The effect of reforestation using Acacia mangium on above- and belowground ecosystem properties

Jenny Vivian

doi:10.25907/01097

In the tropics, Acacia mangium is among the most planted exotic tree species for improving ecosystem services in deforested areas. However, short-term projects often result in minimal management and inadequate long-term monitoring, which limits the current understanding of the potential of A. mangium monocultures to support forest recovery. This study aimed to improve knowledge on early and longer- term effects of A. mangium plantings on ecosystem recovery, focusing on tree diversity, above-ground biomass of living trees, soil fertility (soil organic carbon, total nitrogen, available phosphorus), and microbial communities of the soil and rhizospheres of dominant plants. Four interrelated research objectives were formulated to address this aim: 1) to compare characteristics of soil fungal and bacterial communities across A. mangium plantations at different ages and reference states; 2) to assess the potential of A. mangium plantations to catalyse restoration trajectories towards the tree community composition and carbon stocks of nearby remnant forests; 3) to gain novel insights on the ecosystem modification over time of minimally managed A. mangium plantations by integrating data on the microbial community characteristics, tree species richness and above-ground biomass, and soil fertility; 4) to evaluate the environmental and management conditions under which A. mangium plantations can facilitate forest restoration. We compared sites in 2-, 10-, and 24-year-old A. mangium plantations with remnant forests and degraded Imperata cylindrica grasslands, representing the intended and baseline states before planting, respectively. The plantations were left minimally managed after a few years, enabling natural regeneration to occur. To achieve the first objective, we conducted metabarcoding of the ITS2 and 16S rRNA (V3-V4) gene regions for fungi and bacteria, respectively, using environmental DNA (eDNA) sampled from the bulk topsoil and rhizosphere of the dominant plant species. Fungal and bacterial taxa and functions were identified using genetic databases, and diversities and compositions were compared using PERMANOVA and Principal Coordinate Analysis (PCA). Associations with environmental parameters were tested along with their co-occurrence networks, and stochastic or deterministic processes shaping the bacterial communities were assessed through Mantel correlations. For the second objective, the modification over time of the tree community was evaluated through PERMANOVA, PCA, and Non-Metric Multidimensional Scaling. Differences in tree communities and other abiotic parameters were assessed using Linear Mixed Models. The difference in values of above-ground biomass and topsoil organic carbon (C) stock among landcover types was tested through Generalized Least Squares Models. For the third objective, clustering and classification machine learning techniques were adopted for 184 assessed variables (i.e., including taxonomic and functional traits of microbes with environmental data). The fourth goal was achieved by comparing the potential causes for the invasive or beneficial traits of A. mangium for forest restoration across ten case studies (this study and nine additional) using a Bayesian Belief Network (BBN) and projecting its spatial distribution under future worst-case possible climatic conditions. The taxonomic and functional composition of microbial communities differed among landcover types, but generally not among rhizospheres. Topsoil organic C, pH, and total nitrogen were key factors affecting the composition of microbial communities. Analysis of their co-occurrence networks revealed increasing microbial community resilience over the natural regeneration of the plantations. Stochastic processes predominantly influenced the composition of the bacterial community through different forces, consistent with ecosystem changes. Symbiotrophic fungi and copiotrophic bacteria recovered to the reference forest state, increasing with the age of the plantation. Tree community diversity increased with A. mangium plantation age, transitioning toward that of the remnant forest. With the variation of the tree community, the above-ground C shifted from short- term to long-term storage, reaching an average value of 79.2 Mg/ha. Analysis integrating numerous environmental components unravelled the sequential similarity of the forest ecosystem states, while the individual ecosystem parameters change at different paces. The BBN analysis indicated the most influential conditions promoting forest recovery from A. mangium plantations were a heterogeneous hilly landscape, proximity to natural forests (for their role as seed sources for native species), relatively small size (< 10 ha), and minimal management. The consideration of multiple ecosystem parameters and machine learning techniques highlighted where and how A. mangium plantations act as a catalyst for forest ecosystem recovery. The study also revealed the importance of long-term monitoring of large-scale reforestation efforts using exotic species to forecast impacts on carbon and biodiversity. Through a comprehensive ecosystem-wide analysis, this thesis provides new knowledge on the role of minimally managed A. mangium plantations for ecosystem recovery, validating the efficacy of cutting-edge analytical techniques to unravel trajectories and the effectiveness of forest restoration.

The effect of reforestation using Acacia mangium on above- and belowground ecosystem properties

Files and links (1)

Abstract

Details

Metrics