Degradation and restoration method interact to affect the performance of planted seedlings in tropical rainforest restoration : evidence from plant functional traits

Abstract: A common action to improve or save biodiversity in the tropics is restoration of degraded rainforest. To understand the complex ecological structures of the forest and adapt the restoration actions after them, studies have focused on characterizing species with functional traits. Traits can be placed on a sliding scale between pioneer and climax properties and describe species life history. They can be used to predict species response in different environments, without detailed ecological knowledge for that species. The hypotheses in the study were that forest type and type of restoration treatment affect functional traits and that these traits could predict growth. Furthermore the study investigated if there was a growth-survival trade-off and if it could be predicted with functional traits. Measurements came from a field experiment in Sabah, Malaysia, where seedlings were planted with two methods within three forest types. The two treatments were line planting, where seedlings were planted in cleared lines under the canopy and gap-cluster planting, where seedlings were planted in groups in pre-existing or created gaps. The three forest types were classified based on degradation level and vegetation, forest type A was the most degraded, type B intermediate and type C the least degraded. 32 species of dipterocarps, fruit trees and other climax species were planted within the experiment. Treatments and forest types were thought to create different environmental conditions, e.g. light availability. The effects on growth were tested with a general linear model and the survival with logistic regression. Also species groups were compared with chi-square test to examine the growth-survival trade-off. The traits that were used were total height, stem diameter, crown depth, crown width, leaf stem length, total leaf area, leaf thickness, vein thickness, and leaf damage. Species was the factor that effected both growth and survival the most. Seedlings in line treatment performed similarly in all forest types, but gap-cluster planting was slightly better than line planting in forest type A and clearly outperformed line treatment in forest type B (e.g. in forest type B seedlings in gaps increased total height by average 33 % the last 9 month compared with lines 25 %). Crown depth ratio, total leaf area, and leaf stem length were clearly the traits that best predicted both height and diameter growth. In this study the growth survival trade-off was not obvious, only total height was negatively correlated with survival rate. The differences in traits due to treatments showed that the treatments created different environments for the seedlings and treatment effects were different depending on forest type. Gap-cluster planting gave higher growth in forests there the seedlings were less limited by light availability. The conclusions of this study could contribute to improved effectiveness of restoration actions in the rainforests of Borneo and be of value when planning future projects, especially regarding selection of species in enrichment plantings based on functional traits.