Rodent Detection Monitoring

Following an eradication attempt it is necessary to confirm the success of the operation and to prevent reinvasion. The level of confidence in determining whether an eradication was successful or not, and detecting new invasions is dependent on the type and density of detection devices, duration of deployment, along with the density of rodents present.

Traditional approaches (particularly for aerial eradications) for declaring success have been to wait until at least two rodent breeding seasons (i.e. two years) have passed before undertaking monitoring (Russell and Broome al, 2016). This period allows rodent densities to build up to detectable levels in the event that the operation failed. If no surviving rodents are found after at least two breeding seasons (roughly two years), then the eradication is declared a success. This traditional approach has potential downfalls in that it does not facilitate rapid response to early detection of survivors, and thus obligates repeating the eradication from scratch. An alternative “Rapid Eradication Assessment” approach is to monitor the island at some fixed time soon after the eradication and quantitatively estimate, whether the eradication was successful or not through a spatial-survey model (Samaniego-Herrera et al., 2013). This approach facilitates the early detection and removal of localised survivors in comparison to a complete repeat of the eradication operation. Additionally, if confidence in eradication success is determined earlier, restoration plans can be implemented before the traditional two year mark (Russell and Broome al, 2016).

The differences in scale and topography on LHI for areas treated by aerial application compared to the areas treated by ground based methods (i.e. hand broadcast or bait stations), present an opportunity to implement a combination of the above methods to maximise any chances to remedy a possible failure and increase confidence in eradication success. Therefore rodent detection monitoring will be undertaken in different areas at different scales and intensities on LHI. Monitoring in the following phases is described in more detail below.

  • Initial follow up monitoring
  • Monitoring to declare eradication success
  • Ongoing monitoring for detection of reinvasion from rodents

A range of tools are available for trying to detect rodents at low density however they all have their limitations so in order to maximise the chances of detecting any survivors it is desirable to use a mixture of techniques. Details of these tools are provided below.

Initial Follow up Monitoring

Due to the scale and topography of most of the areas on LHI that are to be treated by aerial broadcast, it is not realistic to try and detect with a sufficient degree of confidence any rodents surviving the eradication in those areas immediately after eradication. A failure in areas treated aerially means that there has been some failure in planning or implementation, and the ability to undertake any meaningful immediate response if survivors were detected in those areas is limited. Therefore very limited monitoring will be undertaken immediately after the eradication (restricted to some easily accessible areas) in areas treated aerially. This will be best achieved after several breeding seasons have passed allowing potentially surviving rodents to build up to detectable numbers.

However, the area around the settlement does offer an opportunity to undertake a high standard of rodent monitoring and to respond promptly to any survivors detected. This is due to the logistical feasibility of the area i.e. size, topography and access. Importantly, this area warrants the extra attention as it has the highest likelihood of failure given it will be treated by a combination of ground methods.

Russell et al. (2016) have tested a statistical model developed for the Rapid Eradication Assessment approach by Samaniego-Herrera et al. (2013) for assessing the probability of detecting surviving rodents and their offspring, using a grid of detection devices to predict eradication success. They found that spacing of detection devices and number of monitoring nights provided the best predictors for eradication success.

Preliminary modelling for LHI undertaken by Samaniego-Herrera and McClelland (2016) using the Rapid Eradication Assessment approach has suggested that a detection device network grid spacing of 30 m x 30m in the settlement area immediately after eradication would produce a median Confidence Interval of 100% (lower CI 99.1% and upper CI 100%) of detection when monitored for at least 30 nights.

Based on the modelling, the settlement area and other easily accessible areas of LHI has been monitored intensively for the presence of rodents throughout the baiting operation. Detection of surviving rodents has been assessed by a combination of detection tools described below. All detection devices are checked frequently so that a targeted response can rapidly be undertaken, i.e. to maximise the likelihood of any rodents that is detected being in the same area so that it can be targeted with the preferred technique –traps or toxicants. Residents are also being asked to report any evidence of rodent activity to the project team. In addition, trained detector dogs and handlers are being deployed throughout the settlement area in the last month of baiting to search for signs of and locate any surviving rodents.

This approach will give a high level of confidence that eradication success has been achieved for the settlement area.

Declaring Eradication Success

During the period immediately following the eradication and in the lead up to declaring eradication success, a the monitoring network will be adjusted in the following ways.

