International collaborative research project, supported by the National Science and Technology Council, Taiwan, NSTC 115-2923-B-002-001-MY3; duration: 2026/04/01-2029/03/31

Biological invasions are a major driver of global biodiversity loss, with their frequency and impact accelerating due to intensifying trade, transport, and human mobility. Although only a subset of introduced species become invasive, those that do impose substantial ecological, economic, and societal costs. A central challenge in invasion biology is that invasion processes and outcomes are strongly context-dependent: they arise from interactions between species' functional traits, environmental conditions, introduction pathways, land-use patterns, and socio-economic drivers. These complexities make it difficult to identify general mechanisms and to design effective early-warning and management strategies. The emerging framework of invasion syndromes provides a promising solution by identifying recurrent socioecological configurations—trait combinations, environmental contexts, and human drivers—that reliably produce similar invasion outcomes. To operationalise this framework, high-quality, harmonised global datasets of plant functional traits are essential.

This NSTC-funded national project contributes to the European-funded BIODIVERSA+ consortium InvaSyn by leading the collection, harmonisation, and ecological interpretation of plant functional trait data. The project integrates Taiwan into an international collaboration that aims to develop the first global typology of invasion syndromes. Specifically, the work focuses on assembling a comprehensive trait dataset from major global resources, including the GIFT database, the TRY Plant Trait Database, floristic monographs, and regional trait compilations. These traits will be standardised across taxonomies, measurement protocols, and units, using authoritative tools such as the GBIF Backbone Taxonomy. Harmonised trait data will be systematically linked to global databases of naturalised and invasive species occurrences (GloNAF), as well as to environmental and socio-economic datasets that shape invasion dynamics. These include global climate layers (TerraClimate, WorldClim, CHELSA), landcover products from the ESA Climate Change Initiative, long-term human landuse reconstructions (HYDE), and indicators of socio-economic development from the World Bank.

Analytically, the project contributes to the statistical identification of invasion syndromes using unsupervised learning approaches, including clustering and dimensionality reduction. The Taiwanese team will provide trait-based ecological interpretation of model outputs, ensuring that identified syndromes correspond to biologically meaningful patterns consistent with established mechanisms of plant invasiveness. Rigorous data-quality procedures—including trait imputation, outlier screening, and dimensionality reduction—will address known challenges arising from heterogeneous global datasets. High-performance computing resources at National Taiwan University will support data processing and modelling. The anticipated outcomes include: (1) a harmonised and openly accessible global trait repository suitable for invasion-syndrome analyses; (2) reproducible analytical workflows developed in R; (3) scientific contributions to consortium manuscripts on trait integration, computational workflows, and the final invasionsyndrome typology; and (4) increased international visibility of Taiwan's ecological research. The project actively applies and promotes FAIR data principles to maximise the long-term value of its outputs.