Sex‐specific floral attraction traits in a sequentially hermaphroditic species

Abstract

●Many angiosperms are hermaphroditic and produce bisexual flowers in which male (pollen export) and female (stigma receptivity) functions are separated temporally. This sequential hermaphroditism may be associated with variation in flower size, color, or pattern, all of which may influence pollinator attraction. In this study, we describe variation in these traits across discrete functional sex stages within and between 225 greenhouse‐grown individuals of Clarkia unguiculata (Onagraceae). In addition, to identify the effects of floral phenotype on pollinator attraction in this species, we examine the effects of these floral traits on pollen receipt in ~180 individuals in an experimental field array. ●Petal area, ultraviolet (UV)‐absorbing nectar guide area, and blue and green mean petal reflectance differ significantly across the functional sex stages of C. unguiculata. Male‐ and female‐phase flowers display significantly different pollinator attraction traits. Petal and UV nectar guide area increase as flowers progress from male phase to female phase, while blue reflectance and green reflectance peak during anther maturation. ●In field arrays of C. unguiculata, female‐phase flowers with large UV nectar guides receive more pollen than those with small nectar guides, and female‐phase flowers with high mean blue reflectance values are more likely to receive pollen than those with low blue reflectance. Female‐phase flowers with green mean reflectance values that differ most from background foliage also receive more pollen than those that are more similar to foliage. These findings indicate that components of flower color and pattern influence pollen receipt, independent of other plant attributes that may covary with floral traits. We discuss these results in the context of hypotheses that have been proposed to explain sex‐specific floral attraction traits, and we suggest future research that could improve our understanding of sexual dimorphism in sequentially hermaphroditic species and the evolution of features that promote outcrossing.

Climate Predicts UV Floral Pattern Size, Anthocyanin Concentration, and Pollen Performance in Clarkia unguiculata

“…However, contrary to our predictions, we found that SumUV had a significant negative effect on proportion nectar guide and nectar guide area. Previous work has shown a significant positive relationship between nectar guide size and pollen receipt in this species (Peach et al, 2020). It is possible that the relationship between latitude and proportion nectar guide observed here reflects adaptation to variation in pollinator availability or assemblage across the state.…”

“…We classified the developmental stages of the flowers of C. unguiculata into three distinct phases: male stage 1 (when the short anthers release their pollen), male stage 2 (when the long anthers release their pollen), and female (stage 3, when the stigma becomes receptive). Previous work has shown that floral sex stage is a significant source of variation in multiple floral traits in this species (Peach et al, 2020). We included floral sex stage as an independent variable in our statistical analyses (described below) so that we could control for its effects on floral phenotype when testing for and characterizing variation among populations.…”

“…Objective measurements of mean UV (300-400 nm), blue (430-500 nm), green (510-530 nm) and red (560-580) reflectance were recorded for each petal analyzed (n = 700). We measured these wavelengths specifically due to their role in pollinator attraction (Peach et al, 2020) and to estimate anthocyanin concentration (see below). The area (mm 2 ) of each region was also recorded.…”

“…We calculated the population mean for each light treatment and functional sex stage. We treated each functional sex stage as distinct because in previous work we found that functional sex is a significant source of variation in these floral traits (Peach et al, 2020). Using these population means (8 populations × 2 light treatments × 3 functional sex stages = 48 mean values) of each of our focal floral traits as a dependent variable, we conducted several stepwise OLS (ordinary least squares) regressions to detect the effects of floral functional sex stage (male stage 1, male stage 2, and female [stage 3]), light treatment (LED vs.…”