Environmental Engineering Reference
In-Depth Information
the key things that will help you to get meaningful results from a dynamic CPUE
model are
●
a long series of data;
●
more variation in effort;
●
monitoring while the system is out of equilibrium (ideally tracking recovery
followed by renewed depletion or
vice versa
);
●
monitoring a period of overexploitation (although from a conservation per-
spective this is obviously not desirable).
In addition, the following will also predispose the analysis to producing useful results:
●
less natural stochasticity in the population;
●
more accurately measured catch and effort;
●
prior information on parameter values (for example, an independent estimate
of carrying capacity, occasional estimates of actual population size, or an
assumption that the initial population was at carrying capacity if harvest was
observed from its outset).
Box 4.2
Fitting a dynamic model to catch and effort data.
In the simplest case of catch proportional to effort, we can predict catch at time
t
on the basis of effort and population size:
C
t
qE
t
N
t
where
q
is the 'catchability' coefficient, defining the proportion of the population
that can be caught per unit effort. We can model population size from one point
in time to the next by adding the net growth and subtracting the catch:
N
t
1
N
t
G
t
C
t
where growth,
G
t
, can be given by any appropriate population growth model; the
logistic is frequently used as a default model in the absence of evidence for an
alternative structure. In this case, the equation becomes:
K
N
t
1
C
t
e
r
max
K
N
t
1
N
t
Other possible structures are discussed in Hilborn and Walters (1992). This
model can be fitted to a time series of catch and effort data in the same way as this
type of data were used over short periods to estimate abundance (Box 2.5), this
time estimating the parameters of the population model as well as abundance.
Hilborn and Mangel (1997) provide an excellent introduction to the art of model
fitting of this kind, while the program CEDA provides an accessible tool for
fitting catch-effort models (weblink in Section 4.4).
