Primary and secondary seed dispersal was investigated for the glacier
lily Erythronium grandiflorum in the Colorado Rocky Mountains. These h
eavy seeds have no obvious adaptations for biotic or abiotic dispersal
, but can be thrown short distances when the dehiscent fruits are shak
en by wind. We used sticky traps to measure primary transport of seeds
up to 1 m away from individual plants. A seed cafeteria experiment ex
amined the role of ants and rodents in secondary seed transport. Prima
ry dispersal by wind was positively skewed and median transport distan
ces were influenced by Variation in plant height. Secondary dispersal
was negligible compared to Viola nuttallii, an elaiosome-bearing speci
es. Thus, seed dispersal was highly restricted in E. grandiflorum, and
a 1 m radius encompassed the modal section of the seed dispersal curv
e. The seed dispersal component of gene flow was quantified and combin
ed with previous measurements of pollen flow to yield a more complete
estimate of Wright's neighborhood size, N-e, for E. grandiflorum. The
lack of a special seed dispersal mechanism in E. grandiflorum is discu
ssed in terms of a source-sink model for seedling establishment with r
espect to distance from the parental plants.