# Genomics-based assessment of the geographic origin of spongy moths (Lymantria dispar) intercepted during vessel inspections, using SpongySeq, an amplicon sequencing panel

## TL;DR

Researchers developed a genomic method to identify the geographic origin of invasive spongy moths intercepted in North American ports, helping prevent their spread.

## Contribution

A novel amplicon sequencing panel (SpongySeq) was developed to accurately assign geographic origins of spongy moths using 283 SNPs.

## Key findings

- The SpongySeq panel achieved 82–97% accuracy in assigning spongy moths to 19 geographic groups using DAPC.
- Most intercepted egg masses in U.S. ports were traced to Japan, indicating a primary source of invasion risk.

## Abstract

Invasive alien species (IAS) are a major threat to native biodiversity, ecosystems services, economic stability and human well-being. The two spongy moths, Lymantria dispar asiatica and L. dispar japonica, native from Asia, are important defoliators of a wide variety of hardwood and coniferous trees, and the risk of their introduction into North America via sea transport is considered high by plant protection regulatory authorities. To prevent such introductions, a cost-effective approach consists in reducing the likelihood that IAS will enter the invasion pathway. This involves identifying the geographic origins of moths intercepted during vessel inspections in North American ports and implementing preventative measures in those foreign ports identified as the sources of moths. In the present work, we designed a genomic-based method for the accurate identification of the geographic origins of intercepted spongy moths. To this end, we developed an AmpliSeq panel, named SpongySeq, using genotyping-by-sequencing-derived SNP obtained from 1156 spongy moths collected at 61 sites in 25 countries.

The 283 SNPs that make up the panel were selected based on their performance to accurately assign spongy moths to one of the 19 geographic groups identified here, through assignment analyses using three different models, i.e., a multivariate approach, discriminant analysis of principal components (DAPC), and two supervised learning methods named Support-Vector-Machine and Naïve Bayes. With the most performant model (DAPC), our SpongySeq panel displayed a high assignment accuracy varying between 82 and 97%, depending on the assignment threshold used. Using this assignment method, an assessment of the origins of 28 egg masses of Asian spongy moths intercepted in different US ports in 2019–20, indicated that the majority were from Japan (18).

This research demonstrates the feasibility to predict provenance and mitigate invasion of an important invasive species using a medium-size subset of selected genetic markers.

The online version contains supplementary material available at 10.1186/s12864-025-11978-z.