  • The network within the settlement area will most likely be maintained but checked at reduced frequency potentially weekly or fortnightly.
  • The permanent rodent detector / biosecurity dog based on the island will sporadically undertake targeted searches of high risk areas.
  • The declaration will be preceded with a thorough search using an additional contract team of rodent detector dogs and handlers to search all accessible areas of the island for rodents.

This methodology will give a high level of confidence to allow declaration of eradication success which will be declared after two years of monitoring with no rodent activity.

Ongoing Rodent Detection Monitoring

The eradication investment will be protected through ongoing rodent detection monitoring on the island aimed at detecting any possible reintroductions. Additionally, there will be pre border inspections at the Birdon wharf and Eastern Air hangar of freight being brought to the island for each voyage of the Island Trader. This will form part of the island’s permanent rodent detection and incursion response system initiated as an integral part of the island’s Biosecurity program which will be upgraded in parallel with the REP. The monitoring network developed for the initial follow-up monitoring and declaration of success will be modified to allow targeted monitoring of high risk reinvasion points. It will include:

  •  A grid of 155 rodent hotels containing rodent detection and kill devices at high risk reinvasion points such as the wharf and airport and potential areas for initial recolonisation. These will be checked at a frequency commensurate with arrivals (i.e. daily at the airport and fortnightly at the wharf coinciding with cargo vessel arrivals) and risk level
  •  The permanent rodent detector / biosecurity dogs and handlers based on the island will routinely screen all incoming cargo and luggage
  • The permanent rodent detector / biosecurity dogs and handlers based on the island will sporadically undertake targeted searches of high risk and random areas

A biosecurity dog handler and dog will undertake pre border inspections of freight at Birdon wharf and Eastern Air hangar for each voyage of the Island Trader.

This methodology will allow a high level of confidence that any reinvasion would be detected. Genetic testing on LHI rodents has been undertaken. In the event that rodents are detected post REP, the genetic samples will allow determination of whether the eradication failed or the detection was a reinvasion.

Detection Response

In the unlikely event that rodents are detected, remedial action will be considered to eliminate them.

In the event that possible sign is detected, trail cameras with a variety of baits will be deployed in the area to try and confirm if a rodent is present. At the same time an array of other detection devices chew cards, wax block, tracking tunnels, bait stations and traps will be deployed in the vicinity and any rodent dogs available will be focused on that area. Any response will need to be carefully planned and implemented as previous experience has shown that if not done properly there is a risk of not locating the animal or even scaring it away from the known area.

Due to the wide number of situations that could involve a rodent being detected /confirmed e.g. unconfirmed sign, single confirmed individual, multiple individuals, animals around in buildings etc., it is not realistic to develop comprehensive response scenarios. Instead a Technical Advisory Group (TAG) will be set up who will be on immediate standby to provide consensus advice on how to respond to any specific situation. The TAG will consist of selected experts in eradication techniques, rodent detection and rodent behaviour. Additionally, there has been a revised Incursion Response Plan developed which includes the location and monitoring schedule for regular periods and escalation plans should sign be detected.

Additional detection and response devices will be held on island to facilitate a rapid response if one is required.
A second eradication attempt using aerial techniques is not part of this proposal.

Detection Tools

Bait Stations

There is a network of bait stations present around much the settlement as part of the eradication. However any rodents surviving the initial eradication operation will have, for some reason, avoided eating bait to which is likely to mean they haven’t entered a bait station either because: a) they have avoided the stations (neophobia), or b) they haven’t had a station within their territory. The only way to use bait stations to detect rodents would be to put additional stations in any possible gaps in coverage i.e. reduce the spacing between stations. It is more effective simply to make sure that the initial coverage is adequate.

While the stations can, depending on the design, be used for the deployment of other devices e.g. tracking tunnels or to protect wax blocks, this would assume that the rodents are not avoiding the stations themselves possibly due to inter or intra species competition. Unless other devices are placed in the stations the only likely rodent activity that will be recorded in the bait stations will be bait being eaten i.e. tooth marks. If there is still bait take continuing at any bait stations then the response is effectively part of the eradication, i.e. keeping the stations topped up.

