Pollen and (or) pollinator limitation may be most likely to occur in s
pring wildflowers or populations of species near their distribution li
mits. From 1990 to 1992, we examined a 2-ha population of a perennial
spring wildflower, Hepatica acutiloba (Ranunculaceae), at the northern
distribution limit of the species. To distinguish between the often c
onfounded effects of pollen and resource limitation, we applied partia
l and whole-plant treatments. Ramets were open-pollinated, cross-polli
nated, open- and cross-pollinated (50% of flowers were open-pollinated
, and 50% were hand-pollinated), or self-pollinated. Although there wa
s year to year variation, plants with cross-pollinated flowers produce
d significantly more viable seeds and heavier seeds than plants with s
elf- or open-pollinated flowers. Comparisons between seed numbers and
masses from complete and partial treatments suggested that there was n
o reallocation of resources among flowers on a plant. From 1990 to 199
2, a minimum of 95% of all visits to a subset of ramets of H. acutilob
a were from Asclera ruficollis (Coleoptera: Oedemeridae), i.e., virtua
lly the only insect active during flowering. Mark-recapture data for A
. ruficollis indicated that even after 6-48 h, 22-40% of the individua
ls remained within 10 m of the same ramet of H. acutiloba. Observation
al data indicated that less than 50% of all flowering ramets of H, acu
tiloba may be visited by A. ruficollis. Pie conclude that pollen limit
ation, mediated by reliance on one sedentary pollinator, was the main
cause of low seed production in open-pollinated flowers of H. acutilob
a.