19.6.4 Globally Applicable Area-Disaggregated General
Ecosystem Toolbox (Gadget) Model
Gadget is a model considered to show the greatest potential to contribute to practical
fishery management advice such as changes to total allowable catch (Plagányi 2007).
It contains a flexible and powerful software tool that has been developed to model
marine ecosystems, including both the impact of the interactions between species
and the impact of fisheries on the populations. Gadget works by running an internal
forward projection model based on many parameters describing the ecosystem and
then comparing the output from this model to observed measurements to get a
likelihood score (Begley 2005; Taylor et al. 2007).
In this model the processes are usually length dependent. The framework allows
the creation of multi-area or multi-fleet models in which it is possible to include
predation and mixed fisheries. Gadget has essentially three components: (a) an
ecosystem simulator; (b) a likelihood function that takes the output (from the
ecosystem simulator) and compares the data; and (c) a function minimizer (optimization routines to find the best set of the model parameters values). To work, Gadget
needs initial estimates of recruitment (age 0) every year and initial abundances by
age in the first year.
19.6.5 Gadget Input Data and Parameters
Particularly, to establish and project forward the population dynamic of the
Blackspot Seabream in the Strait of Gibraltar, Gadget follows this order: fish are
caught by the “voracera” fleet with five different selection patterns (one for Morocco
and four for Spain); afterwards it dies by natural mortality and eventually growths
and ages.
As is stated above, model parameters are estimated minimizing differences
among observations and estimated values provided for the program. Gadget’s
probability processes the result of the ecosystem simulation based on aggregate
dimensions: thus within this module, several datasets can be compared to the model
output using a suite of different types of length distribution. Particularly, in the case
of the Blackspot Seabream fishery, Gadget model takes into account four different
types of data to enter the probability function: length distribution from commercial
fleets (Morocco and Spain), age-length distribution and sex ratio at length (from
biological samplings), and finally fleets’ effort (in fishing days). Thus the likelihood
or probability function included a total of 20 different components.
For a simulation with Gadget, it is necessary to assume that the Blackspot
Seabream is a long-live species, so the maximum age is set at 17 (for males and
females). While the model length range was from 0 to 62 cm, in 1 cm length
intervals, with female population start at 20 cm. The period of simulation considered
starts in 1983 and ends in 2016. On the other hand, in Spain four catch categories
642
J. Gil-Herrera et al.
Ecosystem Toolbox (Gadget) Model
Gadget is a model considered to show the greatest potential to contribute to practical
fishery management advice such as changes to total allowable catch (Plagányi 2007).
It contains a flexible and powerful software tool that has been developed to model
marine ecosystems, including both the impact of the interactions between species
and the impact of fisheries on the populations. Gadget works by running an internal
forward projection model based on many parameters describing the ecosystem and
then comparing the output from this model to observed measurements to get a
likelihood score (Begley 2005; Taylor et al. 2007).
In this model the processes are usually length dependent. The framework allows
the creation of multi-area or multi-fleet models in which it is possible to include
predation and mixed fisheries. Gadget has essentially three components: (a) an
ecosystem simulator; (b) a likelihood function that takes the output (from the
ecosystem simulator) and compares the data; and (c) a function minimizer (optimization routines to find the best set of the model parameters values). To work, Gadget
needs initial estimates of recruitment (age 0) every year and initial abundances by
age in the first year.
19.6.5 Gadget Input Data and Parameters
Particularly, to establish and project forward the population dynamic of the
Blackspot Seabream in the Strait of Gibraltar, Gadget follows this order: fish are
caught by the “voracera” fleet with five different selection patterns (one for Morocco
and four for Spain); afterwards it dies by natural mortality and eventually growths
and ages.
As is stated above, model parameters are estimated minimizing differences
among observations and estimated values provided for the program. Gadget’s
probability processes the result of the ecosystem simulation based on aggregate
dimensions: thus within this module, several datasets can be compared to the model
output using a suite of different types of length distribution. Particularly, in the case
of the Blackspot Seabream fishery, Gadget model takes into account four different
types of data to enter the probability function: length distribution from commercial
fleets (Morocco and Spain), age-length distribution and sex ratio at length (from
biological samplings), and finally fleets’ effort (in fishing days). Thus the likelihood
or probability function included a total of 20 different components.
For a simulation with Gadget, it is necessary to assume that the Blackspot
Seabream is a long-live species, so the maximum age is set at 17 (for males and
females). While the model length range was from 0 to 62 cm, in 1 cm length
intervals, with female population start at 20 cm. The period of simulation considered
starts in 1983 and ends in 2016. On the other hand, in Spain four catch categories
642
J. Gil-Herrera et al.