392 Monitoring Threatened Species and Ecological Communities
outcomes of past actions, and (2) actively, where management is designed to
deliberately target and accelerate learning (McCarthy and Possingham 2007). The
malleefowl AM experiment was initiated in 2013 in collaboration between the
NMRT, the Victorian Malleefowl Recovery Group, Parks Victoria, and the
University of Melbourne, and aims to learn both passively and activity which
threatening process is primarily affecting population trends.
Malleefowl are a suitable candidate for AM because of the uncertain benefits
of potential management actions (Benshemesh and Bode 2011). In particular, the
benefit of the most common, and probably the most expensive, management
action – baiting for predators, especially foxes – is not clear and was the major
source of uncertainty identified at a workshop where views were elicited from
experts (Bode et al. 2017). Field experiments have shown that intense fox baiting
can reduce the mortality of young, captive-reared malleefowl, supporting its use
(Priddel and Wheeler 1997). However, more recent statistical analyses of
monitoring datasets have failed to demonstrate any long-term benefit to
malleefowl populations, despite significant investment in fox control (Benshemesh
et al. 2007; Walsh et al. 2012). Thus, the malleefowl AM experiment currently
focuses on the effectiveness of predator control; however, once this is better
understood, the experiment may be modified to actively learn about the benefits
of managing other threats, such as alternative fire regimes or habitat degradation
by herbivores.
Experimental design
Approximately 41 new monitoring sites (each ~400 ha) are currently being
established for the malleefowl AM experiment (Fig. 31.2). AM sites are grouped in
clusters, containing at least one treatment site and one control site, spanning the
range of the species’ distribution. The experiment is designed so that managers bait
for foxes and/or cats in and around treatment sites, while purposely not baiting in
and around control sites. Importantly, control and treatment sites will be close to
one another so that they are similar in terms of climate and habitat, but sufficiently
far apart to be independent of one another in terms of predator and malleefowl
activity. This ‘paired’ design will allow for any differences in malleefowl activity to
be attributed to predator control alone and not to other spatial or temporal factors
that are known to inf luence breeding activity, such as vegetation or rainfall. The
malleefowl AM experiment can be considered active AM because sites are
arranged purposely to accelerate learning and maximise power to detect the effect
of baiting on breeding activity.
The number of active mounds will be monitored annually in control and
treatment sites by land management agencies and volunteer citizen scientists: they
will become part of the larger mound monitoring program. Mound monitoring
will be used to identify whether baiting improves malleefowl population viability.
To provide information on the casual link between predator control, predator