Annual Fluctuatlons of Sapfrults Productlon and S5rnchronlzatlon wtthtn and lnter Specles ln a Warm Temperate Forest on Yakushlma Island

Abstract

Eight years' fluctuations of fruit production were studied in a warn temperate broad-leaved forest on Yakushima Island, southern Japan. Eight tree species characterized by sapfruits or capsules were studied, and all of them were regarded as animal dispersal. Total numbers of sapfruits fluctuated remarkably in eight years for each species. Cleyera japonica and Eurya japonica bore fruits every year constantly and fruiting fluctuations were less remarkable than the other species. Ternstroemia gymnanthera and Symplocos prunifulia seemed to bear fruits many or few in the alternate years, Fruiting fluctuations of Ardisia sieboldii, Myrsine seguinii, Litsea acuminata and Vaccinium bracteatum werc great; they had few fruits in several consecutive years. The mean temperate of the preceding August was correlated with fruiting number of only C. japonica and the total sunshine time of the preceding August was correlated with fruiting number of C. japonica and M. seguinii. Fruiting fluctuation was also influenced by typhoon. A poor crop of V. bracteatum andA, sieboldii had continued after the damage that trunks or branches were broken by typhoon attacked in 1990. For every species except ?, gymnanthera and L. acuminata, great fruiting years tended to coincide among individuals within a species. As for inter species, great fruiting years also synchronized mildly. Kby Words: evergreen broad-leaved forest / fruiting fluctuations / mast fruiting / seed dispersal / Yakushima Island Fruit production of woody plants is known to fluctuate greatly year by year. The large quantity of fruiting that occurs after an interval is called "mast fruiting" (Kelly, 1994). Several hypotheses have been presented to explain that many individuals of a species bore many fruits simultaneously in "mast fruiting" year. Among them, the resource matching hypothesis (Foster, L982; Norton & Kelly, 1988) and predator avoidance hypothesis (Jarlzen,1971; Silvertown, 1980) are applicable for many species. The iesource matching hypothesis and the predator avoidance hypothesis belong to the different categories of explanation: the former is based on a proximate factor and the latter on an ultimate factor. Resource matching hypothesis mentions that climate is a primary factor: the good photosynthetic condition of a tree leads to mast fruiting and unusual weather restricts the level of investments to fruits, so that the fluctuation of rich or poor harvest becomes uniform within many individuals. Biennial fruiting is well known in the fruit horticulture (Hoblyn et al., !936), and the fluctuation of fruit production in orchards synchronizes entirely regional, or more widely, often in all Japan (Ito er al', 1956)' For an aim to reduce annual harvest fluctuation, physiological mechanism offruit trees has been studied well (Nakagawa, t982). In cases of various fruit trees, the accumulation of starch is related to the decision of flower bud quantity (Davis, 1931; ogaki et al., !963; Sobajima, 1967). Predator avoidance hypothesis explains that the steep fluctuation of fruiting is an adaptation as many plants avoid seed predators in mast year by controlling predator's number low in poor harvest