Limited by the roof of the world: mountain radiations of Apollo swallowtails controlled by diversity-dependence processes

Abstract

Mountainous areas contain a substantial part of the world species richness, but the evolutionary origins and diversification of this biodiversity remain elusive. Diversification may result from differences in clade age (longer time to diversify), net diversification rates (faster speciation rate) or carrying capacities (number of niches). The likelihood of these macroevolutionary scenarios was assessed for six clades of Apollo swallowtails () that diversified mainly in the Himalayan-Tibetan region. The analyses suggest that neither the clade age nor the speciation rate could explain the mountain butterfly diversification. Instead diversity-dependence models were strongly supported for each group. Models further estimated clades' carrying capacities, which approximate to the current number of species, indicating that diversity equilibrium has been reached (or close to being reached). The results suggest that diversification of mountain butterflies was controlled by ecological limits, which governed the number of niches, and provide macroevolutionary justification for regarding mountains as islands.

Spatiotemporal Differentiation of Alpine Butterfly Parnassius glacialis (Papilionidae: Parnassiinae) in China: Evidence from Mitochondrial DNA and Nuclear Single Nucleotide Polymorphisms

“…This is believed to ensure the success of the male by preventing other males from mating, and therefore avoiding sperm competition [19]. Parnassius has diversified into about 60 species, mostly occurring in the Himalayan-Tibetan mountains [20,21]. In China, only one species, Parnassius glacialis, has dispersed into areas south of the Yangtze River, inhabiting lower-altitude mountains (~200 to 2000 m).…”

Parnassius nebrodensis: A threatened but neglected Apollo butterfly species from Southern Europe (Lepidoptera: Papilionidae)

“…In particular, the samples of P. mnemosyne collected in the Western Mediterranean Region and Southern Anatolia and Iran were found to be distant from those sampled through Northern and Eastern Europe and Central Asia (Figs 1-2). Recently, the existence of these specieslevel clades was confirmed using multi-locus time-calibrated phylogenies and species delimitation modeling (Condamine 2018;Condamine et al 2018). It was shown that the Western Mediterranean lineage (=P.…”

“…and Dicentra Bernh. (Papaveraceae) taxa as larval host plants (Korshunov 2002;Michel et al 2008;Condamine 2018). Most species in this group are known to occur in Northern Asia, Central Asia, and North America (Weiss 1999;Korshunov 2002).…”

“…It was shown that the Western Mediterranean lineage (=P. mnemosyne sp.3 sensu Condamine, 2018) was isolated from other populations of P. mnemosyne since the mid-Pliocene (Condamine 2018). Although the evaluation of cryptic species is a task of great importance to resolve a broad array of scientific, conservation, and management issues (Bickford et al 2007;Dincă et al 2011Dincă et al , 2015Platania et al 2020a), these Clouded Apollo lineages are yet to be studied by means of morphological and taxonomic approaches.…”

The origin and drivers of Neotropical plant and tetrapod diversification

“…The most relevant of such models are diversity-dependent models where the number of extant species in a clade affects diversification rates (Etienne et al2012). We did not fit these models for three additional reasons:(i) it is not straightforward to compare diversity-dependent models with other diversification models on the basis of AICc (Etienneet al 2016); (ii) the sampling scheme currently implemented in diversity-dependent models assumes that exactly nspecies are sampled (n -sampling), while other models assume that each species is sampled with a fixed probability (ρ-sampling), and likelihoods associated to these two sampling schemes are not directly comparable (Stadler 2009;Lambert 2017); and (iii) previous simulations have shown that these models tend to over-fit the data when speciation decreases toward the present regardless the cause of decline (Etienne et al 2016;Condamine 2018). This latter artifact might affect more than 78% of clades in our study, since we found most phylogenies are characterized by constant or decreasing rates through time (see Results ).…”