Info

c o o o region were found in P. globosa even among cloned DNA from a single strain, suggesting that it is a species complex composed of at least three species. This observation was also supported by different DNA content among different clones of Phaeocystis (Vaulot et al. 1994). Multiple copies of the ITS1 were also found in a single strain of P. pouchetii, suggesting that this is also a species complex. These species complexes would appear still to be able to interbreed with one another because multiple different haplotypes from different clades can be found within a single strain. Similar findings of multiple haplotypes being present in known hybrids of flowering plant species are quite common (Chase et al 2003) and it is known that in some hybrids of flowering plants, the hybrid can loose one of the parental haplotypes in as few as 12 generations (Bateman, pers. com.) However, among nine P. antarctica strains, only one type of ITS1 haplotype was found per strain, although it was variable among the strains. Using the branching order in the ITS1 tree (Fig. 3) we have attempted to trace the biogeo-

graphic history of the dispersal of strains in Antarctic coastal waters. We see that the first divergence among the Antarctic clade is strain SK 22, which was isolated from the Antarctic circumpolar current (ACC). The second divergence is a strain from Prydz Bay (T4-2) and this is then followed by a nearly simultaneous divergence of the remaining strains. Among these divergences is a cluster of strains from Prydz Bay (DE2, A1-3 and T9-1). The other divergence includes strain CCMP 1374 from the Ross Sea and two Phaeocystis strains from the Weddell Sea (SK 20, SK 23) (Fig. 3) and finally strain D4-5 from Prydz Bay, which shares a last common ancestor with the Weddell Sea strain SK 23, as the final cluster to diverge among the Antarctic strains. Populations of P. antarctica within the continental boundary water masses appear to be well-mixed because currents move around the Antarctic continent rather quickly and may effectively act as a barrier to significant population structure. Strain SK 22 isolated within the ACC, however, is clearly different. An earlier hypothesis, proposed

Fig. 2 Maximum-likelihood phylogeny of the ITS regions showing the multiple sequences from a single strain of P. globosa and P. pouchetii. Each different sequence comes from a different bacterial vector clone. All haplo-types from one strain of Phaeocystis are connected by the arrows. P. antarctica exhibited a single sequence per strain and these are collapsed into a triangle. Redrawn from Lange et al. (2002)

from rDNA data (Medlin et al. 1994), that ancestral populations in the Antarctic gave rise to present day P. antarctica and P. pouchetii populations appears to be supported by ITS1 analysis of the cold-water Phaeocystis strains. P. antarctica and P. pouchetii, both polar, are more closely related to one another than either is to the cold and warm temperate to tropical populations of present-day P. globosa. This suggests that dispersal did not occur from present-day warm-water populations into present-day cold-water populations but that gene flow has occurred from pole to pole across tropical oceans. Arctic P. pouchetii populations thus probably arose by a dispersal event from the south to the north during colder climate periods that allowed populations to survive the crossing of equatorial waters, as has been documented for other organisms (Crame 1993; Darling et al. 2000, 2004; Montresor et al. 2003). A subsequent warm ing event will then isolate the two polar populations. Evidence for this can be found in a study of Antarctic surface-water temperatures since the Cretaceous (Crame 1993, Fig. 4c).

If we follow the branching order in Fig. 1, we hypothesize the following scenario: Phaeocystis likely originated as a warm-water genus because first divergences in our tree are warm-water species. Ancestral populations in the Antarctic were derived from ancestors of the present-day warm-water species, after being isolated in Antarctic waters. The opening of the Drake passage and the formation of the Antarctic circumpolar current (ACC) are the most likely geological events that could have isolated populations in the Antarctic to separate them from warm-water ancestors. It can be inferred from Fig. 3 that presumed descendants of these warm-water ancestors were first entrained in the ACC because these are the first

Fig. 3 (a) Locations of the strains of P. antarctica used in Lange et al. (2002). The location of different clades is indicated by the different patterns in the large circles and correspond to those clades in (b). Prydz Bay locations in E. Antarctica are slightly displaced for visual clarity. (b) Maximum-likelihood tree inferred from ITS1 sequences from P. antarctica with P. pouchetii as outgroup. Bootstrap values are placed at the nodes from a maximum likelihood analysis (100 replicates) a neighbor-joining analysis (500 replicates) and a maximum parsimony analysis (500 replicates). The scale bar corresponds to two changes per 100 nucleotide positions. Redrawn from Lange et al. (2002). Map of Antarctica redrawn from Olbers et al. (1962)

Fig. 3 (a) Locations of the strains of P. antarctica used in Lange et al. (2002). The location of different clades is indicated by the different patterns in the large circles and correspond to those clades in (b). Prydz Bay locations in E. Antarctica are slightly displaced for visual clarity. (b) Maximum-likelihood tree inferred from ITS1 sequences from P. antarctica with P. pouchetii as outgroup. Bootstrap values are placed at the nodes from a maximum likelihood analysis (100 replicates) a neighbor-joining analysis (500 replicates) and a maximum parsimony analysis (500 replicates). The scale bar corresponds to two changes per 100 nucleotide positions. Redrawn from Lange et al. (2002). Map of Antarctica redrawn from Olbers et al. (1962)

divergences. Some of these ancestral populations must have been transported northward and across the Equator shortly after the Drake passage opened because the P. pouchetii populations are sister to the P. antarctica populations. The ACC today encircles the Antarctic continent every 1-2 years. Water is entrained from this current into the major gyres of the continental water masses (Treshnikov 1964). Using the branching order in

Fig. 3 we can trace the dispersal of the clones from the ACC, although the bootstrap support for the branching order is weak to strong among the clades. The first entrainment with a bootstrap support of 99% appears to be into Prydz Bay because strain T4-2 isolated from this bay is the first divergence in our tree. These populations then established themselves in the Eastern Antarctic in Prydz Bay. Subsequent divergences in the tree indicate that populations were then entrained into the Ross Sea and almost simultaneously they were entrained into the Weddell Sea (bootstrap support 54%). Both isolates from the Weddell Sea were the last to diverge before the populations were again entrained back into Prydz Bay from populations in the Weddell Sea because isolates from this bay are some of the last divergences in the tree (bootstrap 54%). The distribution of these isolates in this fashion follows the predominant current patterns of surface waters in the Antarctic today. What we do not know is how diVerent the surface-water circulation was 30 Ma before the ACC was established.

Other studies have also shown the eVect of mixing on the homogenization of the genetic structure of Antarctic populations. Krill species within the Antarctic continental water masses are very similar as documented by both mtDNA (Patarnello et al. unpubl.) and isozyme analysis (Fevolden and Schneppenheim 1989). The mtDNA study also suggested that the formation of the ACC eVectively isolated krill species in Antarctic water masses from those north of the ACC. Calculation of the time of divergence between species groups found either side of the ACC coincided with the timing of the ACC, approximately 30 Ma. Thus, the molecular data is consistent with our hypothesized historical biogeographic reconstruction of the distribution of Phaeocystis based on the circulation patterns developed with the formation of the ACC.

Molecular clock

A molecular clock has been constructed from our 18S rDNA tree and calibrated with fossil dates from the haptophyte coccolithophorid species (Fig. 4). Our molecular clock calculations indicate

Mesocosm Project Ideas
0 0

Post a comment