Simultaneous pulsed flowering in a temperate legume: causes and consequences of multimodality in the shape of floral display schedules

Abstract

Summary1. In plants, the temporal pattern of floral displays, or display schedules, delimits an individual's mating opportunities. Thus, variation in the shape of display schedules can affect the degree of population synchrony and the strength of phenological assortative mating by flowering onset date. A good understanding of the mechanisms regulating the timing of flowering onset has been developed, but we know less about factors influencing subsequent patterns of floral display. 2.We observed unusual multimodal display schedules in temperate populations of the annual legume Chamaecrista fasciculata. Here, we ask whether 'flowering pulses' are simultaneous among individuals and populations and explore potential underlying mechanisms and consequences of pulsed flowering.3. We monitored daily flower production for individual plants from genetically divergent populations during a series of field experiments that manipulated three potential influencers of display schedule shape: average daily temperature, pollinator availability and watering schedules. We measured floral longevity to isolate the contributions of flower retention and flower deployment to display schedules. We assessed relationships between flowering and environmental variables and compared estimates of population synchrony, individual synchrony and the strength of assortative mating with those of 29 unimodally flowering species from the area. 4.We observed simultaneous flowering pulses in all experiments, with peaks aligned among individuals and populations despite variation in flowering onset and/or duration. Pulses were not the result of increases in average temperature, pollinator availability or variation in watering schedules. Seasonal fluctuations in temperature correlated with floral longevity and flower deployment, suggesting that the shape of display schedules may be plastic in response to fluctuations in temperature. Average population and individual synchrony differed only slightly from those of the species with unimodal schedules, while the average strength of assortative mating for flowering onset date was strongly reduced (0.21 in C. fasciculata vs. 0.35 for the 29 other species).5. Synthesis. Researchers should take caution in assuming that components of display schedules are genetically or developmentally correlated with flowering onset. Variation in the shape of display schedules can influence patterns of gene flow within or between populations, with potential effects on the strength of phenological assortative mating and subsequent responses to selection.

Expression of additive genetic variance for fitness in a population of partridge pea in two field sites

“…), and phenological traits (Kelly ; Wadgymar et al. ,b; Wadgymar and Weis ), making it an ideal system for studying capacity for adaptation and G × E . In the fall of 2013, seed pods from 200 maternal plants were collected from a population of C .…”

The success of assisted colonization and assisted gene flow depends on phenology

“…In this experiment and in others, flowering onset for the NC population planted at KSR always occurred after August 28th, when the photoperiod in southern Ontario was 13.75 h, which is similar to the that experienced upon flowering at the NC home site on August 7th (data not shown). This occurred despite warming‐induced advances in phenology (this experiment) and even when planting seeds 2 months ahead of the MN population (Wadgymar et al ., ). This suggests that flowering onset date in C. fasciculata is under at least partial photoperiodic control and that the evolution of genes associated with photoperiodic responses would be necessary for successful long‐term establishment of populations relocated to northern latitudes.…”

“…We estimated the potential for gene flow between populations, as would occur under an assisted gene flow scenario, by analyzing the overlap in flowering schedules to determine the proportion of opportunities for pollen exchange between populations. For the few days with missing flower count data, we used a linear function running from the day before to the day after the missed count to interpolate the expected number of flowers (Wadgymar et al, 2015). Within a treatment, and for each population pairing, we used daily flower counts to construct and n 9 n matrix of pair-wise mating opportunities between all individuals.…”

Temporal population genetic structure in the pollen pool for flowering time: A field experiment with Brassica rapa (Brassicaceae)

“…Several recent studies found temporal structure at neutral loci in natural populations where individual flowering periods are short relative to the flowering season length (Daïnou et al, 2012; Suni and Whiteley, 2015). In contrast, the overlap among individual flowering schedules in most temperate species is too broad to divide populations into multiple mating pools (Rabinowitz et al, 1981; Herrera, 1986; Ison et al, 2014; Weis et al, 2014; Wadgymar et al, 2015). Yet we found that the realized temporal structure was at least as strong, and potentially stronger, than the predicted structure based solely on flowering schedules (Tables 1, 2).…”