Environmental Modelling & Software 89 (2017) 120e130
Contents lists available at ScienceDirect
Environmental Modelling & Software
journal homepage: www.elsevier.com/locate/envsoft
Standardized ecological indicators to assess aquatic food webs: The
ECOIND software plug-in for Ecopath with Ecosim models
M. Coll a, b, c, *, J. Steenbeek a, b, c
a
Institut de Recherche pour le D
eveloppement, UMR MARBEC & LMI ICEMASA, University of Cape Town, Private Bag X3, Rondebosch, Cape Town 7701,
South Africa
b
Ecopath International Initiative Research Association, Barcelona, Spain
c
Institut de Ci
encies del Mar (ICM-CSIC), Barcelona, Spain
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 30 March 2016
Received in revised form
6 October 2016
Accepted 19 December 2016
Ecological indicators are useful tools to analyse and communicate historical changes in ecosystems and
plausible future scenarios while evaluating environmental status. Here we introduce a new plug-in to the
Ecopath with Ecosim (EwE) food web modelling approach, which is widely used to quantitatively
describe aquatic ecosystems. The plug-in (ECOIND) calculates standardized ecological indicators. We
describe the primary functionality of ECOIND and provide an example of its application in both static and
temporal-spatial dynamic modelling, while we highlight several related features including a new taxonomy input database (species traits) and the ability to analyse input uncertainty on output results.
ECOIND adds new capabilities to the widely used EwE food web modelling approach and enables
broadening its applications into biodiversity and conservation-based frameworks to contribute to integrated ecosystem analyses.
© 2016 Elsevier Ltd. All rights reserved.
Keywords:
Software plug-in
Ecological standardized indicators
Environmental status
Food web models
Ecopath with Ecosim
Ecospace
Software and/or data availability
1. Introduction
Name of software Ecopath with Ecosim
Developer Ecopath Research and Development Consortium
Contact address Ecopath International Initiative Research
Association, Barcelona, Spain
Contact phone þ34.685.319.597
Contact email [email protected]
Year first available 1991 (ECOIND plug has been released in
July 2016, version 6.5.14040.0)
Hardware required PC
Software required Windows XP service pack 3 or newer,
Microsoft.NET Framework 4 Full Profile
Availability Public, Open Source (GPLv2), freely available
from www.ecopath.org
Program language Visual Basic.NET, C# Program size: 16 MB
(basic installation) Website: http://www.
ecopath.org
Historical changes in aquatic resources are documented
worldwide (Lotze et al., 2006) due to changes in multiple anthropogenic and climate-related drivers (Halpern et al., 2015). These
drivers alter the structure and functioning of ecosystems (Jackson
et al., 2001) and can affect the ecosystem services that humans
obtain from healthy oceans (Worm et al., 2006). There is a concern
about the environmental status of aquatic ecosystems and a need to
adopt a more integrated view of ecosystem management. This view
should consider not only the dynamics of target species, but also
non-target organisms, trophic relationships and flows, and environmental factors (Pikitch et al., 2004).
To move forward, adaptations and changes to scientific methods
are required, in parallel with changes to the way ecological, social
and economic issues are integrated in management processes
(Browman et al., 2005a, 2005b). Within this context, the scientific
community has developed new methodological tools such as
robust ecological indicators to track the environmental status of
ecosystems and inform management decisions (Cury and
Christensen, 2004). In fact, there has been considerable discussion about ecological indicators to monitor the pressures on, and
status of, exploited aquatic ecosystems with emphasis on marine
ecosystems (Rombouts et al., 2013). Initially, indicators were
ncies del Mar
* Corresponding author. Corresponding author. Institut de Cie
(ICM-CSIC), Passeig Marítim de la Barceloneta, nº 37-49, 08003, Barcelona, Spain.
E-mail address: [email protected] (M. Coll).
http://dx.doi.org/10.1016/j.envsoft.2016.12.004
1364-8152/© 2016 Elsevier Ltd. All rights reserved.
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
developed to include ecological considerations with the goal of
capturing the impact of dominant pressures, such as fishing (Fulton
et al., 2005). Further progress has included the establishment of
criteria and frameworks to (i) guide the selection of indicators that
are used to assess the effects of fishing via trend and threshold
analyses, (ii) define preliminary reference levels and reference directions for selected indicators, and (iii) develop and test evaluation
frameworks (Blanchard et al., 2010; Kleisner et al., 2013; Large et al.,
2013). Recently, the scope of ecosystem indicators has expanded to
include impact on biodiversity, other socio-economic and governance issues, and the cumulative impacts of multiple human activities (Boldt et al., 2014; Large et al., 2015; Levin et al., 2009). Such
ranges of indicators are needed to fulfil the targets of different
National and Transnational management plans and conventions,
such as the European Marine Strategy Framework Directive (MSFD)
seeking to achieve a “Good Environmental Status” (GES) by 2020
for all European seas (EC, 2010) or the Aichi Targets of the
Convention of Biological Diversity (Tittensor et al., 2014).
Ecological models (Fulton, 2010) provide a variety of results that
can be used to calculate meaningful ecological indicators and
inform about the environmental status of ecosystems (Allesina and
Bondavalli, 2004; Valentini and Jord
an, 2010). These models can
take into account the dynamics of commercial and non-commercial
species, their interactions and the main drivers, and are thus useful
tools for marine conservation planning and sustainable use management (Piroddi et al., 2015; Smith et al., 2015). One of such tools is
the ecosystem modelling suite “Ecopath with Ecosim”, or EwE, that
integrates the original Ecopath food web model with the temporal
dynamic and temporal-spatial dynamic modules Ecosim and Ecospace, respectively (Christensen and Walters, 2004a). The EwE
approach was the first user friendly and freely accessible
ecosystem-level simulation model, which has contributed to its
global uptake and its popularity as a key tool for the ecosystemapproach. Currently, more than 400 EwE ecosystem models have
been published, mostly in aquatic ecosystems (Coll et al., 2015a)
This makes EwE an important approach to explore ecosystem
related questions and to link with other modelling techniques
(Cerco et al., 2010). In acknowledgement of its impact, the US National Oceanographic and Atmospheric Administration recognized
Ecopath as one of the ten biggest scientific breakthroughs in the
organization's 200-year history (http://celebrating200years.noaa.
gov/breakthroughs/ecopath/).
A variety of EwE results have been used extensively to calculate
ecological indicators (e.g., Coll and Libralato, 2012; Heymans et al.,
2014). To facilitate this use and enable the standardized calculation
of several of the most widely used ecological indicators, we
developed the new ECOIND plug-in to be released with EwE
version 6.5 (2016).
