ETPP-31 Assessment of Global Risk Analysis for Novel Tree Species Phase 2 Report: Appendix 4 Resources: TRY trait-based resources

Trait-based resources developed to support risk analysis questions of the research project ETPP-31 Assessment of Global Risk Analysis for Novel Tree Species, commissioned by the UK's Department for Environment, Food & Rural Affairs (Defra). ---- Biological characteristics of species can explain differences in invasion success among plant species (van Kleunen et al., 2010). There is strong ecological theory (Westoby, 1998) and empirical evidence relating leaf-height-and seed (L-H-S) traits to different ecological strategies like competitive ability, tolerance of abiotic stress and regeneration after disturbance (Diaz et al., 2016). Traits linked to these strategies have been observed to explain the outcome of competitive interactions among trees (Kunstler et al., 2012). Plant trait databases are among the richest of any taxa. The Phase I report form this project found that species trait information was used in 12 of the 38 risk assessment methods reviewed. Those risk assessments considered traits relating to reproduction, viable seed production, propagule pressure, dispersal capacity and growth rates. In our phase 1 engagements, several participants reinforced these points through examples of value of trait resources and how beneficial traits for forestry, such as vigorous growth and self-regeneration, could themselves become risk factors for invasion requiring careful assessment. To provide a resource to support risk assessors, we extracted all publicly available trait data from the TRY database for 214 Non-native tree species (NNTs) and 31 native species with a taxonomic match to accepted names in the TRY database. For the 10 NNTs selected for light-touch stress testing, we additionally extracted trait data for their congeners at species level – although congeneric-level data were not available until after the stress testing was complete. Species names on the longlist of NNTs were first matched to Kew’s World checklist of Vascular Plants using the R package ‘rWCVP’ (Brown et al. 2023) to identify synonyms and then intersected with the TRY accepted species names (try-db.org/dnld/TryAccSpecies.txt). The resulting TRY Accession numbers and TRY accepted names for these species are included in the supporting files to enable the trait data resource to be periodically updated with new trait data for these species, if required. We also include the TRY Trait ID numbers, so that the same trait data can be extracted for future NNTs subject to the risk analysis (e.g. not yet on the long list we used), though we would recommend extracting all available trait data for the focal NNT in case additional trait information to support the risk analysis can be identified by NNT species experts. The TRY trait-based resources created in Phase II comprise: 1. Data available for each NNT summarising i) the number of traits available in the TRY database that were identified in this project as being of value to support the risk analysis, ii) the total number of traits available in the TRY database, and iii) the total number of individual trait measurements available for each species. This spreadsheet may be used to assess whether the resources will have sufficient information able to be of use for the risk analysis of that species (TRY_data_availability_summary_by_species.csv).). 2. Data available for each trait summarised as number of native and NNT species with relevant trait data, mapped to risk analysis question (TRY_trait_data_availability_summary_by_trait.csv). 3. A folder for each NNT species (NNTs_species_traits/[species name]) with i) all available trait data for each species, which can be filtered by risk analysis question number and trait name and ii) a figure with simple colour-coding of the quantitative traits for the NNT compared to native species: orange indicates the NNT trait value is higher than the 75th percentile of the values among native species; blue indicates the trait value is substantially lower than the 25th percentile of trait values among native species; grey means it falls within the interquartile range of native species trait values (Fig. 1). 4. For species with no trait data for relevant traits, we also provide congener species trait data (in the folder NNTs_congener_traits/[species name]) in case there are suitable congeneric species that can be used as a proxy for likely trait values (though noting that use of congeners should be carefully documented in the protocol: see guidance for risk assessors) 5. For reference, we also provide the interquartile ranges of the quantitative trait values for the 31 native species (TRY_Trait_IQR_native_species.csv). A list of the datasets that contributed to these resources is provided in the file 'TRY_trait_based_resources_datasets_and_references.xlsx'.