Rodent Detection Dogs

Trained rodent dogs are a highly effective tool for locating rodents. They have an advantage over other tools in that they actively seek out the rodent rather than requiring the rodent to come to them. Also the rodent and dog do not need to be in the same place at the same time as the dog will, within limits, detect where the rodent has been. Detector dogs can also cover an area once to get a result whereas all other techniques need to be set up and then checked periodically.

Detector dogs on LHI are being used in the following ways:

  • A team of specialist rodent detector dogs and handlers are being deployed to actively search the settlement area to provide immediate detection capability for any surviving rodents in this area. This is undertaken during the last two months of baiting.
  • Two permanent rodent detection and general biosecurity dogs (and handlers) have been trained and are permanently based on the island since August 2017. The dogs provide a rodent detection capability at the border (airport and wharf) as well as generalised biosecurity capabilities.
  • A team of specialist rodent detector dogs and handlers will be deployed to actively search all accessible areas (particularly the settlement area) two years following the eradication to provide evidence for the declaration of eradication success. The exact number of dogs and handlers will be determined in consultation with the selected tenderer for this service.
  • If the dogs indicate the current or recent presence of any rodents other techniques will be used confirm, locate and kill the individual(s).

Minimum training, accreditation and ongoing certification requirements for both dogs and handlers has been developed prior to implementation of the REP. This includes:

  • Ability to identify target scent and avoidance of non-target species and scent
  • High level obedience and control
  • Good temperament around people and other dogs
  • Initial and ongoing assessment and certification of dog and handler

Figure 20 Rodent Detection Dog Examples

Trail Cameras

Trail cameras / remote cameras come in a variety of specifications e.g. natural light / infrared, stills or movie and are widely used to detect various species of wildlife around the world. Trail cameras have been shown to be very effective for confirming the presence of rodents but are only available in limited numbers for the LHI operation due to the cost ($300 + per camera), and the time required to set up and maintain them, i.e. to check all photos. As such cameras are best used when there is a preferred location, particularly where the presence of a rodent is suspected in an area but unable to be confirmed.

Chew Blocks / Wax Tags

Chew blocks/wax tags are peanut butter flavoured wax blocks with a smooth surface. When an animal bites one it leaves incisor tooth marks which can usually be identified to species (mouse/rat). Chew blocks are cheap to purchase or can be made on site and are easy to deploy and check. However while they have proved very useful for locating rats and have been widely used for mice, potential issues have been raised with identifying mouse chew sign which warrant further investigation. Their low cost and ease of use means that they are a very useful detection tool in the Lord Howe situation.
Chew blocks can be bought commercially from Pest Control Services in New Zealand (NB only peanut butter flavoured tags should be used for rodents) or they can be made by project staff . Blocks are placed 4 cm above the ground to facilitate access by mice.
There can be issues with non-target interference which need to be checked for the site, however with the absence of other mammal species on Lord Howe this is considered to be a very low concern.
Chew cards are pieces of corflute cardboard with peanut butter pressed into the holes. The standard design is a 9 x 18 cm card made of 3 mm white plastic corflute. When the rodent attempts to get to the peanut butter it leaves distinctive chew marks on the corflute which can be identified as rat or mouse. Chew cards are cheap, effective for both target species although somewhat less for mice than rats. Depending on the specifics of the devices used it can also include an ink card to try and get footprints.
Mice and rats have similar bite marks that are mainly distinguished on size. They leave pairs of incisor marks, nearly straight lined on top and more curved underneath. Incisor pairs are about 1 mm across for mice (less than half the width of the corflute channels) and about 2 mm across for rats (more than half the width of a corflute channel). Look for individual bites clear of continuous chewing along card edges. Rats frequently chew large chunks out of the cards leaving a relative cleanly cut edge. Mice usually chew small amounts, sometimes making just small scattered nicks along the edge, or chew short channels between card partitions on just one surface. Continuous mouse chewing along the card edge also tends to be less cleanly cut than for rats, with a short chewed flange attached to the remaining card with numerous light tooth impressions beyond that, as opposed to cleanly cut edges frequently made by rats.

Figure 21 Chew Block

Figure 22 Examples of Chew Cards and Evidence of Rodent Damage


All traps used (snap traps, Good Nature traps) are deployed as an additional detection and kill method as a response to active sign or suspected rodent sightings.

