Buffer or Barrier?

Article by Sandy King

Following the article about “The Fence” from the PFR Governance Group in August’s SIN I thought I’d take a closer look at the difference between a buffer zone and a physical barrier, and the relative effectiveness of each when dealing with invasive predators. Conveniently, there are easily accessible examples of each – found in the work SIRCET are doing from Ackers Point to Golden Bay Road vs a low key DOC project on adjacent Native Island.

The GG article began with these statements:

One of the key principles for undertaking any eradication is that reinvasion can be managed. Some sort of barrier is needed to slow reinvasion, but so far the only barriers that have been shown to work are water (which is why eradications have happened on islands) and predator fences.

The important word to note here is “slow” – the idea that a barrier such as a fence will completely stop all reinvasion is outdated and unrealistic, we now know to expect some reinvasion in most cases (except perhaps where the barrier is extensive, such as the expanse of Southern Ocean surrounding Campbell and Macquarie Islands), and that the systems in place to manage reinvasion are essential and the key to a successful project.

The SIRCET project has no physical barrier, instead the project area has been divided into 3 zones, the Core, the Buffer, and the Community zone – effectively a buffer for the Buffer. I think every local is aware of the persistent hard work and dedication that has gone into the SIRCET trapping programme over the winter months. The Core area, from Ackers Point to Leask Bay/Evening Cove, and the Buffer now have lines 75m apart with a rat trap every 25m along those lines – a 75x25m grid. During an intensive trapping session those traps were checked daily for about 10 days in early autumn, every second day until June, and subsequently twice weekly. 623 rats were caught in the Core and Buffer zones between April and June.

Native Island, on the other hand, does have a physical barrier – albeit a bit “leaky”. The narrow stretch of water separating Native Island from the mainland is easily within swimming distance of rats and reinvasion is a certainty. The question is whether this leaky barrier is enough to slow reinvasion so that occasional re-invaders can be caught faster than they can breed.   Regular SINers might recall an article in January 2014 in which the Native Island project was explained – testing whether Goodnature self-resetting traps would get rid of rats and possums. Locals were invited to join the team on their monthly checks, and recently I took up the invitation and spent a morning with the guys checking traps and chatting about what they’d achieved. They have 146 Goodnature self-resetting rat traps set on lines 100m apart with a trap every 50m (a 100x50m grid), checked and serviced monthly. It is impossible to tell how many rats have been killed, most carcasses are either removed by weka or decompose in the interval between checks. However, they monitor progress using tracking tunnels (tunnels with inkpads and tracking paper, baited for one night each monitoring period. The results are presented as percent of tunnels showing rat tracks). They monitored rat numbers before traps were first set in November 2013 and got 84% tracking, which is “quite high”, and have since been monitoring rat numbers frequently.

The same monitoring technique has been used by SIRCET to monitor rats remaining in the project area for a number of years. In spite of huge numbers of rats killed monitoring still shows high tracking index results. More specifically, the intensified trapping effort in the Core area hasn’t decreased the tracking index to below 10% – a prerequisite target for the introduction of sensitive bird species such as robins (a core goal of SIRCET’s Halfmoon Bay Habitat Restoration Project). The following table compares tracking results for the SIRCET Core and Buffer zones with Native Island.

SIRCET – Core SIRCET – Buffer Native Island
November 2013 No data No data 84%
January 2014 No data No data 46%
March 2014 25% 30% No data
April 2014 12% 70% 17%
June 2014 15% 40% 30%
August 2014 No data No data 6%
Sept 2014 42% 56% No data


It would appear from this that even a “leaky” barrier is enough to slow rat reinvasion to Native Island to the point where it can be managed; even with a wider trap grid, less frequent checking and an unproven type of trap, although it remains to be seen whether the downwards trend continues and is maintained.

SIRCET has achieved a lot over the past 12 years as a result of its trapping effort; during annual bird call counts numbers have consistently been higher inside the project area than out, and monitoring shows improvements in the vegetation. However one of their aims is to reintroduce some of our rare, precious and vulnerable species – to have saddleback/tieke resident in our gardens, and to do this the rat tracking index needs to be consistently 0%, i.e. rat numbers are so low they are undetectable. Saddleback and rats cannot co-exist and there are plenty of examples to prove this. When I talked to Trustees their frustration at not yet being able to attain their ultimate goal was palpable.  The question of whether a barrier would serve their purpose better than a buffer was met with wistful looks and muttered comments about the cost of a fence. However in the long term the initial cost to erect a fence may be comparable with, or even less than, the cost of maintaining the effort that went into trapping from April till June which resulted in an encouraging decrease in rat abundance in the Core zone, not to mention the unquantifiable costs of having your ultimate goal always somewhere beyond the horizon.

Thanks to SIRCET for sharing their data and thoughts, and for their efforts to enhance our local environment, and thanks to DOC for the data.


  • Tom

    Really enjoyed reading this. Quite a lot of useful information here for future pest control programmes around the country!