ‘western’ geographical regions (Figs 2 and 3). Although both ‘eastern’ and ‘western’ haplotypes were found along the east coast of Tasmania, the presence of all ‘eastern’ haplotypes at only eastern geographical locations, © 2009 Blackwell Publishing Ltd

P O S T G L A C I A L R E C O L O N I Z AT I O N I N A U S T R A L I A N B U L L K E L P 2291

Fig. 2 Phylogeographical structure of Durvillaea potatorum in southeastern Australia, based on mitochondrial (COI) DNA sequence data. The phylogenetic tree (inset: lower left) was generated using ML analyses, and shows Bayesian PP values (above the line) and ML bootstraps in parentheses below. Pie charts on the map indicate the relative proportions of haplotypes at each site, with the total number of samples per site given in parentheses after the site name. Pie charts alongside the phylogenetic tree indicate the proportions of each haplotype in each major clade (‘east’ or ‘west’).

and the broad geographical range of ‘western’ haplotypes across western coasts, indicates a general east–west split despite some geographical overlap. The more informative COI data set also revealed considerable diversity and phylogeographical structuring within the ‘eastern’ clade, which comprised a total of six haplotypes (uncorrected distances 0.2–0.8%). Three were restricted to the east coast of Tasmania (haplotypes C-III–C-V), and three were found only at eastern mainland sites (haplotypes C-VI–C-VIII) (Fig. 2). Although rbcL data showed relatively little overall diversity, the genetic division between eastern mainland and eastern Tasmania was still apparent, with a single ‘eastern’ haplotype (R-III) across Tasmanian sites, and another (R-IV) shared by Tathra and Gabo Island samples (Fig. 3). Strong phylogenetic concordance was found between mitochondrial and chloroplast markers, with all specimens allocated to the equivalent clade (‘eastern’ or ‘western’) for each marker. The ‘western’ clade in both COI and rbcL data sets each yielded just two haplotypes, one of which was particularly geographically widespread, extending from Port Fairy on Australia’s mainland to Bruny Island in southern Tasmania © 2009 Blackwell Publishing Ltd

(C-I, Fig. 2 and R-I, Fig. 3). For rbcL, the widespread haplotype R-I (which differs from R-II by merely one transversion) also occurred at Bicheno in a single sample, although the same individual yielded the less common ‘western’ haplotype for COI (C-II). Notably, D. potatorum populations along the western coasts of both the mainland and Tasmania were completely homogeneous for both markers, showing no evidence of genetic division across Bass Strait. The ‘western’ clade also extended into the eastern coast of Tasmania, with haplotypes C-II and R-II found from Pirates Bay northward to The Gardens (Figs 2 and 3). Analysis of the relationships of COI haplotypes resulted in two networks — corresponding to the ‘eastern’ and the ‘western’ clades — that could not be parsimoniously connected at the 95% confidence limit. The ‘western’ network contained only two COI haplotypes, separated by one mutational step, and was therefore largely uninformative. The ‘eastern’ network (Fig. 4) showed two potentially ancestral, central haplotypes, one from the mainland (CVI) haplogroup and one from Tasmania (C-IV). The four remaining haplotypes radiated from an ancestral

2292 C . I . F R A S E R , H . G . S P E N C E R and J . M . WAT E R S

Fig. 3 Phylogeographical structure of Durvillaea potatorum in southeastern Australia, based on chloroplast (rbcL) DNA sequence data. The phylogenetic tree (inset: lower left) was generated using ML analyses, and shows Bayesian PP values (above the line) and ML bootstraps in parentheses below. Pie charts on the map indicate the relative proportions of haplotypes at each site, with the total number of samples per site given in parentheses after the site name. Pie charts alongside the phylogenetic tree indicate the proportions of each haplotype in each major clade (‘east’ or ‘west’).

haplotype via a single mutational step. The ‘ancestral’ haplotypes — and the mainland vs. Tasmania groups — were themselves separated by two mutational steps, linked by a hypothetical undetected haplotype.