2. Methods
2.1. Overview of ECOIND
ECOIND is a compilation of algorithms to quantify ecological
indicators from EwE food web model results. These indicators are
related with the environmental status of ecosystems, mostly
associated with fishing impacts (such as the biomass of commercial
species or the mean trophic level of the catch), but also with
biodiversity impacts and conservation-based issues (such as the
biomass and catch of predators or species at risk) (Coll et al., 2016a;
Shannon et al., 2014) (Table 1). Several of these indicators can be
directly related with management attributes included in national
and international management schemes (i.e. MSFD-GES and CBDArchi Targets). The full description of all the ecological indicators
included in ECOIND and their use and interpretation is beyond the
121
scope of this paper, but interested readers can find relevant information in several key papers about ecological indicators (Cury and
Christensen, 2004; Fulton et al., 2005; Rochet and Trenkel, 2003;
Shin and Shannon, 2010). A table with definitions and reference
information about the indicators is provided in Table S1 (Annex).
ECOIND is freely distributed with EwE software version 6.5 and
can be used to calculate indicators from the snapshot Ecopath
model, the temporal-dynamic module Ecosim, and the temporalspatial module Ecospace. Ecopath is the mass-balance routine
that allows building spatially and time averaged models of the
trophic web, Ecosim is the time dynamic routine, and Ecospace
allows representing temporal and spatial 2D dynamics of trophic
web components. The ECOIND plug-in integrates with the EwE
model execution flow, and when enabled, automatically calculates
indicators by aggregating, integrating and interpreting the results
of different modules of EwE (Fig. 1 and Annex II).
2.2. Data requirements and work flow
The first step to use ECOIND is to develop a food web model
using the Ecopath module (Christensen and Walters, 2004a)
(Fig. 2). Ecopath describes the balance between production of
functional groups and all consumptions within an ecosystem. Each
functional group in an Ecopath model can represent a species, a
sub-group of a species (e.g. juveniles and adults, which are called
multi-stanza groups) or a group of species that have functional and
ecological similarities (Fig. 3). The Ecopath model requires, for each
functional group or species in the food web model, the definition of
biomass, consumption, and production, in addition to the elements
of the diet matrix in the food web, exports and fishing yields, the
fraction of unassimilated food and biomass accumulation terms. A
summary of the main equations and assumptions is provided in
Annex I.
A new feature in Ecopath that has been developed along with
ECOIND is the ability to provide information about the taxonomic
composition of each functional group in the food web model. Once
the taxonomic composition of each functional group is defined,
additional information per species as traits within each functional
group can be entered (Fig. 2 and Annex II).
A second step to using the plug-in is to calculate indicators from
the predictions of the temporal-dynamic module Ecosim (Fig. 2 and
Annex II). On the basis of the system of algebraic equations and
input parameters of Ecopath, a set of differential equations is
defined in Ecosim (Walters et al., 2000). The biomass change at
each model time step for each consumer functional groups can be
expressed as changing over time. This change is in relation to
changes in consumption and production of organisms in the food
web including internal and external drivers (such as fishing and
environmental changes). Ecosim enables consideration of different
mechanisms of flow control, ecological behaviour of organisms and
forcing elements like environmental factors that can become quite
complicated in the case of multiple uses of resources (multiple
predators on the same prey).
Ecosim is frequently used to fit models to historical data to
evaluate the capability of a model to represent past ecosystem
conditions (Walters et al., 2005). Once fitted, a model can be used to
address scenarios of plausible alternative management options. If
confidence intervals of initial input parameters to the Ecopath
model are available (e.g. using the Ecopath “Pedigree” facility or
directly entering available values), Ecosim can run multiple times
using a Monte Carlo simulation analysis (Fig. 2 and Annex II). The
Monte Carlo approach searches for, and computes the impact of,
alternative Ecopath input parameter combinations. The Monte
Carlo routine can be used to test for sensitivity of Ecopath input
parameters to time series outputs, and to improve the fit of the
122
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
Table 1
Ecological indicators included in ECOIND plug-in (a. present values of the Mediterranean food web model in 1978 and b. in 2010, and b/a is the ratio of values from 2010 in
relation to 1978). A detailed description of these indicators is provided in Table S1 Annex.
Indicator
A. Biomass-based
Total B
Commercial B
Fish B
Invertebrates B
Invertebrates/Fish B
Demersal B
Pelagic B
Demersal/Pelagic B
Predatory B
Kempton's Q
B. Catch-based
Total C
Fish C
Invertebrate C
Invertebrates/Fish C
Demersal C
Pelagic C
Demersal/pelagic C
Predatory C
Discards
C. Trophic-based
TL catch
MTI
TL co.
TL co. 2
TL co. 3.25
TL co. 4
D. Size-based
ML of fish co.
ML of fish C
MW of fish co.
MW of fish C
MLS of fish co.
MLS of fish C
E. Species-based
Intrinsic Vul. Index
Endemics B
Endemics C
IUCN species B
IUCN species C
Mammals, birds & reptiles B
Mammals, birds & reptiles C
Description
Units
a.1978
b.2010
b/a
Total biomass (B) (i)
Biomass (B) of commercial species (i)
Biomass (B) of fish species (a)
Biomass (B) of invertebrate species (þ)
Biomass (B) of invertebrates over fish (þ)
Biomass (B) of demersal species (þ)
Biomass (B) of pelagic species (þ)
Biomass (B) of demersal over pelagic species (a)
Biomass (B) of predatory organisms with trophic level 4 (i) (þ)
Kempton's biodiversity index (Q) (i) (þ)
t$km2
t$km2
t$km2
t$km2
117.53
16.44
11.22
5.22
0.47
5.22
13.03
0.40
0.67
4.62
78.85
13.11
6.95
6.16
0.89
6.16
8.10
0.76
0.11
4.02
0.67
0.80
0.62
1.18
1.90
1.18
0.62
1.90
0.17
0.87
Total Catch (C) (þ)
Catch (C) of all fish species (þ)
Catch (C) of all invertebrate species (þ)
Catch (C) of invertebrates over fish (þ)
Catch (C) of demersal species (þ)
Catch (C) of pelagic species (þ)
Catch (C) of demersal over pelagic species (þ)
Catch (C) of predatory organisms with trophic level 4 (i)
Total discarded catch (i)
t$km2$year1
t$km2$year1
t$km2$year1
3.97
3.75
0.23
0.06
0.23
3.95
0.06
0.11
0.28
2.52
1.92
0.59
0.31
0.59
2.21
0.27
0.16
0.27
0.63
0.51
2.63
5.12
2.63
0.56
4.71
1.42
0.96
3.07
3.78
3.12
3.91
1.02
1.03
1.44
2.42
3.69
4.12
1.39
2.40
3.91
4.11
0.97
0.99
1.06
1.00
cm
cm
kg
kg
year
year
16.74
12.09
170.99
89.11
4.47
3.34
17.66
15.85
164.19
158.30
4.61
4.52
1.06
1.31
0.96
1.78
1.03
1.35
t$km2
t$km2$year1
t$km2
t$km2$year1
t$km2
t$km2$year1
49.29
1.E04
2.E08
1.07
0.09
0.42
0.42
47.97
2.E04
6.E07
0.72
0.14
0.29
0.29
0.97
1.73
33.07
0.67
1.61
0.71
0.70
t$km2
t$km2
t$km2
t$km2$year1
t$km2$year1
t$km2$year1
t$km2$year1
Tropic level (TL) of the catch (i)
Marine trophic index, trophic level (TL) of the catch (including organisms
with TL 3.25) (i)
Trophic level (TL) of the community (including all organisms) (i)
Trophic level (TL) of the community (including organisms with TL 2) (i)
Trophic level (TL) of the community (including organisms with TL 3.25) (i)
Trophic level (TL) of the community (including organisms with TL 4) (i)
Mean
Mean
Mean
Mean
Mean
Mean
length (ML) of fish in the community (þ)
length (ML) of fish in the catch (C) (þ)
weight (MW) of the fish in the community (þ)
weight (MW) of fish in the catch (C) (þ)
life span (MLS) of fish in the community (þ)
life span (MLS) of fish the catch (C) (þ)
Intrinsic Vulnerability Index of the catch (þ)
Biomass (B) of endemic species in the community (þ)
Endemic species in the catch (C) (þ)
Biomass (B) of IUCN-endangered species in the community (þ)
IUCN-endangered species in the catch (C) (þ)
Biomass (B) of marine mammals, seabirds and reptiles (þ)
Catch (C) of marine mammals, seabirds and reptiles (þ)
i: initial information from the Ecopath model, þ: additional information needed to compute this indicator (see Table 2).
