Invasive species elimination: delimiting core and buffer management zones

In the world of conservation, there is a growing ambition to do more than just protect a small patch of land from predators; the goal is to completely remove invasive mammals like rats and stoats from vast, open landscapes. This concept, known as landscape-scale elimination, differs from traditional island eradication because the land is not walled off. Instead, it is an open system where animals from the surrounding countryside constantly try to cross back in. To succeed, managers must create a "buffer zone" around the protected core area. This buffer acts as a wide net to catch invaders before they reach the heart of the sanctuary. The challenge has always been figuring out exactly how wide this net needs to be. If the buffer is too narrow, rats slip through; if it is too wide, the cost of trapping and poisoning becomes unsustainable. For decades, planners have had to guess the right distance, often relying on rough estimates of how far a rat might travel.

A team of researchers working on the Predator Free South Westland project in New Zealand decided to stop guessing and start measuring. They managed a massive, unfenced sanctuary covering 101,164 hectares of forest, farmland, and alpine terrain on the country's South Island. Their target was the ship rat, a destructive species that threatens native birds and forests. After successfully removing the vast majority of rats from the area, the team faced the critical task of maintaining that success. They needed to know how deep into the sanctuary rats would travel if they tried to reinvade, so they could design a buffer zone that was both effective and efficient. To find the answer, they deployed a network of 1,254 cameras across the landscape, creating a digital eye that watched for movement day and night.

The researchers tested two different ways of drawing the line for their buffer zone. The first method was simple and traditional: they measured the straight-line distance from the project's outer borders. This approach assumes that a rat could come from any direction and that the distance from the edge is the only thing that matters. The second method was more nuanced, looking at the actual shape of the land. The team recognized that rats do not wander randomly across mountains and rivers; they follow specific paths. They treated coastlines, major rivers, roads, and lake edges as natural corridors that guide movement. By mapping the cameras relative to these landscape features rather than just the project border, they created a "landscape-based" buffer.

The results from a full year of monitoring, spanning from June 2025 to May 2026, revealed a striking difference between the two approaches. When using the traditional border-based method, the team found they would need a buffer zone extending 4,130 meters inward from the edge to catch 90% of the rats that were detected. However, when they used the landscape-based method, the required distance shrank dramatically to just 1,513 meters. This reduction of over 60% means that managers could focus their resources on a much narrower strip of land while still intercepting the same number of invaders. The data showed that rats were not spreading out evenly across the landscape; instead, they were funneling along the rivers, roads, and coastlines. By aligning the buffer with these features, the team could predict where the rats would appear with much greater precision.

The study also looked at what happened when rats were detected deep inside the protected core area, far from the edges. Under the traditional border model, these deep detections appeared random and unexplained, often occurring thousands of meters from the nearest border. This made it difficult for managers to understand why the rats were there or how to prevent them. In contrast, the landscape-based model provided a clear explanation for most of these deep incursions. The researchers found that 70% of the rats found in the core were actually the result of delays in managing the buffer zones near the landscape features. When baiting or trapping was paused near a river or road, rats moved further inland. Once these management gaps were fixed, the unexplained detections dropped significantly. This suggests that the landscape model not only saves money by shrinking the buffer but also helps managers understand the specific reasons for failures in the core area.

Seasonal changes also played a role in how far the rats traveled. The data showed that the distance required to catch the majority of rats increased during the summer and autumn months. This timing coincides with the breeding season for rats and a natural event called a "mast seeding," where trees produce an abundance of seeds, causing rat populations in the surrounding unmanaged areas to explode. During these peak times, the pressure on the sanctuary borders intensified, and rats ventured further inland. However, even during these high-pressure periods, the landscape-based model remained more stable and reliable than the border-based one, suggesting that following the land's natural features provides a consistent guide for management regardless of the season.

The success of this study relied heavily on the technology used to gather the data. The team utilized a network of cameras that could report sightings in near real-time, rather than waiting for field crews to physically collect memory cards from the devices. This allowed them to see patterns as they happened and adjust their management strategies quickly. The data confirmed that rats detected deep inside the sanctuary were usually single individuals passing through, rather than established families. The frequency of sightings dropped sharply the further one moved from the edges, indicating that the core area was effectively rat-free, with only occasional intruders slipping through the gaps.

This work provides a new blueprint for protecting large, open landscapes. It demonstrates that by understanding how animals actually move through the environment—following rivers and roads rather than moving in straight lines—conservationists can design much smarter and cheaper protection zones. The findings suggest that the old way of drawing a uniform circle around a project is less effective than drawing a line that follows the contours of the land. For the Predator Free South Westland project, this means they can maintain a rat-free sanctuary with a more focused effort, ensuring that the rare and endangered wildlife living in the core remains safe from the constant pressure of reinvasion. The study does not claim to have solved the problem of invasive species forever, but it offers a proven, data-driven method to make the job of keeping them out significantly more manageable.