Figure 1. Map showing the arrangement of grassland melomys (Melomys burtoni) monitoring sites on Indian Island, Northern Territory, Australia. Quolls were present at the four monitoring sites in the north of the island and quolls were absent from the three monitoring sites in the south of the island for the duration of the study.
Melomys were monitored at seven independent 1ha (100 m x 100 m) plots (sites 1–7) spread out across Indian Island using a standard mark-recapture trapping regime designed for a monitoring project (Begget al. 1983; Kemper et al. 1987). Sites in the north (quoll-invaded) and south (quoll-free) of the island were between 8.7 and 9.8km apart (Fig. 1; Table S1) and were composed of similar habitat types. The northern and southern sections of Indian Island are divided by mangrove habitat which is inundated at high tide. Cage and camera trapping as well as track surveys confirmed that quolls were present at the “impact” sites and absent from the “control” sites for the duration of the study (Jolly et al. unpub. data).
Each of the seven monitoring sites consisted of 100 Elliott traps (Elliott Scientific Equipment, Upwey, Victoria) spaced at 10 m intervals in a 10 x 10 grid. Most trapping grids were open for four nights, however, the first trapping grid (site 1, May 2017) was open for six nights. After four trap nights, the majority of the melomys population had been captured at least once (Jolly et al. 2019). Traps were baited with balls of peanut butter, rolled oats and honey. These baits were replaced daily for the duration of each trapping session. Traps were checked for captures early each morning and all traps were cleared within two hours of sunrise.
Captured melomys were weighed (g) and sexed. Before release, each melomys was implanted with a microchip (Trovan Unique ID100). On successive mornings, all melomys were scanned (Trovan LID575 Handheld Reader), and any new individuals were microchipped. On the last morning of each trapping session, all melomys caught were retained for behavioural assays. Throughout the study 439 individual melomys were captured and given microchips (melomys caught per site: site 1 = 83; site 2 = 52; site 3 = 63; site 4 = 59; site 5 = 69; site 6 = 59; and site 7 = 54). Of these, 146 (33%) were caught on the final night of trapping and were retained for behavioural trials. Only large, healthy juveniles (n = 11), adult males (n = 58), and adult non-visibly pregnant females (n = 77) were retained for behavioural experiments. Melomys were retained in their respective Elliott traps and taken to the field station for diurnal husbandry. They were provided food and water ad libitum until 2 hours prior to testing. At this point, in an attempt to standardise hunger levels, access to food and water was removed. Indian Island is remote and uninhabited by humans, so all behavioural experiments were conducted in the field under near natural conditions (see Jolly et al. 2019 for detailed experimental procedures).