Discussion Ancient vicariant influences The substantial phylogenetic split between ‘eastern’ and ‘western’ clades of Durvillaea potatorum (Figs 2 and 3) supports the hypothesis that past emergence of the Bassian Isthmus has promoted vicariant divergence in this poorly dispersive shallow-water marine macro-alga (hypothesis i). Although contemporary longitudinal variation in water temperature, salinity and wave action might also be suggested as contributing causes, the presence of both clades across a wide geographical area and latitudinal range, as well as their co-occurrence along the east coast of Tasmania, would appear to refute a causal role for such abiotic factors in generating the deep phylogeographical structuring observed for D. potatorum. This study therefore emphasizes the important role of vicariance in driving

diversification of Australia’s marine temperate biota (e.g. Dartnall 1974; Burridge 2000; Waters et al. 2005; York et al. 2008). Indeed, Waters (2008) proposed that the Bassian Isthmus had likely driven cladogenesis — with cryptic eastern and western lineages — in many components of southern Australia’s littoral and shallow sublittoral communities. The suggestion that southern Australia’s major marine biogeographical break occurs specifically in the vicinity of Wilsons Promontory in southern Victoria (e.g. Dawson 2005; Waters 2008) is not rejected by our findings, with the Phillip Island site immediately to the west of this feature fixed for the common ‘western’ haplotype. Clearly, however, finer-scale research is required to confirm the precise location of the phylogeographical disjunction, although the patchiness of D. potatorum would limit the possibility of geographically intensive sampling. Wilsons Promontory (Fig. 1) lies at the final point of contact between Tasmania and mainland Australia during submerging of the Bassian Isthmus; it represents the site of reunion of western and eastern mainland geminate taxa as sea levels rose and is therefore a natural biogeographical break. A long stretch of uninhabitable sandy coast to the immediate east of Wilsons © 2009 Blackwell Publishing Ltd

P O S T G L A C I A L R E C O L O N I Z AT I O N I N A U S T R A L I A N B U L L K E L P 2293

Fig. 4 Haplotype network diagram of the ‘eastern’ clade of Durvillaea potatorum based on mtDNA (COI), with circle area representing haplotype frequency. ‘Ancestral’ haplotypes are indicated by asterisks.

Promontory (Ninety Mile Beach, Fig. 1) presumably inhibits genetic exchange (or dispersal) across this region (Hidas et al. 2007; York et al. 2008; Ayre et al. 2009), enhancing isolation of eastern vs. western mainland populations. An equivalent point cannot be identified in Tasmania, as D. potatorum does not grow along the relatively sheltered northern coast.

Ancient oceanographic influences The broad genetic homogeneity of the ‘western’ D. potatorum, with one particularly widespread haplotype ranging from Port Fairy (the westernmost site sampled) to Bruny Island in southeastern Tasmania (Figs 2 and 3), contrasts with the relatively high diversity found in the ‘eastern’ clade for COI. The genetic homogeneity observed across western populations is unlikely to be due to anthropogenic translocation (e.g. Voisin et al. 2005), as the range of D. potatorum has apparently changed little since early records (see Harvey 1863). Patterns of low-latitude genetic diversity vs. high-latitude homogeneity are frequently interpreted as evidence of postglacial recolonization in the Northern Hemisphere, with rapid range expansion by leading-edge migrants (Hewitt 1996, 2000; Maggs et al. 2008; Provan & Bennett 2008). Similarly, extensive sea ice during the last glacial maximum (LGM) has been linked © 2009 Blackwell Publishing Ltd

