Global and regional evaluation of Corythucha marmorata distribution under different spatial modeling conditions

A Tiny Bug With a Big Global Footprint

Most people never notice the chrysanthemum lace bug, a speck‑sized insect that feeds on garden flowers and crops. Yet this small hitchhiker can ride trade routes and highways to new continents, damaging plants and costing money to control. This study asks a practical question: where in the world is this invasive pest most likely to thrive now and in the future, and how can scientists best make those predictions so that governments and farmers can act in time?

Why Forecasting Pests Matters

Invasive species increasingly disrupt local wildlife and agriculture, from crowding out native plants to reducing food for insects and birds. Once a species like Corythucha marmorata, the chrysanthemum lace bug, arrives, it is very hard to remove. Because it has already spread from its North American home to Japan, China and South Korea, officials need tools that highlight high‑risk regions before outbreaks explode. The authors use species distribution models, which link known sightings of a species with climate and environmental data to estimate where conditions are suitable. These maps help focus surveillance, quarantine, and control on the areas that matter most.

How the Researchers Built Their Maps

The team compiled over a thousand lace bug records from a global biodiversity database and field surveys, then carefully filtered them so that clusters of points did not bias the results. They combined these occurrences with climate layers that summarize temperature and rainfall patterns across the globe, such as how much it rains in the warmest season or how strongly temperatures swing between seasons. They then ran ten different modeling algorithms, from classic statistical approaches to modern machine‑learning methods, under three ways of defining where the insect is assumed to be absent. Because true absence is rarely known, they created so‑called “pseudo‑absence” points by different rules, such as scattering them randomly, restricting them to areas with very different environments, or placing them in rings at set distances from known occurrences.

Blending Many Models Into a Single Picture

Rather than trust any single method, the authors built ensemble models that combine the outputs of multiple approaches. They tested four ways of averaging the models, including a simple mean, a median, a committee‑style voting system, and a weighted average that gives more influence to better‑performing models. They judged accuracy using two standard scores that measure how well models separate suitable from unsuitable areas. The strongest results came from committee averaging and from weighted averages built with pseudo‑absence points drawn from environmentally contrasting areas. These combinations produced very high accuracy scores, showing that careful choices about how to represent “absence” and how to weight individual models can substantially sharpen predictions.

Where the Lace Bug Is Most Likely to Spread

Using their best‑performing ensembles, the researchers mapped the lace bug’s potential distribution worldwide and zoomed in on South Korea. Globally, the models highlight high suitability not only across the insect’s known range in North America and East Asia but also in yet‑uninvaded regions, including parts of Europe, eastern Australia, Uruguay, and Argentina. Within South Korea, most inland areas appear suitable now and remain so in projections for 2050 under a strong climate‑change scenario. The southern island of Jeju, by contrast, stands out as consistently unsuitable, matching current field observations that the pest has not taken hold there. Areas with intense traffic and roadside vegetation emerge as likely hotspots, reflecting how vehicles help the bug move between patches of host plants.

What This Means for Managing Future Invasions

For non‑specialists, the main takeaway is that how we build prediction tools matters as much as which insect we study. By testing many modeling recipes, this work shows that blending models and carefully choosing where to assume a species is absent can yield reliable, cross‑scale risk maps from global to national levels. For the chrysanthemum lace bug, those maps warn that much of temperate, seasonally wet regions around the world could support future invasions, while confirming that most of South Korea will remain at risk except for a few refuges such as Jeju. More broadly, the study offers a practical blueprint for forecasting other invasive pests before they arrive, allowing monitoring and control efforts to be targeted where they will do the most good.

Citation: Byeon, Dh., Lee, WH. Global and regional evaluation of Corythucha marmorata distribution under different spatial modeling conditions. Sci Rep 16, 13283 (2026). https://doi.org/10.1038/s41598-026-42897-y

Keywords: invasive species, species distribution modeling, chrysanthemum lace bug, climate suitability, ensemble ecological models