MOLECULAR PHYLOGENETICS AND EVOLUTION Molecular Phylogenetics and Evolution 30 (2004) 251–257 www.elsevier.com/locate/ympev Short communication HuxleyÕs line demarcates extensive genetic divergence between eastern and western forms of the giant freshwater prawn, Macrobrachium rosenbergii Mark de Bruyn,* John A. Wilson, and Peter B. Mather School of Natural Resource Sciences, Queensland University of Technology, GPO Box 2434, Brisbane, Qld 4001, Australia Received 5 February 2003; revised 16 April 2003 Abstract Phylogenetic analysis of representatives from 18 wild populations of the giant freshwater prawn, Macrobrachium rosenbergii, utilising a fragment of the 16S rRNA mitochondrial gene, identified two major reciprocally monophyletic clades either side of a well- known biogeographic barrier, HuxleyÕs line. The level of divergence between the two clades (maximum 6.2%) far exceeds divergence levels within either clade (maximum 0.9%), and does not concord with geographical distance among sites. ÔEasternÕ and ÔwesternÕ M. rosenbergii clades have probably been separated since Miocene times. Within-clade diversity appears to have been shaped by dispersal events influenced by eustatic change. Ó 2003 Elsevier Science (USA). All rights reserved. Keywords: Macrobrachium rosenbergii; Decapod crustacean; 16S; mtDNA; HuxleyÕs line; WallaceÕs line; Phylogenetics; Biogeography 1. Introduction south to Papua New Guinea and northern Australia. Gravid females migrate from freshwater to estuarine Prawns of the genus Macrobrachium Bate, 1868 areas, satisfying larval requirements for brackish water (Crustacea: Palaemonidae) are a highly diverse group of for survival and early development, where free-swim- decapod crustaceans found in circumtropical marine-, ming larvae hatch and metamorphose into post-larvae, estuarine-, and fresh-waters. Much debate has sur- before migrating to freshwater after 3–6 weeks (New rounded the systematic relationships of many species and Singholka, 1985). Several studies have reared within this group (e.g., Holthuis, 1950, 1995; Johnson, M. rosenbergii larvae to post-larvae stage in artificial 1973; Pereira, 1997), which has until recently been based seawater (Sandifer and Smith, 1979; Smith et al., 1976), exclusively on comparisons of external morphological and considering this in light of a relatively prolonged characteristics. Molecular genetic approaches to re- larval duration, suggests that marine dispersal may play solving systematic questions in Macrobrachium have a previously unrecognised role in the life-history of this only been applied recently, when Murphy and Austin species. (2002) recognised that species and genus level designa- Two forms of M. rosenbergii (ÔeasternÕ and ÔwesternÕ) tions did not correspond to traditional morphology- have been described independently (De Man, 1879; based classification schemes. Johnson, 1973), although the species is currently con- Macrobrachium rosenbergii, the giant freshwater sidered to be monophyletic. Lindenfelser (1984) anal- prawn, is found in coastal river systems from Pakistan in ysed morphometric and allozyme data, and concluded the west to Vietnam in the east, across SE Asia, and that the boundary for eastern and western M. rosen- bergii forms correspond approximately with WallaceÕs line (although Philippine samples were assigned to the * Corresponding author. Fax: +61-7-3864-1535. eastern form, thus HuxleyÕs line would seem a more E-mail address: [email protected] (M. de Bruyn). appropriate boundary than WallaceÕs line; see Fig. 1). 1055-7903/$ - see front matter Ó 2003 Elsevier Science (USA). All rights reserved. doi:10.1016/S1055-7903(03)00176-3 252 M. de Bruyn et al. / Molecular Phylogenetics and Evolution 30 (2004) 251–257 Fig. 1. Locations sampled for Macrobrachium rosenbergii. Light grey shading indicates—120 m sea-level contour. Pleistocene drainage basins indi- cated on map (Voris, 2000). Haplotype labels correspond to Appendix A. (Map adapted with kind permission Harold K. Voris and the Field Museum of Natural History, Chicago, USA; Voris, 2000.) Malecha (1977, 1987) and co-workers (Hedgecock ium australiense and Macrobrachium lar were used as et al., 1979) recognised three Ôgeographical racesÕ;an outgroup taxa. Tissue samples were incubated overnight eastern, a western, and an Australian Ôrace,Õ based on at 55 °C in 500 ll extraction buffer (100 mM NaCl, allozyme and morphological data. Wowor and Ng 50 mM Tris, 10 mM EDTA, and 0.5% SDS) containing (2001) regard the eastern and western forms of M. 20 llof10lg/ll Proteinase K (Sigma). Total genomic rosenbergii as two distinct species, based on adult DNA was extracted using standard phenol:chloroform morphological characters. Thus, M. rosenbergii as extraction methods. A 472-bp region of the