Canary Grasses (Phalaris, Poaceae): biogeography, molecular dating and the role of floret structure in dispersal

Authors


Abstract

Canary grasses (Phalaris, Poaceae) include 21 species, widely spread throughout the temperate and subtropical regions of the world with two centres of diversity: the Mediterranean Basin and western North America. The genus contains annual and perennial, endemic, cosmopolitan, wild, and invasive species with diploid, tetraploid and hexaploid cytotypes. As such, Phalaris presents an ideal platform to study diversification via historic hybridization and polyploidy events, and geographical dispersal in grasses. We present the first empirical phylogeographic study for Phalaris testing current, intuitive hypotheses on the centres of origin, historic dispersal events and diversification within a geological timeframe. Bayesian methods (beast, version 1.6.2) were used to establish divergence dates, and dispersal–vicariance analyses (rasp, version 2.1b) were implemented for ancestral node reconstructions. Our phylogeographic results indicate that the genus emerged during the Miocene epoch [20.6–8.4 Ma (million years ago)] in the Mediterranean basin followed by dispersal and vicariance events to Africa, Asia and the Americas. We propose that a diploid ancestor of P. arundinacea migrated to western North America via the Bering Strait, where further diversification emerged in the New World. It appears that polyploidy played a major role in the evolution of the genus in the Old World, while diversification in the New World followed a primarily diploid pathway. Dispersal to various parts of the Americas followed different routes. Fertile florets with hairy protruding sterile lemmas showed significant correlation with wider geographical distribution.

Ancillary