Variation of Dispersal Phenology in a Bird‐Dispersed Shrub, Cornus Drummondii

Abstract

Variation in the timing of seed dispersal among individuals of Cornus drummondii was marked at two sites in east—central Illinois. Individuals whose fruit crops were removed most quickly shared few characteristics. Fruit removal rates, dispersal efficiency (percent of total crop removed), and dispersal success (estimated percent of dispersed seeds attributable to each individual) were not closely related to characteristics of the fruits or the immediate environment of the parent plant. Although small fruit crops were often depleted proportionately faster than large crops, dispersal success relative to others in the population was positively related to crop size; i.e., plants with larger crops contributed more seeds to the total pool of dispersed seeds than plants with smaller crops. Over the 4—yr study, individual plants often exhibited between—year differences in ripening time, crop size, and pulp characteristics, but between—year correlations of traits were only sporadically significant. Crop size showed the most consistent between—year correlation (i.e., plants with large crops tended to make large crops in all years), but even this trait was not always correlated between years. Dispersal success was usually correlated between years, but dispersal efficiency was not. Annual (and sometimes site) differences in most dispersal—related plant traits and in dispersal efficiency and dispersal success were common. No plant traits except crop size had consistent effects on dispersal success, although we found some evidence of selective foraging. We suggest that fruit removal may be more a function of avian biology (flocking and social behavior, frequency of foraging interruptions, sun—seeking in cold weather, etc.) than of most characteristics of the fruiting plant. The ecological consequences of differences in dispersal phenology were variable between years. Postdispersal seed predation and germination success showed no consistent seasonal trends. The most important consequence of failure of early fruit dispersal may be a high probability of nondispersal (falling, rotting, being eaten by a predispersal seed predator or poor dispersal agent). Great variation in space and time in the outcome of bird/fruit interactions means that studies directed at the evolution of the interactions need to have both larger samples sizes and longer duration than has been feasible so far.