model to historical data (Christensen et al., 2008). In this case, as a
third step, ECOIND calculates multiple time series for each indicator
from Monte Carlo-generated alternate initial Ecopath model inputs
when running the Ecosim model a large number of times, enabling
further analyses of uncertainty in ecological indicators due to input
uncertainty both in Ecopath and in Ecosim indicators’ outputs
(Fig. 4A and B).
As a fourth step, ECOIND calculates indicators from spatialtemporal predictions of the Ecospace module (Fig. 2 and Annex
II). Ecospace applies the Ecosim equations to a two-dimensional
space over time taking species movement, habitat suitability and
fisheries dynamics into account (Christensen et al., 2014). In Ecospace, the spatial extent of the ecosystem is represented by a grid of
cells; where each can be assigned a ‘depth’ of 0 or higher. Zerodepth cells, historically referred to as ‘land’ cells, are considered
devoid of Ecosystem dynamics, while non-zero depth cells are
‘water’ cells with active ecosystem dynamics. If water, a fixed
habitat type or a habitat suitability value can be assigned to each
cell for each functional group using any combination of environmental parameters. Fishing fleets and other human activities can be
restricted to specific regions, and map areas can be permanently or
temporarily protected from human exploitation (no-take zones).
Moreover, ecological and behavioural parameters including the
average dispersal rates of organisms can be considered and spatialtemporal simulations can be made more realistic by including
relative variations of primary productivity, fishing costs (e.g. in case
distance from port is considered a delimiting cost factor), habitat or
environmental parameters through the spatial-temporal data
framework (Steenbeek et al., 2013). When running Ecospace,
ECOIND calculates all ecological indicators by unit of time and
space, and can be set to automatically save these indicators as series
of geo-referenced maps, which in turn can be visualized and analysed in dedicated spatial data processing software (Figs. 2 and 4C).
The ECOIND plug-in can be set to automatically save results from
Ecopath, Ecosim, Ecospace and Monte Carlo when these are
computed, or can save results from its user interface.
2.3. Calculation and visualization of indicators
Ecological indicators of ECOIND plug-in (Table 1 and Table S1
Annex) are divided in five groups:
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
123
Fig. 1. Integration of ECOIND plug-in in the Ecopath with Ecosim software structure.
Fig. 2. Work flow and main functionalities of ECOIND plug-in. Solid arounds indicate implemented links between each functionality of the plug-in, dotted arrows indicated potential links that can be developed in the future.
124
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
Fig. 3. A. Study area in the Mediterranean Sea, and B. Flow diagram of the study food web where functional groups represented by circles are organised by trophic level (y-axis) and
habitat (x-axis). The quantity of trophic flows between functional groups is highlighted from red (largest) to blue (lowest). The size of the circles is proportional to their biomass in
the ecosystem. This figure has been drawn using Ecopath flow diagram drawing tool. (For interpretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
i) Biomass-based indicators: these indicators are based on the
abundance of organisms in the food web. Biomass data of marine organisms in the ecosystem are basic information to assess
species population trends and status (Hilborn and Walters,
1992).
Indicators include total biomass (Total B) of species in the
ecosystem, biomass of commercial species (Commercial B),
biomass of fish (Fish B), of invertebrates (Invertebrates B) and the
ratio of the two latter (Invertebrates/Fish B), biomass of demersal
(Demersal B) and Pelagic (Pelagic B) organism and their ratio
(Demersal/Pelagic B) and the Kempton's biodiversity index (Q). The
Q index is proportional to the inverse slope of the speciesabundance curve and is a proxy of ecosystem biodiversity
(Ainsworth and Pitcher, 2006).
ii) Catch-based indicators: these indicators are based on the catch
and discard species in the food web. Catch data, or the removal
of organisms from aquatic ecosystems due to fishing activities,
are basic data needed to evaluate the state of fishery resources
and the fisheries sector to support scientific advice and in some
cases can give an idea of the abundance of organisms in the
ecosystem (Zeller and Pauly, 2007).
Catch-based indicators include total catch (Total C) and the
catch of fish (Fish C), invertebrates (Invertebrates C) and their ratio
(Invertebrates/Fish C), demersal (Demersal C) and Pelagic (Pelagic
C) catch of organisms and their ratio (Demersal/Pelagic C), catch of
predatory organisms (Predatory C), defined as organisms with
trophic level (TL) 4, and total discards (Discards).
iii) Trophic-level based indicators: six indicators are based on
the trophic level (TL) concept. The TL identifies the position
of organisms within food webs by identifying the source of
energy for each organism. TL was first defined as an integer
value (Lindeman, 1942), placing species or functional groups
into a simple scheme. The concept was later modified to be
fractional (Odum and Heald, 1975) to account for omnivory.
Following an established convention, fractional trophic
levels are calculated by assigning producers (and often also
detritus) to trophic level 1 (e.g. phytoplankton), and consumers to a trophic level of 1 plus the average trophic level of
their prey, weighted by their proportion in weight in the
predator's diet (Christensen, 1996). As fishing selectively
removes organisms from the food web, the trophic and size
structure of the ecosystem may be altered and thus trophic
level-based (TL-based) indicators can be used to capture this
effect (Shannon et al., 2014).
