Are pollinators and seed predators selective agents on flower color in Gentiana lutea?

Abstract

Animals which interact with plants often cause selective pressures on plant traits. Flower color variation within a species might be shaped by the action of animals feeding on the plant species. Pollinators might exert natural selection on color if flower color is related to their foraging efficiency. For example, some pollinator species might require more time to detect particular colors. If that is the case, flower color might have evolved as a pollination exploitation barrier-ensuring that flowers are more visited by the most efficient pollinators. In addition, non-pollinator agents such as predispersal seed predators may select on flower color, if color indicates food resources (seeds) or if color is related to deterrent compounds. We address selection on flower color in a population of Gentiana lutea where color varies among individuals from yellow to orange. We hypothesize that opposed selection from mutualists (pollinators) and antagonists (predispersal seed predators) maintains flower color variation in this population. By means of path analysis we addressed the role of both interactors in flower color selection. We found that selection acts on flower color, mediated by both pollinators and seed predators. Both agents favored yellow-flowered individuals, thus selection by pollinators and seed predators does not maintain flower color variation in this population.

Flower colour variation in the montane plantGentiana luteaL. (Gentianaceae) is unrelated to abiotic factors

“…If UV variation is unlinked to orange-yellow variation, we may describe colour related to the visible spectra. We did not detect any differences in the UV reflectance between orange-and yellow-flowered plants nor among populations (see Veiga et al 2015). So, we described colour considering the visible range of the spectrum.…”

“…In those cases, colour may indicate toxicity and herbivores may avoid certain colours. We found that seed predators avoided yellow colouration (Veiga et al 2015); thus, it might occur that colour variation relates to toxic compounds and may be the result of selective pressures exerted by herbivores.…”

“…Instead variation may be related to other abiotic factors we have not measured, such as soil/bedrock characteristics or soil nutrients. It could also be explained by natural selection exerted by other forces such as biotic agents (see Veiga et al 2015;Sobral et al In press) or by random processes such as genetic drift or founder effects. Biotic selective pressures on flower colour might be exerted, for instance, by pollinators or herbivores (Irwin and Strauss 2005;Strauss and Whittall 2006;Streisfeld and Kohn 2007;Rausher 2008;Renoult et al 2013).…”

“…This does not mean they cannot detect red colours, but, they may have a reduced ability to detect orange-flowered plants (Rodríguez-Gironés and Santamaría 2004). In fact, recent findings suggest that pollinators exert selective pressures on colour in G. lutea (Veiga et al 2015). The pollinator assemblage varies across the Cantabrian Mountains (Obeso 1992;Ploquin et al 2013), and the spatial variation of pollinators seems to drive part of the geographical structure of flower colour variation in G. lutea (Sobral et al In press).…”

“…Pelargonidin provides orange colour and without its accumulation flowers lack the orange colouration. Additionally, pollinators exert selection on flower colour in this species (Veiga et al 2015) and differences in pollinator community composition are related to variation in flower colour among populations (Sobral et al In press). However, it is not known whether flower colour is also selected by abiotic factors.…”