Light Requirements of Neotropical Tree Seedlings: A Comparative Study of Growth and Survival

(1) Eighteen species of wind-dispersed trees on Barro Colorado Island, Panama, are compared with respect to germination, survival, and growth of seedlings during 1 year under sun and shade conditions in a screened enclosure. The species vary in their mean dry weight of seed from 1.9 mg to 686 mg. (2) High and synchronous germination occurs in both sun and shade for sixteen of the eighteen species. Most species have epigeal germination with leafy green cotyledons. (3) Seedlings of fifteen species survive better in sun than shade; none survive better in shade than sun. Survival in shade and in sun are positively correlated. Shade tolerance varies widely and continuously among the species; it correlates with adult wood density, an indicator of growth rate and successional status, but not with the dry weight of seed reserves. The more shade-tolerant species have a lower proportion of seedlings dying from disease in the shade. (4) Seedling height after 1 year is greater in sun than in shade for all species. In both sun and shade, growth rate and height at 1 year are not correlated with adult wood density or shade tolerance. Seed dry weight correlates with seedling height in both sun and shade at 4 weeks, but with height in shade only at 1 year. Seedlings in shade show no or slow growth after the initial growth has used the seed reserves. (5) In the shade, number and length of leaves show little or no change with time; leaf turnover is negligible. In the sun, leaf number increases and successively younger leaves mature at larger sizes. Leaf size is greater in the sun than in the shade Senescence of the leafy cotyledons is more rapid in sun than shade for all species. (6) This experimental study predicts that all species examined benefit from seed dispersal to light-gaps. Some species are restricted as seedlings to gaps, while others persist, but do not grow appreciably, in the shaded understorey. The implications of these differences in light requirements for the size-age structure and for the spatial pattern of the populations are discussed.