Dancing with the devil: Could native predators inside 'predator free' havens be good for the conservation of threatened native prey species?

As invasive predators continue to drive global biodiversity loss,
predator-free havens are rapidly being established to recover threatened
fauna. An unintended consequence of these efforts is that havened
populations can lose anti-predator traits, making them poorly suited for
reintroductions into landscapes that contain predators. One approach
showing promise in preventing the loss of these critical responses from
havened populations is in situ predator exposure. Here we evaluate this
via experimental translocations to determine whether predation pressure
from a native mesopredator (chuditch; Dasyurus geoffroii) is effective in
retaining behavioural and morphological anti-predator traits in the
Endangered woylie (Bettongia penicillata ogilbyi). We conducted replicated
reintroductions of semi-havened (free from invasive predators but exposed
to chuditch) and non-havened (control) woylies into vacant bushland
containing chuditch, feral foxes (Vulpes vulpes) and feral cats (Felis
catus). We monitored woylie survival, reproduction, physiology (faecal
glucocorticoid metabolites, fGCM), and proxies for behavioural and
morphological anti-predator traits before and for ten months after their
release. We found no effects of source population on weight, pes length,
agitation behaviour, or fGCM. Survival probability was higher for males,
and for individuals with higher fGCM levels, but we could not attribute
differences in woylie survival to whether they originated from a haven.
This suggests that sustained predation pressure from chuditch inside the
haven has been effective in maintaining anti-predator responses and
supporting survival of woylies faced with novel invasive predators.
However, it will be crucial to understand thresholds of in situ predation
pressure across contexts that havened fauna can withstand before
recommending broad scale adoption of this strategy. Authors should follow
a formula in which point 1 sets the context and need for the work; point 2
indicates the approach and methods used; the next 2-3 points outline the
main results; and the last point identifies the wider implications and
relevance to management or policy.