Survival and dispersal data for eastern quolls (Dasyurus viverrinus) reintroduced to Mulligans Flat Woodland Sanctuary, Australian Capital Territory, Australia, 2016-2018

SUMMARY

Threatened species recovery programs are increasingly turning to reintroductions to reverse biodiversity loss. Here we present a real-world example where tactics (techniques which influence post-release performance and persistence) and an adaptive management framework (which incorporates feedback between monitoring and future actions) improved reintroduction success. Across three successive trials we investigated the influence of tactics on the effective survival and post-release dispersal of endangered eastern quolls (Dasyurus viverrinus) reintroduced into Mulligans Flat Woodland Sanctuary, Australian Capital Territory. Founders were monitored for 42 days post-release, and probability of survival and post-release dispersal were tested against trial, origin, sex, den sharing and presence of pouch young. We adopted an adaptive management framework, using monitoring to facilitate rapid learning and to implement interventions that improved reintroduction success. Founders released in the first trial were less likely to survive (28.6%, n = 14) than those founders released the second (76.9%, n = 13) and third trials (87.5%, n = 8). We adapted several tactics in the second and third trials, including the selection of female-only founders to avoid elevated male mortality, and post-mating releases to reduce stress. Founders that moved dens between consecutive nights were less likely to survive, suggesting that minimising post-release dispersal can increase the probability of survival. The probability of moving dens was lower in the second and third trials, for females, and when den sharing with another founder. This study demonstrates that, through iterative trials of tactics involving monitoring and learning, adaptive management can be used to significantly improve the success of reintroduction programs.

STUDY SITE

Mulligans Flat Woodland Sanctuary (MFWS) is a 485 ha reserve containing critically endangered yellow-box Eucalyptus melliodora and Blakely’s red gum Eucalyptus blakelyi grassy woodland and is situated in north-east Canberra, ACT Australia (-35.166543, 149.157946). MFWS is enclosed by predator-proof fencing to exclude non-native animals such as red foxes (Vulpes vulpes), cats (Felis catus), European rabbits (Oryctolagus cuniculus) and European hares (Lepus europaeus), which have been eradicated within the exclosure. The MFWS fence design includes a ‘floppy top’ which prevents introduced predators from climbing into the sanctuary but does not prevent animals from climbing out into the surrounding landscape. MFWS, and the adjoining Goorooyarroo Nature Reserve, are used as an ‘outdoor laboratory’ and form the location of the Mulligans Flat-Goorooyarroo Woodland Experiment (MFGO Experiment, www.mfgowoodlandexperiment.org.au). The experiment aims to restore biodiversity and ecological function to this critically endangered box-gum grassy woodland community.

METHODS

We reintroduced the eastern quoll into MFWS in a series of three trials over three years (Trial 1 in 2016, Trial 2 in 2017, and Trial 3 in 2018). To maximise genetic diversity, founders in the Trials 1 and 2 were selected from both captive-bred and wild populations, and in Trial 3 only wild founders were selected. Captive founders were sourced from Mount Rothwell Biodiversity Interpretation Centre (Mt Rothwell), situated 60 km south-west of Melbourne. Wild founders were derived from free-ranging populations across 14 geographic regions in Tasmania, separated by at least 15 km or a significant geographical barrier to eastern quoll dispersal. To minimise impacts on the source population and maximise genetic diversity in the reintroduced population, no more than two animals in each cohort originated from any one site.

We selected founders that were in fair to excellent body condition (using a subjective assessment of fat and muscle stored between the hips and spine), weighed more than 640g, and were estimated to be 1-2 years old (inferring from tooth condition and wear). They were translocated to the ACT by air and road, where they were anaesthetised and assessed for health and disease. Founders were microchipped (each animal was identified using a four-character microchip code, see S1 Table) and fitted with VHF collars (32g, V6C 163 Zilco, Sirtrack Ltd, Hawkes Bay, New Zealand) or GPS collars (38g, LiteTrack 30 RF, Sirtrack Ltd, Hawkes Bay, New Zealand). Scat, fur, blood and ear (for DNA extraction) samples were collected.

Founders were monitored using VHF collars in Trials 1 and 2 and VHF-enabled GPS collars in Trial 3. Survival and den location were monitored daily for 42 days post-release (the ‘establishment period’) because survival plateaued after this period in Trial 1. We removed collars from males after this period and from females after their young had dispersed. We located collars immediately if a mortality signal was detected and conducted necropsies on all deceased animals that could be located.

We conducted post-release health checks every two weeks, though timing and frequency varied due to the reproductive stage of females, weight fluctuations (influencing collar fit), logistical constraints, and ability to re-trap the targeted animal. We conducted all trapping with wire cage traps (31 cm x 31 cm x 70 cm) that had padded doors, plastic lining (to collect scats), and were covered with a hessian sack. We checked traps before first light to minimise stress and allow animals to find shelter before daylight. Health checks included recording body mass, body condition, head and pes length, pouch occupancy, crown rump length of pouch young (CRL), and collection of fur and scat samples. We conducted health checks without sedation but with procedures to minimise handling time (generally