We use a common prairie perennial, Echinacea angustifolia, to study how flowering phenology can cause reproductive isolation among plants. We have previously shown that flowering phenology can isolate plants so that plants flowering past peak flowering set significantly less seed than peak or early flowering plants. Here we examine individual plants’ flowering phenologies over three years (2005-2007) with the following objectives: 1) compare population level phenology parameters across years, 2) determine if a plant’s flowering phenology in one year predicts its flowering phenology in subsequent years, and 3) quantify an individual plant’s phenological isolation in each year. Adult E. angustifolia plants, like many long-lived perennials, typically do not flower every year so populations must be monitored for several years to characterize flowering phenology both at an individual and population scale. Our study takes place in an experimental common garden in western Minnesota. Flowering phenology was recorded by counting each plant’s total number of anthers and receptive styles every day during 2005 and every other day during 2006 and 2007.

Results/Conclusions

In all three years E. angustifolia flowered from late June to early August with peak flowering occurring in the first or second week of July. The duration of the flowering season was similar across years (57 days in 2005, 49 in 2006, and 47 in 2007). However, the total number of flowering plants varied greatly between years (224 in 2005, 696 in 2006, and 621 in 2007).  A total of 105 plants flowered all three years and 198 plants flowered in 2006 and 2007 but not in 2005. Using non-parametric correlation tests we found significant correlation (p<0.05) but high variability of flowering time parameters (first and last date of flowering, and duration of flowering) for those individuals that flowered multiple years. Even with high variability in the total number of plants in flower each year, the level of phenologically isolation experienced by individual plants is similar among years. The phenological isolation we observed in this common garden experiment is likely also present in the nearby highly fragmented populations of origin, where spatial isolation has been shown to reduce population and individual fitness. Phenological isolation caused by flowering asynchrony could negatively impact fragmented populations more than predicted by spatial isolation alone. We also discuss how phenological isolation might differ from spatial isolation due to the heritability of flowering time.