Phylogenomic incongruence leads to different spatial phylogenetic patterns: an Australian continental scale case study with Isopogon and Petrophile (Proteaceae)

Spatial phylogenetics complement existing biodiversity metrics of species richness and endemism by incorporating evolutionary information from phylogenies with spatial distribution data. These complementary metrics of phylogenetic diversity (PD), phylogenetic endemism (PE), and categorical analysis of neo- and paleo-endemism (CANAPE) have largely been obtained from phylogenies constructed with a few genetic markers from different genomic datasets. Incongruent topologies may affect subsequent spatial phylogenetic results, leading to uncertainty for conservation outcomes. Here, we examined the effects of topological conflicts between plastid and nuclear phylogenies of two Australian plant genera – Isopogon R.Br. ex Knight and Petrophile R.Br. ex Knight (Proteaceae). We constructed the first densely sampled and generally well-resolved nuclear and plastid phylogenies of the two genera using a hybrid capture high throughput sequencing approach. We show that incongruent phylogenies from different molecular datasets have an important effect on downstream spatial phylogenetic analyses. Significant areas of PD and PE and neo- and paleo-endemism for Isopogon and Petrophile include south-west Western Australia, Kangaroo Island and the Sydney Basin. These areas are located along the mesic fringes of the Australian continent, congruent with the peripheral vicariance hypothesis in biogeography. However, these significant areas differ depending on whether the analyses were made with a nuclear or plastid topology. We caution against combining different datasets as is commonly done without detailed assessment of potential incongruence between datasets.