182
1. In order to iteratively run both of the tools and
operate the data format conversion between
them, a first step will be to use an external
third-party tool. As suggested above, R is a
good candidate because of its extended datahandling abilities and its capacity to interact
with other software packages through socalled system calls.
2. It is nevertheless considered to further develop
a module within ISIS-Fish to control MwZ
directly from within simulation iterations,
hence getting quick cost assessments and zoning optimizations to enhance the dynamic
properties of the whole system. The existing
ability of ISIS to connect to R could in particular be used to run scripts controlling the
whole MwZ loop, from data preparation to
processing of output to provide management
zones that can be handled by ISIS. This way,
costs and feature amounts could be updated
from ISIS-Fish to MwZ, and as a feedback,
zoning in ISIS-Fish could be updated according to MwZ outputs.
2.4
Types of Scenarios
to Be Tested
Coupling these two tools would open the opportunity to test a wide range of scenarios regarding
the dynamics of spatial conservation plans. The
following propositions are far from exhaustive
but focus on types of analyses already planned in
the context of this study.
A first and fairly obvious type of analysis
would consist of testing what the dynamics of
the main fleets and fish populations would be
under different conservation scenarios and how
they would influence target achievement (no
feedback, only the evaluation of species abundance proportions within protected areas after
the simulated period) and then testing for the
robustness of the MPA network through the stability in species representation. This would only
require a one-way coupling from MwZ to ISISFish that does not really require explicit and automated translation of a common management
zone output to management scenarios within
ISIS-Fish. Achievement of targets would be easily assessed from ISIS-Fish outputs in terms of
species abundance distributions. More realistic
economic costs, although completely independent at this stage from those used in MwZ, could
also be assessed by comparison of simulated
landings at the beginning and the end of the simulation period.
From the previous analysis, it may be possible to test whether any proposed MPA network
with a particular set of conservation measures is
suitable to ensure medium- to long-term viability of fleets and of the fish populations they harvest. At present, in the Eastern English Channel,
enforcement measures are still to be defined in
most proposed MPAs, and it may be very relevant to use MwZ on its own to provide near
optimal management scenarios within the
already planned MPA network. The addition of
the ISIS-Fish simulations would enable the
evaluation of the management strategy under
which the fishing fleets will remain viable. For
that purpose, the methodology proposed by
Lehuta et al. (2013), based on bioeconomic
indicators and their uncertainty to evaluate management strategies, could be used.
Other types of analyses would require a more
intricate and fully dynamic coupling of the two
tools than the preceding ones. It is, for instance,
planned to test the effect of the chronological
sequence of enforcement and how it could be
optimized. Indeed, it would involve running
ISIS-Fish over a given interval of time (e.g., one
year), then testing which would be the best
enforcement addition to the network using MwZ
and running ISIS-Fish again from where it
stopped, etc. Such an approach would have to be
tested over various time lags.
All these types of analysis would benefit by
also testing whether different near-optimal solutions, which differ noticeably in terms of selected
sets of PU-zone pairs but not in terms of cost,
would lead to different dynamics and viabilities
of fleets and harvested populations. Such analyses would have to account for confidence in data.
From a more general point of view, a sensitivity
analysis on optimization and model parameters
would be necessary.
Y. Reecht et al.
1. In order to iteratively run both of the tools and
operate the data format conversion between
them, a first step will be to use an external
third-party tool. As suggested above, R is a
good candidate because of its extended datahandling abilities and its capacity to interact
with other software packages through socalled system calls.
2. It is nevertheless considered to further develop
a module within ISIS-Fish to control MwZ
directly from within simulation iterations,
hence getting quick cost assessments and zoning optimizations to enhance the dynamic
properties of the whole system. The existing
ability of ISIS to connect to R could in particular be used to run scripts controlling the
whole MwZ loop, from data preparation to
processing of output to provide management
zones that can be handled by ISIS. This way,
costs and feature amounts could be updated
from ISIS-Fish to MwZ, and as a feedback,
zoning in ISIS-Fish could be updated according to MwZ outputs.
2.4
Types of Scenarios
to Be Tested
Coupling these two tools would open the opportunity to test a wide range of scenarios regarding
the dynamics of spatial conservation plans. The
following propositions are far from exhaustive
but focus on types of analyses already planned in
the context of this study.
A first and fairly obvious type of analysis
would consist of testing what the dynamics of
the main fleets and fish populations would be
under different conservation scenarios and how
they would influence target achievement (no
feedback, only the evaluation of species abundance proportions within protected areas after
the simulated period) and then testing for the
robustness of the MPA network through the stability in species representation. This would only
require a one-way coupling from MwZ to ISISFish that does not really require explicit and automated translation of a common management
zone output to management scenarios within
ISIS-Fish. Achievement of targets would be easily assessed from ISIS-Fish outputs in terms of
species abundance distributions. More realistic
economic costs, although completely independent at this stage from those used in MwZ, could
also be assessed by comparison of simulated
landings at the beginning and the end of the simulation period.
From the previous analysis, it may be possible to test whether any proposed MPA network
with a particular set of conservation measures is
suitable to ensure medium- to long-term viability of fleets and of the fish populations they harvest. At present, in the Eastern English Channel,
enforcement measures are still to be defined in
most proposed MPAs, and it may be very relevant to use MwZ on its own to provide near
optimal management scenarios within the
already planned MPA network. The addition of
the ISIS-Fish simulations would enable the
evaluation of the management strategy under
which the fishing fleets will remain viable. For
that purpose, the methodology proposed by
Lehuta et al. (2013), based on bioeconomic
indicators and their uncertainty to evaluate management strategies, could be used.
Other types of analyses would require a more
intricate and fully dynamic coupling of the two
tools than the preceding ones. It is, for instance,
planned to test the effect of the chronological
sequence of enforcement and how it could be
optimized. Indeed, it would involve running
ISIS-Fish over a given interval of time (e.g., one
year), then testing which would be the best
enforcement addition to the network using MwZ
and running ISIS-Fish again from where it
stopped, etc. Such an approach would have to be
tested over various time lags.
All these types of analysis would benefit by
also testing whether different near-optimal solutions, which differ noticeably in terms of selected
sets of PU-zone pairs but not in terms of cost,
would lead to different dynamics and viabilities
of fleets and harvested populations. Such analyses would have to account for confidence in data.
From a more general point of view, a sensitivity
analysis on optimization and model parameters
would be necessary.
Y. Reecht et al.