Recent theory on the degree to which pollen receipt limits plant reproductive output suggests that intraspecific variation in pollen limitation should be a common feature of natural plant populations. However, little is known about how common intraspecific variation in pollen limitation is in nature and about which plant traits underlie intraspecific variation in pollen limitation. Here we focus on flowering phenology because plant reproductive success often depends on the timing of flowering. To address the effect of flowering date and duration on pollen limitation, three species were studied during summer 2014: Hydrophyllum fendleri, Linum lewisii, and Potentilla pulcherrima. Pairs of individuals were tagged as they came into flower, and one plant within each pair was randomly assigned to an open vs. hand pollinated treatment (total of 24–35 pairs of plants per species). Every 2–3 days, flowers were counted on all study plants, and supplemental pollen was added to all open flowers on hand-pollinated plants using anthers from nearby pollen donors. Mature fruits were collected and dissected in the lab to determine the number of viable seeds per fruit. Pollen limitation is reported in fruit set (fruits per flower) and seed set (seeds per plant).
Results/Conclusions
On average across all individuals, H. fendleri fruit set and seed set were both pollen limited. Pollen limitation of fruit set was constant across the population’s flowering period (pollination treatment: p = 0.0051; no interaction with flowering date); however, the magnitude of pollen limitation of seed set increased with flowering duration of individual plants (treatment x flowering duration: p = 0.0091). In L. lewisii, average fruit and seed set were not pollen limited, and this was consistent across individuals (treatment x flowering date: p = 0.63, 0.67, respectively). In P. pulcherrima, fruit set was pollen limited and seed set was not. As for H. fendleri, pollen limitation of fruit set was constant across the population’s flowering period (treatment x flowering date: p = 030). However, seed set was only pollen limited in earlier-flowering individuals (treatment x flowering date: p = 0.068). In summary, pollen limitation may be detected at the population-level and be driven by a subset of individuals with longer flowering durations; conversely, pollen limitation may be absent at the population-level but still be detectable in a subset of earlier-flowering individuals. Therefore, examining pollen limitation as a mean across all study plants can mask important variation in pollen limitation.