An integrative study of species distribution modelling and conservation genetics: Magnolia in Hispaniola

Abstract

On the island of Hispaniola, five endemic species of Magnolia occur, all of which are threatened with extinction. Little is known about their distribution and genetic health, hampering targeted conservation actions. The objective of this study is to assess the potential distribution and the genetic health of the Magnolias of Hispaniola, to inform concrete guidelines for effective conservation management. Using species distribution modelling (SDM), we predict habitat suitability for the Magnolias of Hispaniola by analysing 21 variables, describing climate and landscape features, on 635 occurrences. We genotyped 417 individuals using 16 microsatellite markers, to test for genetic structure and degree of inbreeding. The SDM and genetic data confirm the recognition of the four studied Magnolia species. The known individuals of the three Dominican Magnolias are structured into five populations which show ample genetic diversity and little inbreeding overall. For conservation management, we propose to focus on exploration using the SDM results, and protection and reinforcement using the genetic and occurrence data. The genetic results guide prioritization of species and populations. The SDM results guide spatial prioritization. Installing and/or protecting habitat corridors between populations, starting with the two species with the lowest genetic diversity and relatively nearby populations, is recommended as a durable conservation strategy. Meanwhile, reinforcement efforts can be undertaken to artificially increase gene flow for which we appoint sink and source population pairs, using the genetic data.