Phylogenomic Barcoding of Soil Seed Bank-Persistent and Wind-Dispersed Non-Native Plant Species in South Georgia

Abstract

Climate change and invasive species are leading drivers of biodiversity loss, with island ecosystems being especially vulnerable. South Georgia, a remote sub-Antarctic island, is 170 km long with approximately 30,000 ha of vegetated coastal areas, as snow and ice dominate the inland regions. Human activities on the island have historically introduced non-native species, resulting in 41 introduced vascular plant species compared with only 24 native ones. To address this imbalance, the South Georgia Non-Native Plant Management Strategy was implemented (2016-2020) to control non-native plant populations. We assessed emergent seedlings from South Georgia soil samples and wind-dispersed seeds to determine which species persist in the soil seed bank and contribute to dispersal. Using a molecular barcoding approach, we evaluated traditional markers (rbcL and matK) and optimized a high-throughput Angiosperms353 sequencing pipeline for accurate seedling identification. We generated a reference library covering all native and non-native species and applied this to 1,498 emergent seedlings and 737 trapped seeds. Molecular barcoding identified 21 species, including 10 non-natives and 11 natives. Strikingly, 84% of emergent seedlings were non-native, with Class III invasive species (Cerastium fontanum, Poa annua, Taraxacum officinale) dominating across most sites and in all wind traps. By contrast, Class I and II species occurred rarely and only at a few sites, indicating that management efforts have substantially reduced their spread, though viable seeds persist in the soil. These findings highlight both the continued threat from persistent seed banks of dominant invaders and the value of molecular barcoding for long-term monitoring. Our approach provides a framework for biosecurity and restoration management in South Georgia and other vulnerable ecosystems under climate change pressures.

Keywords: angiosperms353; genome skimming; high‐throughput sequencing; hyb‐seq; invasive; phylogenetics; seedlings identification; target capture sequencing.

FIGURE 1

Experimental design framework to identify seedlings from South Georgia using molecular barcoding: (1) producing a reference library with Angiosperms353; (2A) sampling seeds collected by wind, (2B) sampling seedlings emerging from soil samples, (3A) sequencing individual samples with Angiosperms353 to identify seedlings using phylogenetic topology and patristic pairwise phylogenetic distances, (3B) identifying samples within OTU mixes with genome skimming and plastome SNP density. A353, target capture sequencing using the Angiosperms bait set kit. OTU, operational taxonomic unit.

FIGURE 2

Phylogenetic tree with all angiosperms known to occur in South Georgia (in orange, non‐natives, in dark blue, natives) and relevant species occurring in neighbouring islands (in black) reconstructed using Angiosperms353 sequence data. Numbers on branches represent SH‐aLRT support (%)/ultrafast bootstrap support (%); asterisk * represents 100.

FIGURE 3

Density plots showing the inter‐ and intraspecific phylogenetic distances for several genera and species calculated from a phylogenetic tree including all seedlings from the soil samples (see Figure S6).

FIGURE 4

Analysis of invasive species incidence after having identified all seedlings using seedling morphology and molecular barcoding approaches. (A) Bar plots showing the total number of seedlings by category (invasive, native) per collecting site for soil samples. (B) Bar plots showing the total number of seedlings found for each species in soil samples. (C) Incidence of each species across all studied sites for soil samples. (D) Number of seedlings germinated from seed traps per species.