In ECOIND, the following indicators are included: Tropic level
(TL) of the catch (TLcatch) (Christensen, 1996; Pauly et al., 1998), the
Marine Trophic Index (MTI, or TLc including organisms with
TL 3.25) (Pauly and Watson, 2005), and the TL of the community
including all organisms (TL co), TLco including organisms with
TL 2 (TLco2), TLco including organisms with TL 3.25 (TLco3.25),
and TLco including organisms with TL 4 (TLco 4).
iv) Size-based indicators: these indicators are based on changes
in size or weight of organisms in the food web. Size-based
indicators are used to evaluate changes in marine ecosystems since fishing tend to capture the largest organisms first
(Rochet and Trenkel, 2003).
In ECOIND, the indicators included are Mean length (ML) of fish
in the community (ML of fish co) and in the catch (ML of fish C),
Mean weight (MW) of the fish in the community (MW of fish co)
and in the catch (MW of fish C), and Mean life span (MLS) of fish in
the community (MLS of fish co) and in the catch (MLS of fish C).
However, since EwE models are mainly biomass-based (with the
exception of multi-stanza modelling capabilities within the
biomass model, Walters et al., 2008), these indicators provide a
means to track the weighted average of length, weight and life span
of organisms in the community and in the catch assuming a constant length, weight and life span of organisms in the ecosystem (by
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
125
Fig. 4. Biomass-based ecological indicators resulting from the NW Mediterranean food web model. A. Example of the histogram of ecological indicators results for 1978 (300
Ecopath models were run). Y-axes represent the frequency of occurrence in the 300 model runs, x-axes represent the values of each indicator. B. Example of temporal results
showing time series of biomass-based indicators from 1978 to 2010 (results are based on 300 Ecosim model runs, shaded lines represent single results of each run, blue line
represents the mean of all results). C. Example of temporal-spatial biomass-based indicators in 2010 using the Ecospace model (colour scale represents relative values: higher, in
red, or lower, in blue, relative to the baseline Ecopath model). Additional results for the rest of the indicators are provided in Annex III. (For interpretation of the references to colour
in this figure legend, the reader is referred to the web version of this article.)
default). To vary the length, weight and life span of organisms over
time, extra computations will have to be implemented in the future
(this capacity still needs to be developed).
v) Species-based indicators: ECOIND calculates eight indicators
specifically based on species traits and conservation status.
The Intrinsic Vulnerability Index of the catch (IVIc) is a weighted
mean of the vulnerability of exploited fish species (Cheung et al.,
2007) and has been described to decrease with increasing fishing
pressure if the larger, longer species that are more vulnerable to
collapse are fished out and effort is then directed at smaller, faster
growing, less vulnerable species that are usually located lower in
the food web. The biomass (B) of endemic species in the community (Endemics B) and in the catch (C) (Endemics C) provide a
measure of how abundant endemic species are in the ecosystem
and which proportion of these important species are exploited (Coll
et al., 2012, 2016a, 2015b). The biomass of endangered species in
the community using the IUCN (International Union for Conservation of Nature (IUCN) Red List of species at risk (IUCN, 2015)
(IUCN species B) and in the catch (IUCN species C) provide an idea
on which part of the ecosystem is endangered and what proportion
of endangered species are exploited. Finally, the biomass and catch
of marine mammals, seabirds and reptiles specifically quantify the
abundance and exploitation of these key and iconic groups.
Note that the Ecopath Traits table needs to be populated with
species traits data for the ECOIND plug-in to produce results for all
of the indicators (Table 1). However, the new plug-in can be ran
without this trait information for those indicators for which the
necessary information is already available through the Ecopath
model initial parameterization.
2.4. Computational aspects and algorithms
ECOIND is a software module that is independently developed
of the EwE source code, and ties into the model execution flow via a
number of plug-in points that are provided by the EwE software
(Fig. 1) (Steenbeek et al., 2016). The plug-in system is an integral
part of the EwE software through which programmers can add
extensions to the EwE approach without having to change the EwE
source program code. The EwE software contains a wide range of
plug-in points placed at strategic locations throughout its program
code. External plug-in code modules, in turn, can specify interest in
connecting to any number of these plug-in points. When the EwE
software encounters a plug-in point during its execution, it checks
if there are any external code modules that wish to respond to that
plug-in point, and if so, the external code is activated with access to
data relevant to the plug-in point location in the EwE program flow.
To perform its calculations ECOIND connects to EwE plug-in
points that concern the execution of Ecopath, Ecosim, Ecospace,
126
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
Fig. 4. (continued).
and Monte Carlo (Fig. 1). When the EwE execution encounters one
of these plug-in points, ECOIND receives model estimates from the
EwE model component that just executed, enabling ECOIND to
perform its calculations. Within ECOIND, Ecopath, Ecosim, Ecospace and Monte Carlo indicators are computed via an algorithm
that draws its data from the output of the four different modules.
ECOIND Ecopath indicators are delivered as a single value per indicator, derived from the Ecopath estimates for a balanced model.
Ecosim indicators are computed as time series, with a single value
per indicator per computed Ecosim time step. Ecospace indicators
are computed for every Ecospace map cell with ecosystem dynamics for each computed Ecospace time step. Monte Carlo indicators are delivered in two formats: (i) as single values for each
alternative mass-balanced Ecopath model produced by Monte
Carlo which feed the Monte Carlo histogram (Fig. 4A), and (ii) as
time series, with a single value per indicator, per computed time
step for each Monte Carlo-triggered Ecosim run (Fig. 4B).
Display of indicators in the ECOIND user interface is straight
forward as it follows standard EwE display principles with exception of the Monte Carlo histogram. In the histogram computed
ECOIND indicator values are displayed by number of occurrence,
grouped in 100 equally sized intervals between the lowest to the
highest value of an indicator (Fig. 4A).
3. Results
To exemplify the functionality of ECOIND we used a previously
developed food web model representing the Mediterranean Sea
marine ecosystem of the South Catalan Sea (in the NW Mediterranean Sea, Fig. 3A), which had been previously fitted to time series
of data from 1978 to 2010 using Ecosim (Coll et al., 2013). The NW
Mediterranean Sea is a rich area in terms of marine biodiversity
(Coll et al., 2010) and is important for several marine predatory
species at risk including marine mammals and seabirds (Coll et al.,
2015b). The area is of relatively high productivity due to a combined effect of the northern current and runoff of the Ebro and
Rhone Rivers, and it is an important fishing ground for small pelagic
fish and demersal meso-predators. The food web model is
expressed in terms of biomass as t$km2, and production and catch
as t$km2$year1. It includes 40 functional groups, ranging from
phytoplankton to large predatory species (Fig. 3B), and covers an
area of 5000 km2 with depths from 50 to 400 m. A spatial version of
the food web model using the Ecospace habitat capacity model was
recently developed based on the fitted Ecosim model (Coll et al.,
2016b).
