Phylogenetics and biogeography of the olive family (Oleaceae)

Abstract

Background and aims: Progress in the systematic studies of the olive family (Oleaceae) during the last two decades provides the opportunity to update its backbone phylogeny and to investigate its historical biogeography. We also aimed to understand the factors underlying the disjunct distribution pattern between East Asia and both West Asia and Europe that is found more commonly in this family than in any other woody plant family.

Methods: Using a sampling of 298 species out of ~750, the largest in a phylogenetic study of Oleaceae thus far, with a set of 36 plastid and nuclear markers, we reconstructed and dated a new phylogenetic tree based on maximum likelihood and Bayesian methods and checked for any reticulation events. We also assessed the relative support of four competing hypotheses [Qinghai-Tibet Plateau uplift (QTP-only hypothesis); climatic fluctuations (climate-only hypothesis); combined effects of QTP uplift and climate (QTP-climate hypothesis); and no effects (null hypothesis)] in explaining these disjunct distributions.

Key results: We recovered all tribes and subtribes within Oleaceae as monophyletic, but uncertainty in the position of tribe Forsythieae remains. Based on this dataset, no reticulation event was detected. Our biogeographical analyses support the QTP-climate hypothesis as the likely main explanation for the East-West Eurasian disjunctions in Oleaceae. Our results also show an earlier origin of Oleaceae at ~86 Mya and the role of Tropical Asia as a main source of species dispersals.

Conclusion: Our new family-wide and extensive phylogenetic tree highlights both the stable relationships within Oleaceae, including the polyphyly of the genus Chionanthus, and the need for further systematic studies within the largest and most undersampled genera of the family (Chionanthus and Jasminum). Increased sampling will also help to fine-tune biogeographical analyses across spatial scales and geological times.

Keywords: Chionanthus; Jasminum; Climatic fluctuations; Eurasia; Oleaceae; Qinghai–Tibet Plateau; Tropical Asia; disjunction; divergence time; historical biogeography; phylogenetics; reticulation.

Fig. 1.

Representatives of tribes and subtribes in Oleaceae. Tribe Fontanesieae: (A) Fontanesia fortunei. Tribe Forsythieae: (B) Forsythia intermedia. Tribe Jasmineae: (C) Jasminum kitchingii (©Anton Sieder, CC BY-NC-ND 3.0). Tribe Myxopyreae: (D) Nyctanthes arbor-tristis (©Varun Pabrai, CC BY-SA 4.0). Tribe Oleeae, subtribe Fraxininae: (E) Fraxinus pennsylvanica (©Gerrit Davidse, CC BY-NC-SA 3.0); subtribe Ligustrinae: (F) Syringa vulgaris; subtribe Oleinae: (G) Chionanthus virginicus, (H) Olea europaea (©Nefronus, CC BY-SA 4.0) and (I) Osmanthus fragrans (©Junichiro Aoyama, CC BY 2.0); subtribe Schreberinae: (J) Schrebera minor. Images without copyright licences listed were provided by authors.

Fig. 2.

Simplified topology of the best maximum likelihood tree of Oleaceae and outgroups. This tree results from an analysis of a concatenated dataset of plastid and nuclear regions. Tribe names are in bold. For the full tree, where all accessions are shown, refer to Supplementary Data Fig. S1. For a tree with information on node support, see Supplementary Data Fig. S2.

Fig. 3.

Maximum likelihood ancestral range estimation in Oleaceae, using the best model, DEC+j (model 14 in Supplementary Data Table S3). This tree is a simplified version of our 402-tip tree (Supplementary Data Fig. S3), and it represents tribes and subtribes within Oleaceae. The pie charts at nodes show the relative probability of the possible states (ranges). The boxes on the left indicate the most common ranges depicted in this tree.

Fig. 4.

Summary of dispersal events estimated with biogeographical stochastic mapping in the history of Oleaceae. Contoured areas are the same ones as shown in Fig. 3. The two areas with a bold black contour, Tropical and Temperate Asia, are the most species rich today, and events from those areas are highlighted in pink. The arrows between areas represent the direction and frequency of dispersal events. Only event counts that presented a mean of ≥0.95 (Supplementary Data Table S4) are depicted as arrows here; arrow line thickness is proportional to the mean number of events.

Fig. 5.

Timing and frequency of estimated extinction, vicariance and founder events that are likely to have contributed to disjunctions between Asia and Europe. For the sake of comparison, we show the frequency of such events in the entire family, and the frequency only in the clades with East–West Eurasian disjunctions. Time is represented in five categories: 88–65, 65–33.9, 33.9–20 and 20–5.3 Mya, and 5.3 Mya to present time.