Two components of the floral display of Viburnum opulus L. were manipulated to determine their effect on fruit initiation and maturation. We altered the size of the floral display by planting individual shrubs in groups of 1, 5, or 10 in 1985 and in groups of 2 or 6 in 1986. In addition, we changed inflorescence size by altering the number of sterile accessory flowers per inflorescence: 0, 4, or untreated in 1985 and 0 or untreated in 1986. The sterile-flower treatment did not explain a significant proportion of the variation in fruit initiation or maturation. In contrast, the plant-grouping treatment was significant in 1985 but not in 1986. The number and proportion of fruits initiated were higher in larger groups of plants than in small groups in 1985 (P < 0.0001 for analyses performed at group and plant levels). The proportion of fruits initiated in group sizes 1, 5, and 10 was 6.9, 15.0, and 22.7% per plant, respectively, and 4.3, 9.1, and 19.4% per inflorescence, respectively. The larger groups did not initiate or mature proportionally more fruits in 1986. Twice as much rain fell in 1986 as in 1985, and shrubs produced more fruits overall (32.2% fruit initiation versus 11% in 1985), but it is not clear why group size differences did not have an effect on fruit initiation and maturation in 1986. Within the plant-grouping treatment fruit initiation and maturation were always significantly correlated with flower number (P < 0.0001) in both 1985 and 1986, but the increase in the number of fruits initiated was not proportional to the increase in flower number. Key words: floral display size, fruit initiation, plant grouping, pollination, Viburum opulus.
Self-pollination rate and floral-display size in Asclepias syriaca (Common Milkweed) with regard to floral-visitor taxa
