Long‐term in situ persistence of biodiversity in tropical sky islands revealed by landscape genomics

Abstract

Tropical mountains are areas of high species richness and endemism. Two historical phenomena may have contributed to this: (i) fragmentation and isolation of habitats may have promoted the genetic differentiation of populations and increased the possibility of allopatric divergence and speciation and (ii) the mountain areas may have allowed long-term population persistence during global climate fluctuations. These two phenomena have been studied using either species occurrence data or estimating species divergence times. However, only few studies have used intraspecific genetic data to analyse the mechanisms by which endemism may emerge at the microevolutionary scale. Here, we use landscape analysis of genomic SNP data sampled from two high-elevation plant species from an archipelago of tropical sky islands (the Trans-Mexican Volcanic Belt) to test for population genetic differentiation, synchronous demographic changes and habitat persistence. We show that genetic differentiation can be explained by the degree of glacial habitat connectivity among mountains and that mountains have facilitated the persistence of populations throughout glacial/interglacial cycles. Our results support the ongoing role of tropical mountains as cradles for biodiversity by uncovering cryptic differentiation and limits to gene flow.

Snapshot isolation and isolation history challenge the analogy between mountains and islands used to understand endemism

“…"Darwinian islands" sensu Gillespie & Roderick, 2002) or was TA B L E 2 Effect of Quaternary climate fluctuations on inter-island connectivity and present-day endemism in relationship to the type of archipelago in true islands and mountain islands (Flantua et al, 2019;Simpson, 1974) Galapagos (Ali & Aitchison, 2014) Trans-Mexican Volcanic Belt (Mastretta-Yanes et al, 2018) Large, distant Low degree of inter-island connectivity Mainly changes in surface area Pantepuis (Rull & Nogué, 2007); East African rift (Chala et al, 2017;Sklenář et al, 2014) Ryukyu islands (Wepfer, Guénard, & Economo, 2016); Azores (Rijsdijk et al, 2014) previously part of another landmass from which it separated (i.e., initially not isolated, "fragment islands" sensu Gillespie & Roderick, 2002) is important for understanding the patterns of endemism (Sondaar & Van der Geer, 2005). When the initial level of isolation is high and persists throughout history, evolution has a limited set of lineages to work on.…”

Climate change, extinction, and Sky Island biogeography in a montane lizard

“…Some previous studies (mostly in other regions) have also combined phylogeography with niche modelling to address the impacts of past climate change (e.g., Carstens & Richards, ; Carnaval, Hickerson, Haddad, Rodrigues, & Moritz, ), including some that focused on montane species (Bryson, Murphy, Graham, Lathrop, & Lazcano, ; Gutierrez‐Tapia & Palma, ; Mastretta‐Yanes et al, ) and some that discussed implications for future climate change (Cordellier & Pfenninger, ). Here, we integrate these two approaches (niche modelling, phylogeography) with analyses of distributional, climatic, and physiological data to address the causes of Sky Island distributions and their implications for future climate change.…”

Examining the interglacial high‐elevation refugia scenario in East Asian subtropical mountain systems with the frog species Leptobrachium liui

“…Change in the genetic differentiation between populations often requires long time periods (Ibrahim, Nichols, & Hewitt, 1996;Leng & Zhang, 2011;Palo et al, 2004). It is unlikely that an IBD pattern established over long periods of a continuous distribution (Slatkin, 1993;Wright, 1943) would have been erased by the increased genetic differentiation during the current interglacial of ~0.01 million years (Mastretta-Yanes et al, 2018). In theory, the effect of such increased genetic differentiation on the Mantel correlation test for IBD may be analogous to adding a constant to the genetic-distance variable, which will not affect the detection of a correlation.…”