We use this case study to showcase how the indicators included
in ECOIND can provide additional information about the historic
ecosystem dynamics. For example, the ecological indicators obtained for 1978 and 2010, the initial and final year of the model
calibration (Table 1), highlight important changes in the ecosystem.
In 1978 the Southern Catalan Sea ecosystem was dominated by the
pelagic flows and biomasses, with an overall low trophic level of
the community. Fishing catches of organisms were mainly pelagic
and the trophic level of these catches was low, suggesting a predominance of small pelagic fish, demersal juvenile fish and invertebrates in fishing activities. In 2010 the biomass and catch of
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
127
Fig. 4. (continued).
commercial species had declined, mainly of pelagic fish, while the
biomass of demersal species, mainly invertebrates, increased. The
mean length, mean weight and mean life span of fish in the catch
increased due to the decline of pelagic fish. The biomass of endemic
species, IUCN species and of marine mammals, seabirds and marine
turtles declined as well, while the catch of the first two increased.
Results also evidence that the confidence around some of the
ecological indicators results varies largely due to uncertainty in
model inputs (Fig. 4A and Annex III).
Time series of ecological indicators show that the ecosystem
went through some important changes from 1978 to 2010 (Fig. 4B).
For example, we observe an increase of biomass of demersal organisms, mainly invertebrates, while the biomass of pelagic fish
fluctuate and show a declining trend at the end of the time series.
The biomass of predators and biodiversity indicator Q’ decline at
the beginning of the time series and mainly fluctuates with time.
The rest of indicators also show important ecosystem changes
(Annex III) and confirm the degradation state of the ecosystem
recently described using ecological indicators to track the effects of
fishing (Lockerbie et al., Submitted) and biodiversity analyses (Coll
et al., 2015b; Navarro et al., 2015). Spatial-temporal results evidence that these changes in the ecosystem were heterogenic in
space, and different patterns are seen between the coastal and the
continental shelf and open sea areas following environmental and
exploitation gradients (Fig. 4C and Annex III) (Coll et al., 2015b;
Navarro et al., 2015).
4. Discussion
The development of ECOIND plug-in required the development
of several features within the EwE software. An important addition
is the new capability to define taxa within each functional group of
the food web model. The taxonomic composition can be defined in
the Ecopath model (in the new “Define Taxa” form in Ecopath, Fig. 2
and Annex II). This enables better documentation of species information used to develop ecosystem models and to directly incorporate this data in modelling results. Since biodiversity analyses are
becoming necessary to complement ecosystem-based approaches
(Browman et al., 2005a; Link, 2011), we foresee that this capability
can be interesting to the EwE modelling community. A link between the EwE software and online databases such as FishBase,
WoRMS (the World Register for Marine Species), OBIS (Ocean
Biogeographic Information System) and others will facilitate the
task to populate the new taxonomic data tables. The WoRMS link
was publicly released with EwE version 6.5; other links are in
various stages of development.
The second development is the capability to associate species
traits and conservation status information with each species within
each functional group (in the new “Traits” form under Ecopath
Tools, Fig. 2 and Table 2, and Annex II). Information about the
128
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
Table 2
Descriptions of traits included in the new “Traits” form of Ecopath with Ecosim.
Entries
Description
Species
Organism type
Ecology
Name of the species
Type of the organism (Bacteria, Fungi, Algae, Plants, Invertebrates, Fishes, Birds, Mammals, Reptiles, Other)
Ecology of the organism (Bathydemersal, Bathypelagic, Benthic, Benthopelagic, Demersal, Pelagic, Pelagic-neritic, Pelagic-oceanic,
Reef-associated, Land-based)
Origin of the organisms (Native, Introduced, Endemic, Questionable)
Proportion of biomass of the species within the functional group. If there is only one species, then this is ¼ 1
Proportion of catch of the species within the functional group. If there is only one species, then this is ¼ 1
IUCN category of the species from the Red List (Not evaluated, Data deficient, Least concern, Near threatened, Vulnerable, Endangered,
Critically endangered, Extinct in the wild, Extinct)
Exploitation status of the exploited species from regional assessments (Not exploited, Underexploited, Moderately exploited, Fully exploited,
Overexploited, Depleted, Recovering)
Intrinsic extinction vulnerabilities of marine fishes to fishing sensu Cheung et al., 2005 [0, 100]
Average length (cm) of the species in the population
Maximum length (cm) of the species in the population
Average weight (g) of the species in the population
Mean life span (years) of the species in the population
Origin
Proportion of biomass
Proportion of catch
IUCN conservation status
Exploitation status
Vulnerability index
Mean length (cm)
Max length (cm)
Mean weight (g)
Mean life span (year)
species ecology (e.g. type of organism, such as invertebrate or fish),
species biological traits (e.g. mean length), species conservation
status (e.g. IUCN status) and species exploitation status (i.e.
moderately exploited, fully exploited) can be now included. This
broadens the capability of future EwE ecological analyses and indicators, needed to contribute to integrated assessments ()(such as
the MSFD and the Intergovernmental Platform on Biodiversity and
Ecosystem Services, IPBES frameworks, EC, 2010; IPBES, 2012).
Trait-based indicators can also contribute to the application of EwE
tools in the development of environmental impact analyses (Coll
et al., 2015a).
A third important new capability is the link between ECOIND
and the Monte Carlo simulation routine. This link enables the
calculation of multiple sets of values for a single indicator, both
from Ecopath and from Ecosim results, and facilitates to develop
further analyses of uncertainty in ecological indicators within EwE.
The ability to assess the impact of input uncertainty on modelling
results is an acknowledged need for the modelling community
(Bastin et al., 2013; Coll et al., 2015a; Lassalle et al., 2014). Future
software developments should focus on uncertainty of Ecospace
results.
Technical limitations of ECOIND include limited capability of
EwE models to represent size-based changes in food webs, and
current lack of variation in the proportion of species in catch and
biomass of each functional group if species are not parameterized
as individual groups or size groups. These limitations can currently
be partially overcome by parametrizing models using individual
species as functional groups or multi-stanza groups (Walters et al.,
2008). Further development could enable variation of sizeparameters and proportions of catch and biomass using forcing
functions. Data shortage in taxonomic resolution of some groups
(e.g. invertebrates) can limit the applicability, too. Therefore, the
interpretation of results from the ECOIND plug-in should take these
limitations into account.
Despite these limitations, the new capabilities of ECOIND open
the door to future applications of the EwE software to develop
automated analysis that can be used to test indicators in a standardized way and look at the responses of aquatic food webs to
single or multiple environmental factors and human activities.
