Effect of novel predator on survival
When we assessed the impact of quolls on melomys survival between trapping sessions the best supported model was one in which survival rates between sessions were lower at quoll-invaded sites than at quoll-free sites, and recapture rates were session-dependent (see Table S3). All other models were more than 4 AIC units from this best model, and so clearly inferior descriptions of the data. From the best-supported model, estimates of apparent survival (S ) for the intervals between the capture sessions were substantially higher at quoll-free sites (S2017–2018 = 0.368;S2018–2019 = 0.225) than at quoll-invaded sites (S2017–2018 = 0.207;S2018–2019 = 0.091; Fig. 2). The differing survival probability between sessions is largely explained by the time difference between intervals (2017–2018 = 9 months vs. 2018–2019 = 13 months; Fig. 3).