Landcare Research - Manaaki Whenua

Landcare-Research -Manaaki Whenua

Using radio frequency identification technology to measure possum interaction rates with traps

RFID collar on a possum. Image - Sam Brown.

RFID collar on a possum. Image - Sam Brown.

All possum control and monitoring devices require possums to interact with them if they are to function as intended. In the field, it is obvious when a possum is captured, but not how often an individual visits a trap but avoids capture. Increasing the number of visits that result in capture is of obvious practical importance. To answer this question, Samantha Brown and her colleagues have been researching encounter and interaction rates of possums with traps and monitoring devices set in areas containing both uncontrolled (naïve) and controlled (potentially shy) populations.

The probability of capturing a possum is determined by the likelihood of it encountering a trap and then interacting with it. The probability of a possum encountering a trap is principally determined by trap density within that animal’s home range. Samantha focused her research on the probability of a possum interacting with a trap or monitoring device once it has encountered it and what can be done to ensure successful interactions follow encounters.

A possum fitted with an RFID collar at one of the trap sites.Initial research focused on how to increase the probability of capturing a possum in a leghold trap following an encounter. The study was conducted on farmland in South Canterbury that had been subject to infrequent, low intensity possum control. Sixteen possums were trapped and fitted with active RFID (radio frequency identification) collars. Collared possums were detected by an RFID scanner (developed by Landcare Research) and a PIR (passive infrared) movement sensor when they were 12 m and 3 m away from modified leghold traps set at 21 sites throughout the study area, and when they triggered the traps. The traps were locked open so possums were unable to be caught, which allowed them to make repeated visits to multiple trap sites. The behaviour of possums around traps was observed using motion-activated trail cameras. The interaction probability and behaviour of collared possums was compared between standard trap sets (i.e. that met the National Pest Control Agencies’ (NPCA) monitoring protocol); hazed (side fenced) trap sets, and covered trap sets.

All three types of trap sets had similar encounter rates; on around 60% of the occasions that possums came within 12 m of a trap, they came closer to investigate it (Table). The average number of trap encounters each night was 1.6, but individual possums visited traps up to 5 times per night. Eighty-one percent of the collared possums encountered traps within 3 nights of them being laid and set.

Table. Encounter and interaction probabilities for standard, hazed and covered leghold trap-sets.

Trap set Probability of an encounter given a detection at 12 m Probability of a capture given an encounter Probability of a capture given one or more encounters per night

(Encounter / Detection) (Interaction / Encounter) (Interaction / Encounter night)

Standard NPCA protocol1

0.66 (79/119)

0.21 (6/29)

0.33 (6/18)

Hazed (fenced)2

0.63 (95/151)

0.34 (10/29) 0.43 (10/23)


0.60 (73/121)

0.40 (10/25)

0.48 (10/21)

1 Trap hand-width from tree lured with flour
2 Trap with fence either side to focus possum onto trap
3 Trap covered with soil so it’s barely visible

The number of nights between a possum first encountering a trap and being ‘captured’ varied between 0 (i.e. caught on its first visit) and 6 (Fig.). However, the majority of possums (77%) got ‘caught’ on the first or second night they encountered a trap. The two possums that got ‘caught’ 4 and 6 nights after first encountering a trap (see Fig.) both visited the same trap on three different nights before ‘capture’.

The capture rate for each trap-set includes all encounters up to a possum’s first ‘capture’; subsequent ‘captures’ are ignored as in reality possums would not have the opportunity to be recaptured. Hazing and covering the trap increased the capture rate from 0.21 for standard sets to 0.34 and 0.4 for hazed and covered traps, respectively (Table). However, because of the small sample sizes these differences were not statistically significant and further trials are underway to test this relationship more rigorously. For a trapper, it is unimportant whether a captured possum has one or more encounters on any one night as long as it is caught, but taking multiple encounters before a capture into account still showed that hazed and covered traps had a higher capture rate than standard sets (Table).

Many possum visits to trap sites do not result in a capture, but altering the trap-set by hazing or covering it may increase the capture rate. Further work to better understand possum behaviour when encountering and interacting with detection and trapping devices will improve the cost effectiveness of control and minimise numbers of device-shy animals. Future trials will look at the effect of population density on capture rates; i.e. do encounter and interaction probabilities change when possum densities are reduced. The new information on improving capture and detection rates, particularly at low population densities, will contribute significantly to achieving the aspiration of Predator-Free New Zealand.

This study is funded by the Ministry of Business, Innovation and Employment.

Samantha Brown, Bruce Warburton, Jagath Ekanayake & Steve Hough