ECOIND computes standardized valuable information that can be
useful to the application of EwE approach for environmental and
biodiversity analyses. For instance, it can be foreseen that the
fishing policy search tool of EwE (Christensen and Walters, 2004b)
could be further developed so as to optimize against biodiversity
indicator values computed by ECOIND. Currently, this is not yet
feasible as indicators are computed by external code modules and
cannot be delivered back to the EwE approach (Fig. 2, dashed lines).
We foresee an extension to the EwE system that not only brings
ECOIND indicators back to EwE for driving its dynamics, but that
can also benefit information produced by many other EwE plug-in
modules such as Network Analysis, Value Chain, or EcoTroph
(Steenbeek et al., 2016) (Fig. 2). ECOIND indicators can also be used
in external analyses using R packages, for example, expanding the
network analysis capabilities (e.g. enaR, Borret and Lau, 2014).
Acknowledgements
MC was partially funded by the European Commission through
the Marie Curie Career Integration Grant Fellowships e PCIG10-GA2011-303534 - to the BIOWEB project. This study is a contribution
to the projects ECOTRANS (CTM2011-26333, Ministerio de Economía
y Competitividad, Spain), BlueBRIDGE (EC 2020-EINFRA-9-2015RIA-Proposal number: 675680) and SafeNET (EU-DGMARE MARE/
2014/41).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.envsoft.2016.12.004.
References
Ainsworth, C.H., Pitcher, T.J., 2006. Modifying Kempton's species diversity index for
use with ecosystem simulation models. Ecol. Indic. 6 (3), 623e630.
Allesina, S., Bondavalli, C., 2004. WAND: an ecological network analysis userfriendly tool. Environ. Model. Softw. 19 (4), 337e340.
Bastin, L., Cornford, D., Jones, R., Heuvelink, G.B., Pebesma, E., Stasch, C., Nativi, S.,
Mazzetti, P., Williams, M., 2013. Managing uncertainty in integrated environmental modelling: the UncertWeb framework. Environ. Model. Softw. 39,
116e134.
Blanchard, J., Coll, M., Cotter, J., Link, J., Trenkel, V., Vergnon, R., Yemane, D., Shin, Y.J., 2010. Trend analysis of indicators: a comparison of recent changes in the
status of marine ecosystems around the world. ICES J. Mar. Sci. 67, 732e744.
Boldt, J.L., Martone, R., Samhouri, J., Perry, R.I., Itoh, S., Chung, I.K., Takahashi, M.,
2014. Developing ecosystem indicators for responses to multiple stressors.
Oceanography 27 (4), 116e133.
Borret, S.R., Lau, M.K., 2014. enaR: an R package for ecosystem network analyses.
Methods Ecol. Evol. 5, 1206e1213.
Browman, H.I., Cury, P.M., Hilborn, R., Jennings, S., Lotze, H.K., Mace, P.M.,
Murawski, S., Pauly, D., Sissenwine, M., Stergiou, K.I., Zeller, D., 2005a. Perspectives on ecosystem-based approaches to the management of marine resources. Mar. Ecol. Prog. Ser. 274, 269e303.
Browman, H.I., Stergiou, K.I., Agardy, T., Fluharty, D., Hirshfield, M.F., Livingston, P.A.,
Misund, O.A., Skjoldal, H.R., Rice, J.C., Rosenberg, A.A., 2005b. Politics and socioeconomics of ecosystem-based management of marine resources. Mar. Ecol.
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
Prog. Ser. 300, 241e296.
Cerco, C.F., Tillman, D., Hagy, J.D., 2010. Coupling and comparing a spatially-and
temporally-detailed eutrophication model with an ecosystem network model:
an initial application to Chesapeake Bay. Environ. Model. Softw. 25 (4),
562e572.
Cheung, W.W.L., Watson, R., Morato, T., Pitcher, T.J., Pauly, D., 2007. Intrinsic
vulnerability in the global fish catch. Mar. Ecol. Prog. Ser. 333, 1e12.
Cheung, W.W.L., Pitcher, T.J., Pauly, D., 2005. A fuzzy logic expert system to estimate
intrinsic extinction vulnerabilities of marine fishes to fishing. Biol. Conserv. 124,
97e111.
Christensen, V., 1996. Managing fisheries involving predator and prey species. Rev.
Fish Biol. Fish. 6 (4), 417e442.
Christensen, V., Coll, M., Steenbeek, J., Buszowski, J., Chagaris, D., Walters, C.J., 2014.
Representing variable habitat quality in a spatial food web model. Ecosystems
17 (8), 1397e1412.
Christensen, V., Walters, C., 2004a. Ecopath with Ecosim: methods, capabilities and
limitations. Ecol. Model. 72, 109e139.
Christensen, V., Walters, C., 2004b. Trade-offs in ecosystem-scale optimization of
fisheries management policies. Bull. Mar. Sci. 74 (3), 549e562.
Christensen, V., Walters, C., Pauly, D., Forrest, R., 2008. Ecopath with Ecosim Version
6. User Guide - November 2008. Lenfest Ocean Futures Project 2008, 235 pp.
Coll, M., Akoglu, E., Arreguín-S
anchez, F., Fulton, E.A., Gascuel, D., Heymans, J.J.,
Libralato, S., Mackinson, S., Palomera, I., Piroddi, C., Shannon, L.J., Steenbeek, J.,
Villasante, S., Christensen, V., 2015a. Modelling dynamic ecosystems: venturing
beyond boundaries with the Ecopath approach. Rev. Fish Biol. Fish. 25 (2),
413e424.
Coll, M., Libralato, S., 2012. Contributions of food-web modelling for an ecosystem
approach of marine resource management in the mediterranean sea. Fish Fish.
13, 60e88.
Coll, M., Navarro, J., Palomera, I., 2013. Ecological role of the endemic Starry ray Raja
asterias in the NW Mediterranean Sea and management options for its conservation. Biol. Conserv. 157, 108e120.
Coll, M., Piroddi, C., Albouy, C., Ben Rais Lasram, F., Cheung, W., Christensen, V.,
Karpouzi, V., Le Loc, F., Mouillot, D., Paleczny, M., Palomares, M.L., Steenbeek, J.,
Trujillo, P., Watson, R., Pauly, D., 2012. The Mediterranean Sea under siege:
spatial overlap between marine biodiversity, cumulative threats and marine
reserves. Glob. Ecol. Biogeogr. 21 (4), 465e480.
