Floral uniformity through evolutionary time in a species‐rich tree lineage

Abstract

Changes in floral morphology are expected across evolutionary time and are often promoted as important drivers in angiosperm diversification. Such a statement, however, is in contrast to empirical observations of species-rich lineages that show apparent conservative floral morphologies even under strong selective pressure to change from their environments.Here, we provide quantitative evidence for prolific speciation despite uniform floral morphology in a tropical species-rich tree lineage. We analyse floral disparity in the environmental and phylogenetic context of Myrcia (Myrtaceae), one of the most diverse and abundant tree genera in Neotropical biomes.Variation in floral morphology among Myrcia clades is exceptionally low, even among distantly related species. Discrete floral specialisations do occur, but these are few, present low phylogenetic signal, have no strong correlation with abiotic factors, and do not affect overall macroevolutionary dynamics in the lineage.Results show that floral form and function may be conserved over large evolutionary time scales even in environments full of opportunities for ecological interactions and niche specialisation. Species accumulation in diverse lineages with uniform flowers apparently does not result from shifts in pollination strategies, but from speciation mechanisms that involve other, nonfloral plant traits.

A stable pollination environment limits current but not potential evolution of floral traits

“…Note that selection on floral traits is not absolutely restricted to pollinators, and herbivores and abiotic factors can also be agents of selection (reviewed by Caruso et al 2018). Regardless of the source of selection, our findings contribute to explain the macroevolutionary patterns of floral evolution where novel phenotypes are ubiquitous (exceptions are often related to very generalised pollination that is stable over evolutionary time, see Vasconcelos et al 2019). Populations can experience stable conditions with undetectable directional selection, but at the same time harbour genetically based variability to evolve under new conditions.…”

Schrödinger's phenotypes: Herbarium specimens show two‐dimensional images are both good and (not so) bad sources of morphological data

“…These differences could bias biological interpretations (Yezerinac et al, 1992) of studies relying on acquisition of morphological data, such as phenotypic evolution, disparity, modularity, phylogenetic inference and others (e.g. Castiglione et al., 2019; Catalano, Goloboff, & Giannini, 2010; Dellinger et al., 2019; Gallaher et al., 2019; Guillerme & Cooper, 2018; Klingenberg & Gidaszewski, 2010; Vasconcelos et al., 2018; Wiens, 2003; Zaragüeta‐Bagils & Bourdon, 2007).…”

Schrödinger’s phenotypes: herbarium specimens show two-dimensional images are both good and (not so) bad sources of morphological data

“…These differences could bias biological interpretations (Yezerinac et al, 1992) of studies relying on acquisition of morphological data, such as phenotypic evolution, disparity, modularity, phylogenetic inference, and others (e.g. Wiens, 2003; Zaragüeta-Bagils and Bourdon, 2007; Catalano et al, 2010; Klingenberg and Gidaszewski, 2010; Vasconcelos et al, 2018; Castiglione et al, 2019; Dellinger et al, 2019; Gallaher et al, 2019; Guillerme and Cooper, 2018).…”