individuals: e.g. parrots are preferentially captured as nestlings for the pet trade
(Beissinger and Bucher 1992); eggs of the Maleo (Macrocephalon maleo) are
collected for food, while the adults are left alone (Argeloo 1994). The effect of
harvest of a particular class of birds on the population growth rate depends on the
reproductive value of that class (Kokko et al. 2001). Thus, where life-history
traits are strongly class-dependent (as in long-lived birds), or where exploitation
targets a particular class of birds, a structured population model (Caswell 2001)
is needed to assess the impact of exploitation. Field studies designed to estimate
age-, size-, sex-, or stage-specific survival and reproductive rates are required to
provide the parameters for this type of model (see Chapters 3 and 5).
13.7 Addressing uncertainty
13.7.1 Motivation
The advantage of an approach like the PBR formula is that it is simple to apply
and requires minimal information about the species being exploited. The disad-
vantage is that it produces a very conservative estimate of sustainable harvest, in
order to guard against the large amount of uncertainty about the life-history
dynamics and effects of harvest. Where there is strong motivation for increased
exploitation, managers and biologists can make more precise, less guarded,
assessments of harvest potential by increasing their knowledge about the dynamics
of the species in question.
Two critical uncertainties that need to be resolved to improve management of
any exploited species are the effects of density-dependence and exploitation
on life-history parameters. There are two approaches to resolving these uncer-
tainties in the context of management: management experiments and adaptive
management.
13.7.2 Management experiments
Management experiments involve applying different management treatments to
multiple experimental units (e.g. separate populations) in a randomized and
controlled statistical design (Walters and Holling 1990). The advantage of this
approach is that results can be obtained fairly quickly, and the inference that can
be made from the results is strong. That is, because of the random application
of treatments to experimental units, conclusions about cause and effect can be
made. Further, the methods for designing such studies and the techniques for
analysis are standard. The disadvantage is that this approach is sometimes viewed
as risky from a conservation standpoint: the range of treatments may need
to include options that have not been previously tried, or which are greater in
322 |Exploitation