Current stewardship practices in invasion biology limit the value and secondary use of genomic data

Abstract

Invasive species threaten native biota, putting fragile ecosystems at risk and having a large-scale impact on primary industries. Growing trade networks and the popularity of personal travel make incursions a more frequent risk, one only compounded by global climate change. With increasing publication of whole-genome sequences lies an opportunity for cross-species assessment of invasive potential. However, the degree to which published sequences are accompanied by satisfactory spatiotemporal data is unclear. We assessed the metadata associated with 199 whole-genome assemblies of 89 invasive terrestrial invertebrate species and found that only 38% of these were derived from field-collected samples. Seventy-six assemblies (38%) reported an 'undescribed' sample origin and, while further examination of associated literature closed this gap to 23.6%, an absence of spatial data remained for 47 of the total assemblies. Of the 76 assemblies that were ultimately determined to be field-collected, associated metadata relevant for invasion studies was predominantly lacking: only 35% (27 assemblies) provided granular location data, and 33% (n = 25) lacked sufficient collection date information. Our results support recent calls for standardized metadata in genome sequencing data submissions, highlighting the impact of missing metadata on current research in invasion biology (and likely other fields). Notably, large-scale consortia tended to provide the most complete metadata submissions in our analysis-such cross-institutional collaborations can foster a culture of increased adherence to improved metadata submission standards and a standard of metadata stewardship that enables reuse of genomes in invasion science.

Keywords: biological invasion; invasion genomics; metadata; reference genomes; sequencing.

FIGURE 1

Genome assembly frequency and taxonomic coverage. Bars represent the relative frequency (n  = 1–9) of individual genomes identified within the NCBI repository for 89 terrestrial arthropods, with species coloured by taxonomic order according to the provided key; Inner plot shows the proportion of species comprising each taxonomic order between the reference genome (inner ring; n = 89) and assembly (outer ring; n = 199) datasets.

FIGURE 2

Frequency of genome assembly deposition to repositories and their relative spatiotemporal metadata entry completeness for invasive terrestrial arthropods between 2010–2022. (a) Data were separated into ‘other’—defined as the combination of lab, commercial, and unknown origins—and ‘field‐collected’ based on metadata associated with 199 assemblies, using submission date for each assembly as a classifier. (b) Completeness of spatiotemporal metadata in the assembly dataset by submission year, where ‘complete’ refers to spatial data that includes collection status (field, laboratory), provision of geographic location (to at least country), and temporal (to at least year of collection) information.

FIGURE 3

Metadata trends for reference genome (n = 89) and assembly (n = 199) datasets. (a) Assembly completeness (chromosome, scaffold, or contig level); (b) Sample origin (wild, laboratory, commercial, and managed colony); and (c) Completeness of spatiotemporal metadata (location and/or collection date provided).

FIGURE 4

Metadata trends for field‐collected species assemblies. The proportion (from a total of n = 49; green inner circle) of species with field‐collected data, including those with single or multiple assemblies (shades of yellow), and those that included information on native or expanded range (shades of purple). Includes two assemblies of Aphis gossypii, where the origins of the species were unknown but were specified as field‐collected.

FIGURE 5

Level of geographic metadata provided for field‐collected sample assemblies. Sampling sites for field‐collected sample assemblies were reported at three different resolutions, as indicated by the provide key.