mitochon- currently recognised taxonomically may be polytypic drial 16S ribosomal gene was amplified using primers both regionally and perhaps even within biogeographic 16SAR and 16SBR (Palumbi et al., 1991). DNA se- regions. Hence, the goal of the present study was to quencing was conducted at the Australian Genome examine the evolutionary relationships among wild M. Research Facility, Brisbane, Australia; using an ABI 377 rosenbergii stocks at a regional scale, using 16S ribo- automated DNA sequencer. Both strands of the PCR somal RNA mitochondrial DNA (mtDNA) sequences, product were sequenced. Because mtDNA sequences and relate the findings to the biogeographical history were invariant among five individuals from each of four of the region. sampling sites (Mekong and Dongnai, Vietnam; Wen- lock, Australia; Plandez/Pulilan, Philippines; de Bruyn et al., unpublished data), a single sequence from each 2. Materials and methods sampling site was considered to be representative for phylogenetic analyses. 2.1. Specimens, DNA extraction, amplification, and sequencing 2.2. Phylogenetic analyses Prawns used in this study were collected from local- Consensus sequences were aligned using ClustalX ities indicated in Appendix A and Fig. 1. Macrobrach- (Thompson et al., 1997). A total of 472 bp were aligned M. de Bruyn et al. / Molecular Phylogenetics and Evolution 30 (2004) 251–257 253 for analysis (see Appendix A for GenBank accession ern and western haplotypes observed indicates the coa- numbers). Saturation of nucleotide substitutions in the lescence for these two clades was probably of mid to late dataset was tested. A bootstrapped (1000 pseudorepli- Miocene origin, and approximates 5.3–11.7 million cates) maximum parsimony (MP) and neighbour-joining years before present (BP), based on 16S rRNA molec- (NJ) phylogeny was constructed using MEGA version ular clocks calibrated for porcelain crabs (0.53%/MY; 2.1 (Kumar et al., 2001), based on Kimura 2-parameter Stillman and Reeb, 2001) and fiddler crabs (0.96%/MY; distances (Kimura, 1980). A quartet-puzzling maximum- Sturmbauer et al., 1996; these values represent upper- likelihood tree using the Hasegawa–Kishino–Yano and lower-bound extremes for crustacean 16S rRNA (HKY) sequence evolution model (Hasegawa et al., molecular clocks identified in a literature search). This 1985) was constructed in TREE-PUZZLE (Strimmer estimate should be approached with caution, however, and von Haesler, 1996), using 1000 iterations of the due to the rejection of clock-like behaviour of the puzzling process. Finally, a log-likelihood ratio test was dataset. carried out in TREE-PUZZLE that compared trees WallaceÕs line has long been recognised as a major generated under the assumption of a molecular clock, to biogeographical barrier. Huxley (1868) modified Wal- trees unconstrained by any such assumption (Felsen- laceÕs line by extending it into the Philippines, based on stein, 1988). zoological data, linking the island of Palawan to the western (Oriental) group, and the rest of the Philippine Archipelago to the eastern (Australasian) group. Data presented here clearly links a region of the Philippines 3. Results (Luzon) to the eastern group. Tree topology indicates that the Australian OTUs (Operational Taxonomic A total of 472 bp of the 16S mitochondrial gene Units) are basal to the remaining eastern OTUs exam- were amplified successfully for 18 M. rosenbergii indi- ined. The unexpectedly low degree of divergence (1– viduals and two outgroup species. Of these, 90 variable 2 bp) between the Philippine OTU and the rest of the sites were detected, of which 59 were phylogenetically eastern OTUs suggests recent gene flow has occurred. informative. All sequences were found to be AT-rich This has presumably been facilitated by larval marine (62.9%). Nucleotide substitutions (excluding out- dispersal, as the Philippine and Australian/New Guinea groups) favoured transitions over transversions, yield- landmasses have been geographically distant since at ing a transition/transversion ratio of 3.3. No evidence least Miocene times (Hall, 1996). Tree topology indi- of saturation was evident. Kimura 2-parameter se- cates that gene flow has occurred from a southerly quence divergences ranged from 5.1 to 6.2% between (Australian) to northerly (Philippines) direction, con- haplotypes from eastern and western M. rosenbergii sistent with major ocean current movements in the re- samples, 0.0–0.6% among western samples, and 0.0– gion (South Equatorial Current; Gordon and Fine, 0.9% among eastern samples. A single deletion was 1996), although this remains to be rigorously tested with observed in the dataset, for the M. australiense out- a more comprehensive dataset. Similar genetic signa- group sequence. The log-likelihood