Monitoring Threatened Species and Ecological Communities

31 – The national malleefowl monitoring effort^393

density and malleefowl viability, motion-triggered cameras will be deployed at
control and treatment sites to determine whether baiting affects predator activity
(foxes, cats and dogs) (van Hespen 2015). These cameras will measure the response
of other species to predator baiting, such as kangaroos, rabbits and goats, which
will also provide information about the broader effect of predator control on the
ecosystem (e.g. competition with native or invasive herbivores).

Power analysis

To ensure the malleefowl AM experiment is well designed and to better understand
how long it might take to learn about the effectiveness of predator baiting, a power
analysis was conducted on the 41 candidate sites (Lahoz-Monfort and Hauser
2014). This involved simulating mound activity data that was expected to be
collected given historic records and an expected effect of baiting at treatment sites
(e.g. a 50% increase in the mean number of active mounds). The power analysis
showed that, with the proposed 41 sites, an effect of baiting will very likely be
detected after only 1 year of monitoring if the benefits offered by baiting are large
(specifically, a 65% increase in active mounds at baited sites) (Fig. 31.3). As the AM
sites are monitored for more years, the statistical power to detect an effect of
baiting increases. For example, after 4 years of running the experiment, a smaller
beneficial effect of baiting (specifically, a 25% increase in mound activity with a
probability of around 80%) can be expected to be detected. It should be noted,
however, that this analysis assumes all sites are established at the same time, and
there is no time lag between baiting and the effect of baiting on mound activity.
Delays in the establishment of experimental sites, and delays in the impacts of
baiting will increase the length of the experiment required.

ϭ͘Ϭ

Ϭ͘ε

Ϭ͘Θ

Ϭ͘ϳ

Ϭ͘ς

Ϭ͘ρ

Ϭ͘κ

Ϭ͘ϯ

Ϭ͘Ϯ

Ϭ͘ϭ

Ϭ͘Ϭ

Ϭй ϭϭй ϮϮй

ĞīĞĐƚƐŝnjĞ;йŝŶĐƌĞĂƐĞĐŽŵƉĂƌĞĚƚŽŶŽŶͲďĂŝƚĞĚͿ

ϯρй κεй ςρй ΘϮй

ƉŽǁĞƌ

dсϭLJƌ

dсϮLJƌ

dсϯLJƌ

dсκLJƌ

dсρLJƌ

Fig. 31.3. Statistical power (y-axis) to detect a percentage increase in the number of active mounds in
treatment sites compared with control sites (x-axis). Coloured lines represent years of monitoring, ranging
from 1 (blue) to 5 (black). The horizontal blue line indicates 80% power.