Coll, M., Piroddi, C., Kaschner, K., Ben Rais Lasram, F., Steenbeek, J., Aguzzi, J.,
Ballesteros, E., Nike Bianchi, C., Corbera, J., Dailianis, T., Danovaro, R., Estrada, M.,
Froglia, C., Galil, B.S., Gasol, M., Gertwagen, R., Gil, J., Guilhaumon, F., Kesnerpez-Fe
Reyes, K., Kitsos, M.-S., Koukouras, A., Lampadariou, N., Laxamana, E., Lo
de la Cuadra, C.M., Lotze, H.K., Martin, D., Mouillot, D., Oro, D., Raicevich, S.,
Rius-Barile, J., Saiz-Salinas, J.I., San Vicente, C., Somot, S., Templado, J., Turon, X.,
Vafidis, D., Villanueva, R., Voultsiadou, E., 2010. The biodiversity of the Mediterranean Sea: estimates, patterns and threats. PLoS ONE 5 (8) doi:10.1371.
, M.J., Bundy, A., Akoglu, A.G.,
Coll, M., Shannon, L.J., Kleisner, K., Juan Jorda
Banaru, D., Boldt, J.L., Borges, M.F., Cook, A., Diallo, I., Fu, C., Fox, C., Gascuel, D.,
Gurney, L.J., Hattab, T., Heymans, J.J., Jouffre, D., Knight, B.R., Kucukavsar, S.,
Large, S.I., Lynam, C., Machias, A., Marshall, K.N., Masski, H., Ojaveer, H.,
Piroddi, C., Tam, J., Thiao, D., Thiaw, M., Torres, M.A., Travers-Trolet, M.,
Tsagarakis, K., Tuck, I., van der Meeren, G.I., Yemane, D., Zador, S.G., Shin, Y.-J.,
2016a. Ecological indicators to capture the effects of fishing on biodiversity and
conservation status of marine ecosystems. Ecol. Indic. 60, 947e962.
Coll, M., Steenbeek, J., Ben Rais Lasram, F., Mouillot, D., Cury, P., 2015b. “Low hanging
fruits” for conservation of marine vertebrate species at risk in the Mediterranean Sea. Glob. Ecol. Biogeogr. 24 (2), 226e239.
Coll, M., Steenbeek, J., Sole, J., Palomera, I., Christensen, V., 2016b. Modelling the
cumulative spatial-temporal effects of environmental factors and fishing in a
NW Mediterranean marine ecosystem. Ecol. Model. 331, 100e114.
Cury, P., Christensen, V., 2004. Quantitative ecosystem indicators for fisheries
management. ICES J. Mar. Sci. 62 (3), 307e310.
EC, 2010. Decision on criteria and methodological standards on good environmental
status of marine waters. Decision 2010/477/EU. Official J. Eur. Union L 232,
14e24.
Fulton, E.A., 2010. Approaches to end-to-end ecosystem models. J. Mar. Syst. 81,
171e183.
Fulton, E.A., Fuller, M., Smith, A.D.M., Punt, A.E., 2005. Ecological Indicators of the
Ecosystem Effects of Fishing. Final Report. Australian Fisheries Management
Authority. Repor. R99/1546. 239 pp.
Halpern, B.S., Frazier, M., Potapenko, J., Casey, K.S., Koenig, K., Longo, C.,
Lowndes, J.S., Rockwood, R.C., Selig, E.R., Selkoe, K.A., Walbridge, S., 2015.
Spatial and temporal changes in cumulative human impacts on the world's
ocean. Nat. Commun. 6, 7615. http://dx.doi.org/10.1038/ncomms8615.
Heymans, J.J., Coll, M., Libralato, S., Morissette, L., Christensen, V., 2014. Global
patterns in ecological indicators of marine food webs: a modelling approach.
PLoS ONE 9 (4), e95845.
Hilborn, R., Walters, C., 1992. Quantitative Fisheries Stock Assessment. Choice, Dynamics and Uncertainty. Kluwer Academic Publishers.
IPBES, 2012. Intergovernmental Platform on Biodiversity and Ecosystem Services
(IPBES). World Wide Web electronic publication. http://www.ipbes.net/.
IUCN, 2015. IUCN Red List of Threatened Species. Version 2014.3. www.iucnredlist.
org. Accessed on 17 February 2015.
Jackson, J.B.C., Kirby, M.X., Berger, W.H., Bjorndal, K.A., Botsford, L.W., Bourque, B.J.,
Bradbury, R.H., Cooke, R., Erlandson, J., Estes, J.A., Hughes, T.P., Kidwell, S.,
Lange, C.B., Lenihan, H.S., Pandolfi, J.M., Peterson, C.H., Steneck, R.S., Tegner, M.J.,
129
Warner, R.R., 2001. Historical overfishing and the recent collapse of coastal
ecosystems. science 293 (5530), 629e638.
Kleisner, K.M., Longo, C., Coll, M., Halpern, B.S., Hardy, D., Katona, S., Le Manach, F.,
Pauly, D., Rosenberg, A., Samhouri, J., Scarborough, C., Sumaila, U.R., Watson, R.,
Zeller, D., 2013. Exploring patterns of seafood provision revealed in the global
ocean health index. AMBIO a J. Hum. Environ. 42 (8), 910e922.
Large, S.I., Fay, G., Friedland, K.D., Link, J.S., 2013. Defining trends and thresholds in
responses of ecological indicators to fishing and environmental pressures. ICES
J. Mar. Sci. J. du Conseil 70 (4), 755e767.
Large, S.I., Fay, G., Friedland, K.D., Link, J.S., 2015. Quantifying patterns of change in
marine ecosystem response to multiple pressures. PLoS ONE 10 (3), e0119922
e0119922.
at, B., Rochette, S., Niquil, N., 2014. A toolbox to
Lassalle, G., Bourdaud, P., Saint-Be
evaluate data reliability for whole-ecosystem models: application on the Bay of
Biscay continental shelf food-web model. Ecol. Model. 285, 13e21.
Levin, P.S., Fogarty, M.J., Murawski, S.A., Fluharty, D., 2009. Integrated ecosystem
assessments: developing the scientific basis for ecosystem-based management
of the ocean. PLoS Biol. 7 (1), e1000014.
Lindeman, R.L., 1942. The trophic-dynamic aspect of ecology. Ecology 23, 399e418.
Link, J., 2011. Ecosystem-based Fisheries Management: Confronting Tradeoffs.
Cambridge University Press, Cambridge.
Lockerbie, E., Coll, M., Shannon, L.J., Jarre, A., Submitted. The Use of Indicators for
Decision Support in Northwestern Mediterranean Sea Fisheries. Journal of
Marine Systems.
Lotze, H.K., Lenihan, H.S., Bourque, B.J., Bradbury, R.H., Cooke, R.G., Kay, M.C.,
Kidwell, S.M., Kirby, M.X., Peterson, C.H., Jackson, J.B.C., 2006. Depletion,
degradation, and recovery potential of estuaries and coastal seas. science 312
(5781), 1806e1809.
ndez, A.M., Bellido, J.M., 2015. The relative
Navarro, J., Coll, M., Cardador, L., Ferna
roles of the environment, human activities and spatial factors in the spatial
distribution of marine biodiversity in the Western Mediterranean Sea. Prog.
