Production rates of reproductive parts in Aesculus turbinata forests, with special reference to pollen and seeds.

To determine the amount of photosynythates flowing to reproductive parts in trees bearing large and heavy seeds and whether insect-pollinated trees produce less pollen than wind-pollinated trees, a six-year study was done in two horse chestnut (Aesculs turbinata Blume) forests at Ashiu, Kyoto prefecture. The main results were as follows. (1) Annual production rates of all reproductive parts were 2t/ha.yr for a mast year, about 1.2t/ha.yr on average, and 0.5t/ha.yr for poor fruiting. The production rates for horse chestnut were not as high as the values for other trees such as beech and chinquapin. In terms of seeds alone, horse chestnut appears to be one of the most productive trees. (2) Dry-matter production was concentrated on female parts (65%-85%) derived from hermaphrodite flowers, except in poor crop years (40%). (3) Small annual fluctuations in the numbers of both male and hermaphrodite flowers were recognized. (4) There were differences between stands in the sex ratio (ratio of number of hermaphrodite flowers to total), being 0.013-0.037 for one stand and 0.048-0.085 for the other. (5) Flowers before anthesis had 246-321*10(3) pollen grains and a weight of 1.56-1.88mg. (6) Pollen production rates estimated by multiplication of the total number of flowers produced in a l-ha stand by the amount of pollen per flower before anthesis were in the range of 5.29-12.7 (mean: 9.13)*10(3) grains/ha.yr and 32.0-87.4 (mean: 55.9) kg/ha.yr. The other stand produced 20% less than these values. (7) Numbers of pollen grains produced by insect-pollinated horse chestnut were approximately equal to or more than those of wind-pollinated trees. (8) Mast fruiting occurred in years of good hermaphrodite flower crop, and was partly brought about by high seed-maturing ratios (numerical ratio of mature seeds to hermaphrodite flowers). Year-to-year trends in seed-maturing ratios parallelled pollen production, except in one year. Numbers of pollen grains scattered per ovule or ratios of pollen grains to ovules were in the range of 1.3-3.2*10(6) for one stand and half of this value for the other.