218 Monitoring Threatened Species and Ecological Communities
recovery/monitoring.html). Unfortunately, no monitoring design is optimal across
all objectives, and trade-offs must be made in design if more than one objective is
to be met. For example, it may be easier to engage volunteers in cockatoo counts if
they are offered count locations with high abundance. However, sending all
volunteers to places of high abundance would bias population size estimates, so a
trade-off between statistical and social benefits must be made. Similarly, money
spent on estimating the size of the population, or the range of the species, is money
that could have been spent on understanding its feeding preferences or estimating
its breeding success. Every monitoring program design involves numerous such
trade-offs. This is why the most pressing knowledge needs must be clearly
identified and prioritised to ensure the most appropriate type of monitoring
framework is used (Fig. 16.3).
What to monitor – threats or species?
For monitoring to deliver useful information about the effectiveness of management
in conserving species, it is commonly advised that monitoring should focus on
something that is close to the fundamental objective of avoiding extinction, such as
population size, population growth rate, survival rates, fecundity or range size
(Wintle and Lindenmayer 2008). Although tracking proximal variables may reveal
trends or patterns necessary to trigger emergency actions or highlight the need for
closer investigation, learning about causes of population f luctuations or the
effectiveness of management usually requires information about variation in agents
of change such as threats and environmental covariates. Unless there is complete
knowledge about the relationship between a species’ abundance or vital rates and
threats and stressors, it is almost always necessary to monitor and learn about the
relationship between the two in order to support good management decisions. For
example, if a manager is removing foxes to reduce the extinction risk of bandicoots,
it may not be sufficient to simply monitor the change in fox abundance to assess the
success of the program because there may be a commensurate rise in cat abundance
that in turn has a significantly negative inf luence on bandicoots – the true species of
interest (Ritchie and Johnson 2009). To properly evaluate the effectiveness of fox
culling, it will probably be necessary to monitor foxes and bandicoots, as well as
other likely covariates such as cats. After a period of monitoring and learning, the
state of knowledge may evolve to the point where it is possible to reduce or remove
the need for monitoring bandicoots because there exists a robust model describing
the relationship between fox, cat and bandicoot abundance. In short, what should be
monitored will be predicated by the state of knowledge at the time – if the
relationship between a threatened species and its threats and stressors is very well
known for a particular region, it may be sufficient to simply monitor the threats,
especially if that is cheaper and easier than monitoring the threatened species itself.