to patterns of high-latitude homogeneity in Durvillaea antarctica in the Southern Hemisphere (Fraser et al. 2009b). The monomorphism observed here in western D. potatorum most probably reflects recent recolonization following postglacial retraction of the subtropical convergence and cold waters that affected western Tasmania during the last glacial period (Fig. 1; Wells & Okada 1996; McGowran et al. 1997; Nürnberg et al. 2004) (hypothesis ii). This postglacial expansion might have been facilitated by the strong Zeehan Current (a continuation of the Leeuwin Current; Ridgway & Condie 2004) that flows east from the southern coast of mainland Australia to southern Tasmania during interglacial periods (Fig. 1). In contrast, the southern extension of the warm EAC during glaciations (Nürnberg et al. 2004) presumably allowed temperate marine biota — such as D. potatorum — to persist along the eastern coasts of Tasmania. If the genetic homogeneity of western D. potatorum does indeed reflect severe glacial population decline and subsequent postglacial range expansion, the divergence observed between mainland and Tasmanian representatives of the ‘eastern’ clade almost certainly predates the LGM. This pre-LGM time frame of molecular divergence is also supported by the findings of Fraser et al. (2009b), who reported distinct D. antarctica haplotypes from ‘refugial’ islands in the New Zealand subantarctic. Despite the absence of reliable mutation rate estimates for brown algae (Hoarau et al. 2007) precluding precise dating of the time of D. potatorum recolonization, the near-absolute genetic homogeneity of the ‘western’ samples certainly seems consistent with a post-LGM timeframe. For the ‘eastern’ clade, by contrast, the presence of two apparently ‘ancestral’ haplotypes (Fig. 4), one on the mainland (found at both Tathra and Gabo Island) and one in southern Tasmania, suggests persistence of multiple eastern populations through the last glacial period, consistent with the warming influence of the EAC (hypothesis ii). To date, few marine phylogeographical studies have directly assessed the importance of palaeoceanographic features (but see Benzie 1999; Wares 2001). Although some previous studies have hinted at their role (e.g. Muss et al. 2001; Marko 2004), the current analysis represents, to our knowledge, one of the first studies to provide compelling evidence for lasting biological effects of Pleistocene oceanographic conditions.

Ecology and biology of D. potatorum Although this study assesses only genetic variation, a previous phylogenetic analysis of morphological and ecological data (Cheshire & Hallam 1989; Cheshire et al. 1995) has suggested a similar east–west split within D. potatorum based on morphological and ecological characteristics. The character states distinguishing these morphotypes were,

2294 C . I . F R A S E R , H . G . S P E N C E R and J . M . WAT E R S ‘western’ populations in northwestern Tasmania growing intertidally (J.M. Waters, personal observation), as well as intertidal plants at Tathra (‘eastern’ clade, C.I. Fraser, personal observation). Nonetheless, the dominance of ‘western’ individuals in our eastern Tasmanian collections, combined with the absence of any ‘western’ haplotypes from the subtidally collected samples at King George III Reef (southern Tasmania), suggests that the ‘eastern’ clade may grow lower in the littoral and sublittoral zones. Future studies will assess the genetic, morphological, ecological, and taxonomical characteristics of D. potatorum to rigorously test the hypothesis that it comprises two biologically distinct taxa. Ideally, future genetic analyses should also include a suite of variable markers (e.g. microsatellite loci) to further assess levels of gene flow among D. potatorum populations.

Acknowledgements Fig. 5 Photograph of Durvillaea potatorum retrieved from 10-m depth at King George III Reef by SCUBA, showing secondary stipes (s.s.) [photo: JMW].

however, somewhat variable (Table 1). Nonetheless, when combined with our phylogeographical results, the differentiation between ‘eastern’ and ‘western’ clades seems likely to have a morphological, as well as genetic, basis. Interestingly, several of the character states listed by Cheshire et al. (1995) as ‘infrequent’ in eastern populations, compared to entirely absent in western populations (Table 1), were indeed observed in our ‘eastern’ samples, such as secondary stipes emerging from the primary stipe (Fig. 5) and stipes longer than 0.5 m. Perhaps the apparent variability of these characteristics reflects the simultaneous presence of both ‘eastern’ and ‘western’ genetic clades along the east coast of Tasmania. If these clades correspond to biologically distinct taxa, Cheshire & Hallam (1989) are likely to have measured both sympatric forms in eastern Tasmanian. Further morphological and physiological work, combined with genetic identification, is clearly required to determine whether these two clades represent separate species, although their apparent morphological distinctness, and their cohabitation along eastern Tasmania, certainly suggests that they form separate taxonomic entities. Although both clades are clearly sympatric in eastern Tasmania, it is possible that their ecological characteristic may differ somewhat, as noted for cryptic Durvillaea taxa in New Zealand (Fraser et al. 2009a). Cheshire & Hallam (1989) claimed that the ‘western’ form of D. potatorum is never intertidal, compared to the ‘eastern’ form which is occasionally intertidal. In contrast, we observed some