Oceanogr. 131, 126e137.
Odum, W.E., Heald, E.J., 1975. The detritus-based food web for an estuarine
mangrove community. In: Cronin, L.E. (Ed.), Estuarine Research, vol. 1. Academic
Press, New York.
Pauly, D., Christensen, V., Dalsgaard, J., Froese, R., Torres, F., 1998. Fishing down
marine food webs. science 279 (5352), 860e863.
Pauly, D., Watson, R., 2005. Background and interpretation of the ‘marine trophic
Index’as a measure of biodiversity. Philosophical Trans. R. Soc. B Biol. Sci. 360
(1454), 415e423.
Pikitch, E.K., Santora, C., Babcock, E.A., Bakun, A., Bonfil, R., Conover, D.O., Dayton, P.,
Doukakis, P., Fluharty, D., Heneman, B., Houde, E.D., Link, J., Livingston, P.A.,
Mangel, M., McAllister, M.K., Pope, J., Sainsbury, K.J., 2004. Ecosystem-based
fishery management. science 305 (5682), 346e347.
Piroddi, C., Teixeira, T., Lynam, C.P., Smith, C., Alvarez, M.C., Mazik, K., Andonegi, E.,
Churilova, T., Tedesco, L., Chifflet, M., Chust, G., Galparsoro, I., Garcia, A.C.,
€ma
€ri, M., Kryvenkof, O., Lassalle, G., Neville, S., Niquil, N., Papadopoulou, N.,
Ka
Rossberg, A.G., Suslin, V., Uyarra, M.C., 2015. Using ecological models to assess
ecosystem status in support of the. Eur. Mar. Strategy Framew. Dir. 58, 175e191.
Rochet, M.-J., Trenkel, V.M., 2003. Which community indicators can measure the
impact of fishing? A review and proposals. Can. J. Fish. Aquatic Sci. 60, 86e99.
Rombouts, I., Beaugrand, G., Artigas, L.F., Dauvin, J.-C., Gevaert, F., Goberville, E.,
Kopp, D., Lefebvre, S., Luczak, C., Spilmont, N., 2013. Evaluating marine
ecosystem health: case studies of indicators using direct observations and
modelling methods. Ecol. Indic. 24, 353e365.
Shannon, J.L., Coll, M., Bundy, A., Shin, Y.J., Travers-Trolet, M., Gascuel, D.,
Kleisner, K., Tam, J., Piroddi, C., Heymans, J.J., Lynam, C.P., 2014. Trophic levelbased indicators to track fishing impacts across marine ecosystems. Mar. Ecol.
Prog. Ser. 512, 115e140.
Shin, Y.J., Shannon, L.J., 2010. Using indicators for evaluating, comparing and
communicating the ecological status of exploited marine ecosystems. 1. The
IndiSeas project. ICES J. Mar. Sci. 67 (4), 686e691.
Smith, M.D., Fulton, E.A., Day, R.W., 2015. Using an Atlantis model of the southern
Benguela to explore the response of ecosystem indicators for fisheries management. Environ. Model. Softw. 69, 23e41.
Steenbeek, J., Buszowski, J., Christensen, V., Akoglu, E., Aydin, K., Ellis, N., Felinto, D.,
Guitton, J., Lucey, S., Kearney, K., Mackinson, S., Pan, M., Platts, M., Walters, C.,
2016. Ecopath with Ecosim as a model-building toolbox: source code capabilities, extensions, and variations. Ecol. Model. 319, 178e189.
Steenbeek, J., Coll, M., Gurney, L., Melin, F., Hoepffner, N., Buszowski, J.,
Christensen, V., 2013. Bridging the gap between ecosystem modeling tools and
geographic information systems: driving a food web model with external
spatialetemporal data. Ecol. Model. 263, 139e151.
Tittensor, D.P., Walpole, M., Hill, S.L.L., Boyce, D.G., Britten, G.L., Burgess, N.D.,
Butchart, S.H.M., Leadley, P.W., Regan, E.C., Alkemade, R., Baumung, R.,
Bellard, C., Bouwman, L., Bowles-Newark, N.J., Chenery, A.M., Cheung, W.W.L.,
Christensen, V., Cooper, H.D., Crowther, A.R., Dixon, M.J.R., Galli, A., Gaveau, V.,
€ ft, R.H., Januchowski-Hartley, S.R.,
Gregory, R.D., Gutierrez, N.L., Hirsch, T.L., Ho
Karmann, M., Krug, C.B., Leverington, F.J., Loh, J., Lojenga, R.K., Malsch, K.,
Marques, A., Morgan, D.H.W., Mumby, P.J., Newbold, T., Noonan-Mooney, K.,
Pagad, S.N., Parks, B.C., Pereira, H.M., Robertson, T., Rondinini, C., Santini, L.,
Scharlemann, J.P.W., Schindler, S., Sumaila, U.R., Teh, L.S.L., van Kolck, J.,
Visconti, P., Ye, Y., 2014. A mid-term analysis of progress toward international
biodiversity targets. science 346 (6206), 241e244.
n, F., 2010. CoSBiLab Graph: the network analysis module of
Valentini, R., Jorda
130
M. Coll, J. Steenbeek / Environmental Modelling & Software 89 (2017) 120e130
CoSBiLab. Environ. Model. Softw. 25 (7), 886e888.
Walters, C., Martell, S.J.D., Christensen, V., Mahmoudi, B., 2008. An Ecosim model for
exploring Gulf of Mexico ecosystem management options: implications of
including multistanza life-history models for policy predictions. Bull. Mar. Sci.
83 (1), 251e271.
Walters, C., Pauly, D., Christensen, V., Kitchell, J.F., 2000. Representing density
dependent consequences of life history strategies in aquatic ecosystems: EcoSim ii. Ecosystems 3 (1), 70e83.
Walters, C.J., Christensen, V., Martell, S.J., Kitchell, J.F., 2005. Possible ecosystem
impacts of applying MSY policies from single-species assessment. ICES J. Mar.
Sci. 62, 558e568.
Worm, B., Barbier, E.B., Beaumont, N., Duffy, J.E., Folke, C., Halpern, B.S.,
Jackson, J.B.C., Lotze, H.K., Micheli, F., Palumbi, S.R., Sala, E., Selkoe, K.A.,
Stachowicz, J.J., Watson, R., 2006. Impacts of biodiversity loss on ocean
ecosystem services. science 314, 787e790.
Zeller, D., Pauly, D., 2007. Reconstruction of marine fisheries catches for key
countries and regions (1950-2005). Fish. Centre Res. Rep. 15 (2), 163.