Revenant clades in historical biogeography: the geology of New Zealand predisposes endemic clades to root age shifts

Abstract

AbstractAimOur aim was to elucidate the effect of mass extinctions on inferred crown ages of terrestrial clades endemic to ancient islands. We thereby assessed the potential for mass extinction events, such as the Zealandian marine incursion episode in the Oligocene, to skew the interpretation of the evolutionary history of clades of various sizes.LocationSimulation study focusing on New Zealand.MethodsClades of various sizes were simulated under a birth–death model with variable parameters for diversification, invoking a prolonged extinction event, using the R packageTreeSim. We measured the scaled root shiftJbetween actual (i.e. first cladogenetic event) and inferred (i.e. most recent common ancestor of extant taxa) clade divergence times in simulated phylogenies incurred by extinction processes.ResultsA pulse of extinction followed by a prolonged period of low cladogenetic potential – modelled after the geological history of Zealandia in the Oligocene – produced large root shifts in clades of all sizes. Small clades with high net diversification rates were especially prone to belying pre‐Oligocene history. In simulations invoking mass extinctions wherein a pre‐extinction root was retained, phylogenies of extant taxa were characterized by anti‐sigmoidal log lineage‐through‐time plots that mimicked an upturn in diversification rate after the extinction period.Main conclusionsNon‐selective mass extinctions can engender large discrepancies in actual and inferred root ages, particularly in small, old clades. The evolutionary histories of lineages that survive mass extinctions are difficult to distinguish from scenarios of rapid radiation. This outcome challenges previous interpretations of post‐Oligocene crown ages for clades endemic to New Zealand as sufficient evidence for rejecting a pre‐Oligocene evolutionary history. As a corollary, our results suggest that the extant size of a particular clade is the foremost indicator of its potential for historical biogeographical inference. We therefore review the hypothesis of the total submersion of Zealandia, highlighting empirical cases of lineages with demonstrable pre‐Oligocene history that refute the hypothesis of total submersion.