Reproductive Consequences of Insect Seed Predation in Hamamelis Virginiana

Abstract

Through the use of populations of marked individuals over a 4—yr period, I studied the demographic impact on a temperate forest shrub, witch—hazel (Hamamelis virginiana), of predispersal seed predation by a host—specific weevil (Pseudanthonomus hamamelidis). The general consequences of seed predation for witch—hazel recruitment are dependent upon the interaction of seed predation with the pattern of annual fruit production. Fruit production was poor in 1977 and 1978, and in these 2 yr, seed predation intensities (percent fruits attacked per crop) averaged 80—90%. In 1979, fruit crop sizes increased dramatically, resulting in only 45% seed predation on average. Fruit production declined again in 1980, and the weevils were able to attack nearly 100% of the available fruit. Thus at the population level, successful seed production is episodic, with large numbers of seeds being produced in the occasional good fruiting year through satiation of seed predators. The host specificity of Pseudanthonomus provides a mechanism for direct limitation of weevil population size in poor fruiting years. I also evaluated the effect of seed predation on the relative contributions of different witch—hazel individuals to successful seed production. The distribution of individual plant fecundity in any given year is highly skewed, with only a few plants in the population large fruit crops. In any given year, more fecund plants incur neither higher nor lower percent fruit attack on average than less fecund individuals. Seed predation therefore does not change the relative contributions of different—sized individuals to annual seed production; more fecund plants produce more viable seeds than do less fecund plants, and the few highly fecund individuals in the population make a disproportionately large contribution. Between—plant variation in witch—hazel fecundity is thus significant demographically, even though it may not be subject to natural selection from seed predation pressure in these populations. The spatial distribution of witch—hazel is highly clumped, with few individuals being very isolated from conspecifies. Spatial isolation reduced the probability of a plant's location by weevils only in the case of small, low—fecundity individuals; large individuals, in contrast, were likely to be found regardless of distance from other conspecifics. Among the larger plants there were slight but nonsignificant tendencies for percent fruit attack to be lower with increased distance from conspecifics. Spatial isolation in this system seems important in lowering seed predation intensity only in interaction with individual plant fecundity.