Patterns of wing size variation in seeds of the lily Cardiocrinum cordatum (Liliaceae)

We examined the patterns of variation in wing-loading and its related characteristics in Cardiocrinum cordatum to clarify the factors that determine the variation in seed dispersal ability in this species. The square root of wing-loading of a seed of a plant was not significantly correlated with basal stem diameter of a plant, indicating that large plants did not necessarily produce seeds with high dispersal ability. This result was inconsistent with the hypothesis that large plants produce seeds with high dispersal ability to avoid high mortality of seeds and seedlings in the vicinity of the parents. On the other hand, the square root of wing-loading of a seed of a fruit was negatively dependent on seed number of a fruit. Thus, many-seeded fruits produced seeds with high dispersal ability. This was because the projected surface area per seed was large in large fruits and large fruits contained large numbers of seeds. The cost per seed of producing fruit structures was small for many-seeded fruits. Thus, high dispersal ability of seeds in many-seeded fruits may be a result of an effective resource allocation pattern in which a high proportion of resources are allocated to those many-seeded fruits, enabling seeds to develop large wings and thus reducing the structural cost of fruits per seed.     

Evolutionarily stable resource allocation for production of wind-dispersed seeds

We developed a game-theoretic model for wind-dispersed seed production to examine the seed mass–dispersal ability relationship and the evolutionarily stable distance of seed dispersal in terms of exploitation of safe sites. We assumed trade-offs between masses of the embryo (including albumen) and the wind-dispersal structures per seed, and also between seed mass and number of seeds per parent. We showed that ESS wing-loading is independent of embryo mass; that is, heavy seeds are not poor dispersers if the cost of producing wind-dispersal structures per unit area is constant. The ESS embryo mass per seed depends only on the factors which determine the probability of a seedling being established from a seed. However, wing-loading was found to increase with embryo mass when the change in length was isometric and there was a negative correlation between seed mass and dispersal ability. Thus, the area–mass relationship in wind-dispersal structures may have large effects on the ESS production of wind-dispersed seeds. On the other hand, given that only a limited number of adults can be established at a safe site, the ESS seed dispersal distance depends on the relative degree of sib to non-sib competition. A parent disperses its seeds over a wide area to exploit many safe sites if sib competition is strong. However, it disperses its seeds within a narrow area if the mean number of parents per unit area is large, or if non-sib competition is strong. Thus, in addition to an upper limit on the number of adults per safe site, the degree of sib and non-sib competition may be important for the ESS dispersal distance in wind-dispersed seeds. This revised version was published online in July 2006 with corrections to the Cover Date.