Traps have the major advantages over the other techniques of both killing the survivor and providing a body which can then be examined for species, age, sex and breeding status i.e. a female that has bred is of far more concern than a lone male. However they have several disadvantages:

  • They are labour intensive, both to set up and to monitor – NB every trap has to be set with care as each one needs to be considered as THE trap that will catch the rodent.
  • They are generally more expensive to purchase.
  • They are generally species specific i.e. rat versus mouse, so you need to effectively set pairs of traps.
  • There is a non-target risk with kill traps, particularly with rat traps i.e. they need to be set under covers to reduce the risk.

The most common and simplest kill traps are snap traps which are lightweight and relatively cheap. There are concerns that some rats may escape from snap traps, however most of the concerns with snap traps relate to large Norway Rats (Rattus norvegicus) so are not a significant issue for LHI.

There are multiple variations of the snap trap and care is taken to select the most suitable one – the Victor treadle trap with a yellow plastic treadle is the preferred option as unlike most other traps the rodent only needs to inspect the bait to set it off, whereas for most snap traps the animal needs to actively chew on the bait. The double spring on the rat trap also gives greater killing power.

When used inside, as long as there is no risk to children, the traps can be set without a cover. It is important to use a cover when setting kill traps outside to minimise the risk to non-targets. The covers on LHI are made of plywood with exclusion bars over the main entry to prevent access by birds and other non target species. Woodhen, currawong and banded rail have potential to interfere with traps in order to access the bait so that even if they are not caught they will make the trap non-functional until it is reset. Also it is likely that once any of these birds learns that the traps are a food source they will target them. Bait for the traps is highly variable but peanut butter with fish oil and rolled oats to bind it is the standard bait. No non target species have been affected by traps on LHI.

There is currently a self-resetting trap available (Good Nature A24). While these have major benefits (targeting multiple individuals as they don’t need to be reset) they are very expensive. As such, the few have been utilised on LHI are used only in strategic locations.

Tracking Tunnels

Tracking tunnels come in a variety of designs from semi-permanent wooden structures to lightweight plastic. Rodents are known for entering tunnels but the tracking tunnels are usually baited/lured to act as an added attractant. Inside the tunnel is a plain card with an ink source- either inked card or an ink tray set up so that any animal that walks through it will leave footprints on the card which can then be identified to species. To reduce the risk of neophobia the tunnels, minus tracking cards should be put in place a couple of weeks prior to the planned activation period.

There is a likelihood of currawong, rail and any woodhen interfering with the tunnels to access the bait after they have been released from captivity. The design of the tunnel is set to reduce non-target interference while still allowing easy access to rodents. This is important as making the entrance small enough to prevent entry by non-targets may effectively deter the target species.

The cost of using tracking tunnels is in a large part dependent on the servicing regime as they can be left for several days between checks if required, however this reduces the likelihood of being able to mount an effective response to any detection as the individual may have moved prior to detection.

Figure 25 Example Tracking Tunnel and Foot Print Evidence

Improved Biosecurity

To improve Biosecurity on the island more generally and to protect the rodent eradication investment, the LHIB is updating the Island’s Biosecurity system concurrently with the proposed REP. In 2015 a consultant was engaged to review and update the LHI Biosecurity Strategy. Recommendations from the updated Strategy (AECOM, 2015) include:

  • reducing risk at the Port Macquarie wharf
  • increasing education and awareness for residents and visitors pre arrival to LHI
  • increasing inspection regimes for all pathways
  • pursuing legislative declaration of LHI as a Special Biosecurity Zone under the Biosecurity Act 2015
  • increasing residents’ awareness of biosecurity risks of plants, animals and diseases both before and after import
  • being prepared to react quickly to new incursions through early detection and rapid response
  • continuing with on ongoing management and eradication programs
  • ensuring biosecurity is adequately resourced with realistic cost and resource estimates

Specifically, in relation to rodents and in addition to the ongoing rodent detection measures described in Section 2.6.3, the following measures are being applied:

  • Employment of a dedicated on island biosecurity officer who will have primary responsibility for the ongoing rodent detection network. This role may be combined with the rodent / biosecurity detector dog handler
  • Upgrades to the shipping contract to increase emphasis on rodent prevention including requirements to:
    o have in place a Biosecurity Management Plan
    o maintain rodent baiting at the point of mainland departure
    o maintain rodent baiting and de-ratting certificates on the cargo vessel
    o report biosecurity risk cargo and incidents prior to arrival.