Abstract
Pollinators with different vision are a key driver of flower coloration. Islands provide important insights into evolutionary processes, and previous work suggests islands may have restricted flower colors. Due to both species richness with high endemism in tropical–subtropical environments, and potentially changing pollinator distributions with altitude, we evaluated flower color diversity across the mountainous island of Taiwan in a comparative framework to understand the cause of color diversity. We sampled flower color signaling on the tropical–subtropical island of Taiwan considering altitudes from sea level to 3300 m to inform how over-dispersion, random processes or clustering may influence flower signaling. We employed a model of bee color space to plot loci from 727 species to enable direct comparisons to data sets from continental studies representing Northern and Southern Hemispheres, and also a continental mountain region. We observed that flower color diversity was similar to flowers that exist in mainland continental studies, and also showed evidence that flowers predominantly had evolved color signals that closely matched bee color preferences. At high altitudes floras tend to be phylogenetically clustered rather than over-dispersed, and their floral colors exhibited weak phylogenetic signal which is consistent with character displacement that facilitated the co-existence of related species. Overall flower color signaling on a tropical–subtropical island is mainly influenced by color preferences of key bee pollinators, a pattern consistent with continental studies.
Highlights
Floral color is a key functional trait that affects plant–pollinator interactions (Oberrath and Böhning Gaese, 1999; Bradshaw and Schemske, 2003; Ômura and Honda, 2005), contributing to biodiversity and structure of plant communities (Sargent and Ackerly, 2008; Schemske et al, 2009)
How flowers evolved color signals in different environments is important for understanding the plant–pollinator interactions that have occurred in the past and driven evolutionary processes (Fenster et al, 2004), and what might happen in the future with changing climatic conditions and/or pollinator distributions (Hegland et al, 2009; Shrestha et al, 2018)
We considered the uniquely placed island of Taiwan and used a hexagon color space model to plot flower loci and subsequently calculate floral color diversity (FCD) as the minimum convex polygon (MCP) encapsulating all loci
Summary
Floral color is a key functional trait that affects plant–pollinator interactions (Oberrath and Böhning Gaese, 1999; Bradshaw and Schemske, 2003; Ômura and Honda, 2005), contributing to biodiversity and structure of plant communities (Sargent and Ackerly, 2008; Schemske et al, 2009). Even on larger oceanic islands, floras in New Zealand (NZ) mountains consist of constrained flower color signals, exhibiting much lower FCD (Bischoff et al, 2013b) when compared to its adjacent continental island of Australia (Webb and Kelly, 1993; Dyer et al, 2012). It’s worth noting that the low species richness of these oceanic islands, e.g., MI (37 angiosperm species, de Salas and Baker (2015) and JFI (152 species), may more or less determine their low FCD, as plant communities with larger species richness would often, but not always, have greater FCD (Arnold et al, 2009b; Shrestha et al, 2014; Bergamo et al, 2018; Gray et al, 2018)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