We gratefully acknowledge the following organizations and individuals for additional sample collections: Lynda Avery and S. Beaton of Infauna Data for collections from Peterborough; Allan Millar for advice and a sample from Pirates Bay; Matt Edmunds for collections from Port Fairy and Warrnambool; Roz Jessup for collections from Phillip Island; Simon Talbot for SCUBA collections from Tasmania; Wayne Kelly, Barry Rumbold and Sue Jones (School of Zoology, University of Tasmania) for logistical support in Tasmania. Thanks to Raisa Nikula for comments on the manuscript; to Margaret Clayton and Cameron Hay for advice; and to Tania King for laboratory assistance. This work was funded by Department of Zoology funding allocations to C.I.F.; and University of Otago Research grants to J.M.W.

References Ayre DJ, Minchinton TE, Perrin C (2009) Does life history predict past and current connectivity for rocky intertidal invertebrates across a marine biogeographic barrier? Molecular Ecology (in press). Baines PG, Edwards RJ, Fandry CB (1983) Observations of a new baroclinic current along the western continental slope of Bass Strait. Australian Journal of Marine and Freshwater Research, 34, 155–157. Benzie JAH (1999) Genetic structure of coral reef organisms: ghosts of dispersal past. American Zoologist, 39, 131–145. Bernardi G, Findley L, Rocha-Olivares A (2003) Vicariance and dispersal across Baja California in disjunct marine fish populations. Evolution, 57, 1599–1609. Burridge CP (2000) Biogeographic history of cirrhitoids (Perciformes: Cirrhitoidea) with east–west allopatric distributions across southern Australia, based on molecular data. Global Ecology and Biogeography, 9, 517–525. Chenoweth SF, Hughes JM, Keenan CP, Lavery S (1998) When oceans meet: a teleost shows secondary intergradation at an Indian–Pacific interface. Proceedings of the Royal Society B: Biological Sciences, 265, 415–420. Cheshire AC, Hallam ND (1988) Morphology of the southern bull-kelp (Durvillaea potatorum, Durvilleales, Phaeophyta) from King Island (Bass Strait, Australia). Botanica Marina, 31, 139–148. © 2009 Blackwell Publishing Ltd

P O S T G L A C I A L R E C O L O N I Z AT I O N I N A U S T R A L I A N B U L L K E L P 2295 Cheshire AC, Hallam ND (1989) Morphological differences in the Southern bull-kelp (Durvillaea potatorum) throughout southEastern Australia. Botanica Marina, 32, 191–197. Cheshire AC, Conran JG, Hallam ND (1995) A cladistic analysis of the evolution and biogeography of Durvillaea (Phaeophyta). Journal of Phycology, 31, 644–655. Clayton MN, Ashburner CM (1994) Secretion of phenolic vesicles following fertilization in Durvillaea potatorum (Durvillaeales, Phaeophyta). European Journal of Phycology, 29, 1–9. Clement M, Posada D, Crandall KA (2000) tcs: a computer program to estimate gene genealogies. Molecular Ecology, 9, 1657–1660. Cresswell G (2000) Currents of the continental shelf and upper slope of Tasmania. Papers and Proceedings of the Royal Society of Tasmania, 133, 21–30. Dartnall AJ (1974) Littoral biogeography. In: Biogeography and Ecology in Tasmania (ed. Williams WD), pp. 171–194. Dr W. Junk, The Hague, The Netherlands. Darwin CR (1872) On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life, 6th edn. John Murray, London. Dawson MN (2005) Incipient speciation of Catostylus mosaicus (Scyphozoa, Rhizostomeae, Catostylidae), comparative phylogeography and biogeography in south-east Australia. Journal of Biogeography, 32, 515–533. Felsenstein J (1985) Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39, 783–791. Fraser CI, Hay CH, Spencer HG, Waters JM (2009a) Genetic and morphological analyses of the southern bull kelp Durvillaea Antarctica. (Phaeophyceae: Durvillaeales) in New Zealand reveal cryptic species. Journal of Phycology, 45, 436–443. Fraser CI, Nikula R, Spencer HG, Waters JM (2009b) Kelp genes reveal effects of subantarctic sea ice during the Last Glacial Maximum. Proceedings of the National Academy of Sciences, USA, 106, 3249–3253. Harvey WH (1863) Phycologica Australica, Vol. 5. Lovell Reeve, London. Hay CH (1994) Durvillaea (Bory). In: Biology of Economic Algae (ed. Akatsuka I), pp. 353–384. SPB Academic Publishing, The Hague, The Netherlands. Herbert TD, Schuffert JD, Andreasen D et al. (2001) Collapse of the California Current during glacial maxima linked to climate change on land. Science, 293, 71–76. Hewitt GM (1996) Some genetic consequences of ice ages, and their role in divergence and speciation. Biological Journal of the Linnean Society of London, 58, 247–276. Hewitt GM (2000) The genetic legacy of Quaternary ice ages. Nature, 405, 907–913. Hidas EZ, Costa TL, Ayre DJ, Minchinton TE (2007) Is the species composition of rocky intertidal invertebrates across a biogeographic barrier in south-eastern Australia related to their potential for dispersal? Marine and Freshwater Research, 58, 835–842. Hoarau G, Coyer JA, Veldsink JH, Stam WT, Olsen JL (2007) Glacial refugia and recolonization pathways in the brown seaweed Fucus serratus. Molecular Ecology, 16, 3606–3616. Huisman JM (2000) Marine Plants of Australia. Australian Biological Resources Study, University Of Western Australia Press, Canberra, Australia. Kirkman H, Kendrick GA (1997) Ecological significance and commercial harvesting of drifting and beach-cast macro-algae and seagrasses in Australia: a review. Journal of Applied Phycology, 9, 311–326. © 2009 Blackwell Publishing Ltd

