Environmental Engineering Reference
In-Depth Information
because the determinant of harvesting effort is
profitability
. This is made up both
of revenues obtained from hunting and costs incurred, which are related both to
the size of the harvested population (which we are attempting to estimate) and also
to the economic system in which hunting is embedded. Instead, effort is used as an
index to standardise catch rates, giving the index
Catch per Unit Effort
(CPUE).
The more abundant a resource is, all else being equal, the easier it is to harvest.
Following from this, the greater the CPUE is, the more abundant the population
is likely to be. Thus, in principle, CPUE can act as an
index of population abun-
dance
, and could be monitored to detect declines in the same way as abundance
itself. This is attractive because catch and the effort put into harvest are relatively
easy to measure, compared to biological parameters such as population size,
productivity or carrying capacity.
This approach relies on the very strong assumption that CPUE is strictly
pro-
portional to abundance
. There are several reasons why this might not be the case,
which are detailed in Section 2.3.5.1. Unfortunately, given catch and effort data
alone, there is no way to test the assumption of proportionality, and using these
data as a monitoring tool therefore needs to be treated with extreme caution.
As we saw in Figure 1.1, the theory of harvesting predicts that there is a domed
relationship between equilibrium catch and effort. Thus, in theory, if we have a
series of catch and effort data for widely varying levels of effort, plotting the
catch-effort curve may enable us to define a domed response that can be used to
define whether a population is overexploited. If current effort and catch are on the
right-hand side of the peak, the population is overexploited. This is extremely mis-
leading, however, because it assumes
equilibrium
, whereas data will always be from
dynamic systems, with varying levels of effort and populations lagging behind in
their responses to changing harvest. If catch and effort from a time series of data
from a single location are simply plotted against one another in order to define a
maximum, MSY will almost always be overestimated. Using data from several spa-
tially separated populations that have been harvested at contrasting rates could
allow this method to work, in principle, but only if each population is close to its
equilibrium state, having been harvested at more or less constant rates for a con-
siderable period of time. In practice, such data are very hard to find.
The solution to this problem is to use a
dynamic model
to estimate the crucial
parameters. For this, rather than assuming that the population is at equilibrium,
one uses the catch and effort time series to model the underlying changes in popu-
lation size by fitting a dynamic population model to the data (Box 4.2). While this
approach is potentially powerful, it is very constrained by the quantity and quality
of available data. At least four parameters need to be estimated (intrinsic rate of
increase, carrying capacity, catchability and initial population size), and a reason-
ably long CPUE time series is needed to resolve all of these parameters. Not only
that, but the data also needs to contain good
contrast
. This means there should be
a lot of variation in the underlying population size over time, and in the amount of
effort applied. If you monitor a system where everything is at a fairly steady state,
the data will contain no useful information, no matter how long the time series. So
