Use of functional traits to classify tree species planted for forest restoration into successional guilds

Abstract

Since forest ecosystem restoration aims to accelerate forest succession, selecting tree species of appropriate successional status is vital. Traditionally, ecologists have divided tree species into light-dependent pioneers and shade-tolerant climax species. However, since many species do not fit into this binary classification, a gradient of successional guilds is more appropriate. This project aimed to distinguish successional guilds by using functional traits, i.e., species-specific characteristics that influence species performance or fitness. Species-specific values of 27 functional trait variables, linked to successional status, of 28 tree species, planted in restoration plots in 1998, were derived from FORRU’s databank, online databases and by direct measurement. The data were first subject to decentralized standardization. Then closely correlated variables were identified, using Pearson’s correlation coefficient. One variable of each closely correlated pair was discarded, following recommendations from literature. The remaining 13 non-correlated traits were: growth, mortality, half-life, wood density, leaf area, specific leaf area, leaf dry-matter content, leaf nitrogen and phosphorus concentration, dry seed mass, median length of dormancy, seedling type, and photoblastic germination. Early growth and survival data (1998 to 2000), were combined with long-term data, collected during this study (2017 to 2020). Two classification systems were compared. Rank-scores were calculated by assigning negative (0 to -2) and positive scores (0 to +2) to traits, associated with early and late succession, respectively. The successional status of a species was therefore expressed as the sum of all the plant trait scores—increasingly negative sums indicating stronger pioneer status and increasingly positive ones indicating stronger climax status. This method allowed distinction of 7 successional guilds as follows; Super Pioneer, Pioneer, Intermediate Pioneer, Intermediate, Intermediate Climax, Climax, and Super Climax species. The second method, applied hierarchical cluster analysis using the uncorrelated 13 traits. The cluster dendrogram, grouped the species into 6 successional guilds: Pioneer 1, 2, and 3, Climax 1, 2, and 3. These two different methods of determining the successional status were in very close agreement with each other. Such close agreement provides cross-verification of the assumption that functional traits can be used to indicate successional status, which can be used to design optimum species mixes, maximize recovery of forest structure and biodiversity during restoration projects, and mimic natural forest succession.