A new phylogeny for the genus Picea from plastid, mitochondrial, and nuclear sequences.

Studies over the past ten years have shown that the crown groups of most conifer genera are only about 15-25 Ma old. The genus Picea (spruces, Pinaceae), with around 35 species, appears to be no exception. In addition, molecular studies of co-existing spruce species have demonstrated frequent introgression. Perhaps not surprisingly therefore previous phylogenetic studies of species relationships in Picea, based mostly on plastid sequences, suffered from poor statistical support. We therefore generated mitochondrial, nuclear, and further plastid DNA sequences from carefully sourced material, striking a balance between alignability with outgroups and phylogenetic signal content. Motif duplications in mitochondrial introns were treated as characters in a stochastic Dollo model; molecular clock models were calibrated with fossils; and ancestral ranges were inferred under maximum likelihood. In agreement with previous findings, Picea diverged from its sister clade 180 million years ago (Ma), and the most recent common ancestor of today's spruces dates to 28 Ma. Different from previous analyses though, we find a large Asian clade, an American clade, and a Eurasian clade. Two expansions occurred from Asia to North America and several between Asia and Europe. Chinese P. brachytyla, American P. engelmannii, and Norway spruce, P. abies, are not monophyletic, and North America has ten, not eight species. Divergence times imply that Pleistocene refugia are unlikely to be the full explanation for the relationships between the European species and their East Asian relatives. Thus, northern Norway spruce may be part of an Asian species complex that diverged from the southern Norway spruce lineage in the Upper Miocene, some 6 Ma, which can explain the deep genetic gap noted in phylogeographic studies of Norway spruce. The large effective population sizes of spruces, and incomplete lineage sorting during speciation, mean that the interspecific relationships within each of the geographic clades require further studies, especially based on genomic information and population genetic data.