Lambeck K, Chappell J (2001) Sea level change through the last glacial cycle. Science, 292, 679–686. Lane CE, Mayes C, Druehl LD, Saunders GW (2006) A multi-gene molecular investigation of the kelp (Laminariales, Phaeophyceae) supports substantial taxonomic re-organization. Journal of Phycology, 42, 493–512. Maggs CA, Castilho R, Foltz D et al. (2008) Evaluating signatures of glacial refugia for North Atlantic benthic marine taxa. Ecology, 89, S108–S122. Marko PB (2004) ‘What’s larvae got to do with it?’ Disparate patterns of post-glacial population structure in two benthic marine gastropods with identical dispersal potential. Molecular Ecology, 12, 597–611. McGowran B, Li Q, Cann J, Padley D, McKirdy DM, Shafik S (1997) Biogeographical impact of the Leeuwin Current in southern Australia since the late middle Eocene. Palaeogeography, Palaeoclimatology and Palaeoecology, 131, 19–40. Millar AJK (2007) The Flindersian and Peronian Provinces. In: Algae of Australia: Introduction (eds McCarthy PM, Orchard AE), pp. 554–559. ABRS Canberra: CSIRO Publishing, Melbourne, Australia. Muss A, Robertson R, Wirtz P, Bowen B, Stepien CA (2001) Phylogeography of the genus Ophioblennius: the role of ocean currents and geography in reef fish evolution. Evolution, 55, 561– 572. Nelson JS, Hoddell RJ, Chou LM, Chan WK, Wang VPE (2000) Phylogeographic structure of false clownfish, Amphiprion ocellaris, explained by sea-level changes on the Sunda shelf. Marine Biology, 137, 727–736. Nürnberg D, Brughmans N, Schönfeld J, Ninnemann U, Dullo C (2004) Paleo-export production, terrigenous flux and sea surface temperatures around Tasmania — implications for glacial/ interglacial changes in the Subtropical Convergence Zone. Geophysical Monograph Series, 151, 291–317. Papenfuss GF (1964) Catalogue and bibliography of Antarctic and Sub-Antarctic benthic marine algae. In: Bibliography of the Antarctic Seas, 1 (ed. Lee MO), pp. 1–76. American Geophysical Union, Washington, DC. Posada D, Crandall KA (1998) ModelTest: Testing the model of DNA substitution. Bioinformatics, 14, 817–818. Provan J, Bennett KD (2008) Phylogeographic insights into cryptic glacial refugia. Trends in Ecology & Evolution, 23, 564–571. Rambaut A, Drummond AJ (2007) Tracer, Version 1.4 (computer program). Available from URL: http://beast.bio.ed.ac.uk/Tracer Ramírez ME, Santelices B (1991) Catálogo de las algas marinas bentónicas de la costa temperada del Pacífico de Sudamérica. Monografías Biológicas, 5, 1–437. Reid DG, Lal K, Mackenzie-Dodds J, Kaligis F, Littlewood DTJ, Williams ST (2006) Comparative phylogeography and species boundaries in Echinolittorina snails in the central Indo-West Pacific. Journal of Biogeography, 33, 990–1006. Ridgway KR, Condie SA (2004) The 5500-km-long boundary flow off western and southern Australia. Journal of Geophysical Research, 109, C04017. Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19, 1572– 1574. Sanmartin I, Enghoff H, Ronquist F (2001) Patterns of animal dispersal, vicariance and diversification in the Holarctic. Biological Journal of the Linnean Society, 73, 345–390. Saunders GW (2005) Applying DNA barcoding to red macroalgae: a preliminary appraisal holds promise for future applications.

