Top predator restricts the niche breadth of prey: effects of assisted colonization of Tasmanian devils on a widespread omnivorous prey

Abstract

Few landscape-scale experiments test the effects of predators on the abundance and distribution of prey across habitat gradients. We use the assisted colonization of a top predator, the Tasmanian devil ( Sarcophilus harrisii ), to test the impacts of predation on the abundance, habitat use and temporal activity of a widespread prey species, the omnivorous common brushtail possum ( Trichosurus vulpecula ). Before introduction of devils to Maria Island, Tasmania, Australia, in 2012, possums were abundant in open grasslands as well as forests. Predation by devils caused high mortality of possums in grasslands, but individuals with access to trees had a higher survival probability. Possum abundance declined across the whole island from 2012–2016, as possums disappeared almost completely from grasslands and declined in drier forests with more open understorey. Abundance remained stable in wet forests, which are not preferred habitat for possums but provide better refuge from devils. Abundance and habitat use of possums remained unchanged at a control site on the adjacent Tasmanian mainland, where the devil population was low and stable. This study demonstrates how spatial variation in predator-caused mortality can limit both abundance and habitat breadth in generalist prey species, excluding them entirely from certain habitats.

Coat colour in marsupials: genetic variants at the ASIP locus determine grey and black fur of the brushtail possum

“…Conversely, possum is a prey species of the non-arboreal devil [ 78 , 79 ], indicating a further potential driver of black fur may be improved concealment from predators. In support of this as a significant selective force, introduction of the devil to Maria Island in 2012 to establish an insurance population, led to high possum mortality [ 80 , 81 ].…”

Fear of the human ‘super predator’ in native marsupials and introduced deer in Australia

“…Because all five playback treatments were broadcast at each site (see below), each site represents a stand-alone replicate of the experiment [19,32], and the experiment was thus replicated 13 times (the maximum number of sites that fit on the property). The >1 km spacing between sites was selected to ensure that different individuals were sampled at the different sites, which was almost certain concerning wallabies, pademelons and possums, and also highly probable regarding kangaroos and deer because the home range sizes of the smaller species are all <10 ha [47,48], and based on telemetry studies in south-east Australia and New Zealand those of the kangaroos and fallow deer are likely <100 ha [49][50][51].…”

Broad‐scale acoustic monitoring of koala populations suggests metapopulation stability, but varying bellow rate, in the face of major disturbances and climate extremes

“…We used stacked N‐mixture models to assess associations between koala bellow rate (number of 10 min periods with a bellow night −1 site −1 to improve data independence following the approach for camera traps by Scoleri et al., 2023 ) and covariates. Using such an approach, data collected in subsequent years are stacked so that unique site‐year combinations are treated as separate sites.…”

“…We used stacked N-mixture models to assess associations between koala bellow rate (number of 10 min periods with a bellow night −1 site −1 to improve data independence following the approach for camera traps by Scoleri et al, 2023) and covariates.…”