Pollinator Limitation Ofcytisus Scoparius(scotch Broom), an Invasive Exotic Shrub

Introductions of exotic species provide unique opportunities to study the demographic significance of species interactions, but as yet there is little information on how mutualistic interactions affect the invasion process. A shortage of mutualists could potentially limit the rate of population growth for an invading species. The introduced shrub Cytisus scoparius (Scotch broom, Leguminosae) is a pest plant on the west coast of North America. It produces flowers that are “tripped” open when pollinated and has a nearly obligatory relationship with resident bumble bees and honey bees. Experiments in the state of Washington showed that <1% of untripped flowers produced fruits and that outcross-pollinated flowers yielded fourfold more fruit than self-pollinated flowers, revealing apparent inbreeding depression. Mean pollinator visitation rate, as determined by the proportion of flowers tripped, varied among three years and among four populations but was low (3–30%) in every case. Two urban populations (Magnuson Park and Discovery Park in Seattle) received higher numbers of visits than two native prairie populations (Johnson Prairie and 13th Division Prairie in southwestern Washington). Hand-pollination experiments revealed significant pollinator limitation in all populations in both 1993 and 1994, with the mean increase in fruit production ranging from 280 to 2620%. Prairie populations were more pollinator limited than urban populations. Resources available for per-flower fruit production appeared to be equally available in all populations in 1993 but more available in prairie populations than in urban populations in 1994. The relationship between natural visitation and proportion fruit set per branch was a saturating curve in 1993 but a linear function in 1994 and 1995. Significant correlations were found between pollinator visitation and variation in whole-plant fruit production in all 3 yr. No evidence was found for either (1) reallocation of resources between branches within a season, or (2) a cost of reproduction between seasons in plants receiving supplemental pollen. Demographic analysis showed that a very large cost of reproduction would be required to counterbalance the increase in fecundity achieved with full pollination. Simulations of new populations invading over a short time scale (10–30 yr) demonstrated little effect of pollen limitation in the slow-growing urban populations, but a potentially large effect of increasing pollinator visitation in the rapidly invading prairie populations.