2296 C . I . F R A S E R , H . G . S P E N C E R and J . M . WAT E R S Philosophical Transanctions of the Royal Society of London. Series B, Biological Sciences, 360, 1879–1888. Spencer HG, Waters JM, Eichhorst TE (2007) Taxonomy and nomenclature of black nerites (Gastropoda: Neritimorpha: Nerita) from the South Pacific. Invertebrate Systematics, 21, 229–237. Tiffney BH (1985) The Eocene North Atlantic land bridge: its importance in Tertiary and modern phytogeography the Northern Hemisphere. Journal of the Arnold Arboretum, 66, 243–273. Voisin M, Engel CR, Viard F (2005) Differential shuffling of native genetic diversity across introduced regions in a brown alga: aquaculture vs. maritime traffic effects. Proceedings of the National Academy of Sciences, USA, 102, 5432–5437. Wares JP (2001) Biogeography of Asterias: North Atlantic climate change and speciation. Biological Bulletin, 201, 95–103. Waters JM (2008) Marine biogeographic disjunction in temperate Australia: historical landbridge, contemporary currents, or both? Diversity and Distributions, 14, 692–700. Waters JM, O’Loughlin PM, Roy MS (2004) Cladogenesis in a starfish species complex from southern Australia: evidence for vicariant speciation? Molecular Phylogenetics and Evolution, 32, 236–245. Waters JM, King TM, O’Loughlin PM, Spencer HG (2005) Phylogeographic disjunction in an abundant high-dispersal littoral gastropod. Molecular Ecology, 14, 2789–2802. Wells P, Okada H (1996) Holocene and Pleistocene glacial palaeoceanography off southeastern Australia, based on foraminifers and nannofossils in Vema cored hole V18–222. Australian Journal of Earth Science, 43, 509–523.

Wen J (1999) Evolution of eastern Asian and eastern North American disjunct distributions in flowering plants. Annual Review of Ecology and Systematics, 30, 421–455. York KL, Blacket MJ, Appleton BR (2008) The Bassian Isthmus and the major ocean currents of southeast Australia influence the phylogeography and population structure of a southern Australian intertidal barnacle Catomerus polymerus (Darwin). Molecular Ecology, 17, 1948–1961.

C.F. is a postgraduate student at the University of Otago. Her interests include the ecology and phylogeography of Southern Ocean marine biota, with current research focusing on southern bull kelps (Durvillaea). She is particularly interested in applying molecular techniques to broad-scale questions of dispersal and climate change. H.S. is Professor of Zoology at the University of Otago and a Principal Investigator with the Allan Wilson Centre for Molecular Ecology & Evolution. His research covers aspects of evolutionary biology, from theoretical population genetics and phenotypic plasticity to the phylogenetics of birds and the biogeography of molluscs, as well as the history of genetics. J.W. is an Associate Professor of Zoology at the University of Otago. His research focuses on the phylogeography and evolution of Southern Hemisphere marine and freshwater biota. Current projects are centered on the use of freshwater vicariant events to calibrate molecular clocks, and the importance of macroalgae as facilitators for oceanic rafting.

© 2009 Blackwell Publishing Ltd