A 10-year evaluation of the functional basis for regeneration habitat preference of trees in an African evergreen forest

The spatial distribution of tree juveniles in relation to light environments may reflect species differences in growth, survival, and functional traits and will shape the nature of forest regeneration. Long-term field experiments are important to evaluate this issue because of the potentially very long juvenile period in trees. Here, we combine a 10-year seedling survival–growth data with the results of community ordination and multivariate analyses of functional traits to ask how observed juvenile light guilds are related to species functional traits and seedling performance. We transplanted seedlings at a standardized height of 11 cm into the shaded understory and quantified their growth and survival for 10-years. Using the community-wide stem distribution data, we categorized 33 species including the focal 11 species to understory vs. gap/edge guilds. Then, we determined differences between the two guilds in seedling survival, growth, as well as seed size, adult height, and a series of leaf traits, including toughness and chemical traits (fiber, protein, phenolics, tannins, alkaloids, saponins). Among the 11 non-pioneer species whose seedlings were planted into the understory, there was no significant difference in 10-year survival between light guilds, but species in gap/edge guild tended to achieve greater height than species in the understory guild. The leaf chemical traits of 33 species did not differ between the two juvenile light guilds, but gap/edge species had smaller seeds, taller adults, and tougher leaves than understory species. We used logistic regression as a complementary approach to assess the extent to which plant traits varied between light guilds and the most parsimonious model based on AICc ranking included only leaf toughness and had an Akaike weight of 0.52. In addition, across the 11 species planted as seedlings, these traits were not significantly related to survivorship or growth over 10 years. A Principle Components Analysis illustrated associations among traits. We conclude that light guilds in terms of juvenile stem distribution could not be explained by long-term field performance of post-establishment seedlings alone. Earlier seedling stage or later sapling stage may be more important in differentiation of light guilds. For the species examined difference in growth rates could be linked to seed size and adult stature, but not to the adult leaf chemical traits considered. These results suggest the importance of examining ontogenetic shifts and relationships among functional traits for a better understanding of regeneration strategies of tropical trees.