Spatiotemporal Patterns of Invasive Mammals in a Wetland Environment: Trialling an Automated Radio Telemetry System to Track Rattus spp.

Invasive species are a core driver of global biodiversity loss, with a disproportionate negative effect on island ecosystems. This is notably true in Aotearoa New Zealand, where every terrestrial ecosystem has been detrimentally impacted by invasive mammalian species. Out of the 25 actively managed invasive mammal species in Aotearoa, Rattus spp. inflict widespread damage. Although widely studied in some habitats, little is known about invasive mammals in wetland environments. This thesis aimed to characterise the community of invasive mammals in a wetland environment with a specific focus on the behaviour and movement of Rattus spp., while trialling a novel Automated Radio Telemetry System (ARTS).

This study first utilised two passive monitoring techniques, tracking tunnels and camera traps, to characterise the community of invasive mammals present in Wairio Wetland, Wairarapa. Trap catch data from the past decade indicated high abundance of hedgehogs, rats, and mice, but low abundance of possums, mustelids, and feral cats. Data gathered in the present analysis contradicted these findings, with high incidence of possum, mice, and ship rats but low incidence of hedgehogs. These differences can likely be attributed to monitoring methodology, where previous indications of abundance were determined through trap data, while these analyses using cameras were more robust and catered to the entire community. There were large differences in the time-of-day species were active between seasons, most species becoming active earlier in the night over winter but continuing their activity later into the morning in summer. Canopy cover did not influence differences in community throughout the wetland, but the presence of differences in community composition suggests intricate ecological relationships shaped by multiple unmeasured factors.

Chapter Three investigated the construction and functioning of an ARTS. Manufactured by Cellular Tracking Technologies (CTT), an ARTS was deployed as a network of 20 receiving and transmitting nodes, a SensorStation, antennae, and a power supply. This was trialled in the wetland environment and stationary transmitting CTT PowerTags were tested in situ to determine signal strength calibration, derive a method for localising tags, and measuring the accuracy and precision of tag localisations. The ARTS was sufficiently powered by an 80W solar panel and a 12V battery and was able to withstand local weather conditions. However, pest proofing the system was required after damage from rats. The error on tag localisation points were 51.87 ± 5.45 metres (80% CI). The root mean square error for localisation accuracy was 22.31m to 30.57m. It is recommended that a greater number of node and tag calibration trials are conducted in future studies to increase the site-specific accuracy and precision of localisations.

Chapter Four employed the ARTS to track ship rats (Rattus rattus) and Norway rats (Rattus norvegicus). After extensive live trapping one individual of each species was collared with an activated CTT PowerTag. Data was automatically received by the SensorStation and uploaded to an online portal for post-processing. Rat #1 (ship rat) yielded 53,681 transmitter detections over 1 day, while Rat #2 (Norway rat) yielded 271,492 transmitter detections over 11 days. Combining transmissions into minute-by-minute intervals and using those minutes where three or more nodes detected a transmitter, estimates of spatial locations were determined. From these fixes, Kernel Density Estimates (KDE) of home ranges were calculated, of Rat #1 90% KDE was estimated at 3,360m2 and for Rat #2 at 17,664m2. Clearly identifying areas of the individuals spatial use within the wetland and providing the first known home ranges estimates of Rattus spp. in an Aotearoa wetland.

This study is one of very few to characterise the community of mammals present in a New Zealand wetland, while providing insights into Rattus spp. movement and home ranges in such an environment. The results have important implications for effective, targeted pest control management. The novel CTT technology that was trialled here has the potential to enhance conservation research in Aotearoa, if improvements in technology and data reliability can be developed, and enhanced in situ calibration of signal strength is implemented.