New research predicts wildlife trafficking hotspots

Airports are places most of us associate with family trips, business travel, or maybe the occasional lost suitcase. But for traffickers in the illegal wildlife trade, airports are something very different: lifelines in a vast global network that quietly moves rare plants, exotic animals, and their body parts across continents.

A study published in Communications Earth & Environment, a journal from the Nature portfolio, shines a spotlight on this hidden world. Led by Hannah Murray and Bistra Dilkina of the University of Southern California, together with Meredith Gore of the University of Maryland, the research takes on a puzzle that has long stumped conservationists and policymakers: how to find the real choke points in wildlife trafficking when so much of the trade remains invisible.

Their approach blends network science (think maps of how airports connect like webs of spider silk) with machine learning, allowing them not just to analyze where trafficking has been caught, but to predict where it’s likely happening undetected. The results reveal surprising hotspots, including major airports in the United States, Italy, and China, that rarely show up in official seizure data.

Before diving into the technical sleuthing, let’s step back. The illegal wildlife trade, or IWT for short, is a multibillion-dollar black market. It involves everything from pangolin scales and elephant ivory to rare orchids and parrots. This isn’t just a conservation issue, it’s part of what the United Nations calls the “triple planetary crisis,” alongside climate change and pollution.

When species vanish, ecosystems unravel. Local communities lose sustainable livelihoods. Criminal networks profit while governments lose control of natural resources. And because the trade often overlaps with other crimes, like smuggling drugs, timber, or even people, it has ripple effects on national security and global stability. The tragedy? There’s little evidence the illegal wildlife trade is slowing down.

If we want to stop traffickers, the obvious first step is to follow their footprints. But here’s the catch: most of the data we have on wildlife trafficking comes from seizures, basically, when authorities get lucky enough to catch something in transit.

That’s like trying to understand global shipping routes by only looking at the few cargo containers customs decides to check. It creates a distorted picture: some countries look like major players simply because they report more, while others disappear from the map because there’s little monitoring.

Only a third of recorded incidents even include enough detail to trace the journey from origin to transit point to destination. The rest are fragments. So, what can science do with such patchy, biased data?

This is where Murray, Gore, and Dilkina brought in tools more often used to study drug cartels or even social media networks. Instead of just tallying incidents, they mapped airports as nodes in a global web of flights, then asked: which airports matter most for traffickers?

Two key ideas come into play here. Degree centrality means how many direct connections an airport has. Big hubs like Bangkok or Addis Ababa naturally have lots. Betweenness centrality instead measures how often an airport acts as a bridge between two others. Think of Dubai or Hong Kong as middlemen that connect faraway routes.

Their analysis confirmed some familiar patterns: Africa supplies much of the wildlife, Asia drives demand, and places like Bangkok, Hong Kong, and Nairobi stand out as major nodes. But here’s the twist: some airports that don’t show many seizures, like Moscow or Luanda, still turn out to be critical bridges in the network. Shut those down, and you could throw a wrench into traffickers’ operations even if you never catch a single shipment there.

The researchers didn’t stop there. They trained a machine learning model using data from nearly 2,000 airports worldwide, feeding it not just flight connections but also information about crime levels, governance, and the presence of groups that can hold governments accountable.

The model’s predictions were striking. It flagged 307 airports as likely trafficking hubs that don’t currently show up in seizure data. Among the most confident calls: Dallas–Fort Worth and Denver in the U.S., Milan’s Il Caravaggio in Italy, and airports in China, Indonesia, the Philippines, and Mexico.

Some of these predictions were later backed up by independent reports, for instance, rhino horn shipments caught at Shanghai’s Hongqiao Airport, which wasn’t in the original dataset.

This research could be a game changer for policymakers or a wildlife enforcement officers. Instead of playing whack-a-mole with seizures, they can start targeting airports that matter most in the bigger picture.

That could mean prioritizing scarce resources for inspections where they’ll do the most good, building partnerships with civil society groups to increase oversight in vulnerable regions, and recognizing that even countries not traditionally seen as trafficking hotspots, like the U.S. or Italy, play bigger roles than previously acknowledged.

In short, the study offers a roadmap for smarter, more proactive interventions.

What’s powerful about this work is not just the findings but the method. By “encoding” the wildlife trade into a network and then “decoding” it with predictive modeling, the team demonstrates a way around the limits of messy, incomplete data.

If you are traveling soon, when at the airport watching planes taxi and passengers line up for boarding, remember: that same web of routes can be a lifeline for traffickers, or, with the right strategies, a net that catches them.

If you want to learn more, read the original article titled "Encoding and decoding illegal wildlife trade networks reveals key airport characteristics and undetected hotspots" on Communications Earth & Environment at http://dx.doi.org/10.1038/s43247-025-02371-5.