2011). The highest percentage of interspecific differentiation was attained with

the dam gene (2.6%), that also exhibited the highest level of intraspecific divergence within G. oceanica (1.5%, equal to the divergence within this morpho-species shown by petA despite a lower level of polymorphism (pi = 3.15 × 10−3 and 8.37 × 10−3 for dam and petA, respectively; Table 1). The dam gene also exhibited the highest selleck screening library intraspecific polymorphism within E. huxleyi (pi = 10.51 × 10−3, Table 1). Cox1 (short and long) exhibited the highest intraspecific polymorphism within G. oceanica (pi = 10.33 × 10−3 and 8.77 × 10−3, respectively; Table 1) with a lower level of polymorphism in E. huxleyi (pi = 5.15 × 10−3 and 4.90 × 10−3, respectively; Table 1). Cox3, rpl16 and dam all exhibited 0.8%–0.9% intraspecific variability within E. huxleyi, but the largest intraspecific find more divergences for this morpho-species were exhibited by the plastidial tufA (long) and petA markers (1.2% and 1.1% respectively; Table 1). With their lack or relatively low rate of nucleotide substitution, the 18S, 28S (nuclear), and 16S (plastidial) rDNA and the rbcL genes were not suitable for constructing phylogenies. Other markers exhibited a phylogenetic signal, in some cases by exclusively selecting parsimonious

informative sites. Overall, plastidial and mitochondrial markers generated partially congruent phylogenetic scenarios, with full monophyletic delineation of morpho-species only achieved with the mitochondrial markers (Fig. 1 and Figs. S3–S6 in the Supporting Information). For the plastidial markers, four statistically supported clades were defined the tufA topology, similar to the clades inferred in Cook et al.

(2011), while three clades were formed in the petA topology, but in both cases with a clear paraphyletic pattern, with G. oceanica strains partly distributed within E. huxleyi-dominated clusters check details (Fig. S3). In detail, the tufA GO clade (Fig. 1) is composed exclusively of G. oceanica strains, tufA I contains strains of E. huxleyi and G. oceanica corresponding to groups 3 and 5 defined by Cook et al. (2011), while tufA II and tufA III contain exclusively E. huxleyi and correspond, respectively, to group 1 and groups 2 and 4 of Cook et al. (2011). For both petA and tufA, the phylogenies did not correspond to geographical origin of strains or morpho-species delineation. By contrast, the five mitochondrial markers tested herein displayed consistent phylogenetic patterns with three statistically supported clades and clear morpho-species delineation. Clade γ (Fig. 1) exclusively contains G. oceanica strains and is highly diverse in cox1. Clades α and β contain the 84 E.