Elsevier

Ecological Modelling

Volume 331, 10 July 2016, Pages 100-114
Ecological Modelling

Modelling the cumulative spatial–temporal effects of environmental drivers and fishing in a NW Mediterranean marine ecosystem

https://doi.org/10.1016/j.ecolmodel.2016.03.020Get rights and content

Highlights

  • A spatial–temporal food web model of a NW Mediterranean ecosystem was used.

  • Individual and joint historical effects of environment and fishing were quantified.

  • Species distributions closely matched available information with the capacity model.

  • Impacts of environment and fishing on commercial species were mainly synergistic.

  • Fishing, distribution of primary producers and depth had highest impact on results.

Abstract

To realistically predict spatial–temporal dynamics of species in marine ecosystems it is essential to consider environmental conditions in conjunction with human activities and food web dynamics. In this study, we used Ecospace, the spatial–temporal dynamic module of Ecopath with Ecosim (EwE) food web model, to drive a spatially explicit marine food web model representing the Southern Catalan Sea (NW Mediterranean) with various environmental drivers and with fishing. We then evaluated the individual and joint effects of environmental conditions and fishing in various compartments of the food web. First we used a previously developed EwE model fitted to time series of data from 1978 to 2010 as a baseline configuration. The model included 40 functional groups and four fishing fleets. We first ran the original Ecospace spatial–temporal dynamic model using the original habitat configuration, in addition to fishing, and we predicted species distributions and abundances. Afterwards, we ran the new habitat foraging capacity model using the most important environmental drivers linked with the Ebro River delta dynamics (salinity, temperature, and primary production), in addition to depth, substrate and fishing, and we compared results with those from the original implementation of Ecospace. Three commercial species, European hake (Merluccius merluccius), anchovy (Engraulis encrasicolus) and sardine (Sardina pilchardus), were used to analyse results. Species distributions more closely matched the empirical information available from the study area when using the new habitat capacity model. Results suggested that the historical impacts of fishing and environmental conditions on the biomass and distributions of hake, anchovy and sardine were not additive, but mainly cumulative with a synergistic or antagonistic effect. Fishing had the highest impact on spatial modelling results while the spatial distribution of primary producers and depth followed in importance. This study contributes to the development of more reliable predictions of regional change in marine ecosystems of the Mediterranean Sea.

Introduction

Biodiversity, natural resources, and ecosystem services change in response to human stressors, such as climate change and fishing activities that often have cumulative effects (Crain et al., 2008, Darling and Côté, 2008, Halpern et al., 2008). The need to consider changes in the environment as well as human activities when analysing and managing marine ecosystems highlights the necessity to perform integrated analyses (Link, 2011). Since the productivity of marine resources depends on the ecological state of communities and ecosystems, and on external drivers, the dynamics of target species in conjunction with the dynamics of non-target organisms, trophic relationships and energy flows, environmental drivers and human impacts have to be considered to manage marine resources properly (Christensen et al., 2011).

In the last decade, the scientific community has made substantial progress in the identification and quantification of multiple human threats that impact marine diversity, habitats and ecosystems (Jackson et al., 2001, Lotze et al., 2006, Halpern et al., 2008). Currently, there is an increasing knowledge on the quantification of these multiple drivers, also in the Mediterranean Sea (Coll et al., 2012, Micheli et al., 2013). However, the way these drivers may interact and combine to impact patterns of marine ecosystems is not well known (Crain et al., 2008, Darling and Côté, 2008). Multiple drivers may interact and their effects may cause impacts that accumulate in an additive way (thus the sum of their individual impacts is equal to the total impact) or may act synergistically or antagonistically (when the sum of individual impacts are greater or lower than the separated effects, respectively) at different ecological levels (Mackinson et al., 2009, Ainsworth et al., 2011). Despite their importance, a comprehensive understanding of these impacts and their interactions is mostly lacking. Therefore, there is a growing need to develop and use novel methodologies of data integration, assimilation and modelling human impacts at different scales, taking into account spatial–temporal dynamics (Coll et al., 2013a).

Species distribution models and ecosystem models are two methodologies that are developed to study the spatial and temporal dynamics of marine resources and ecosystems (Guisan and Zimmermann, 2000, Fulton, 2010). The main aim of statistical modelling is to use the information about where a species occurs and the relationship with associated environmental conditions to predict how likely the species is to occur in non-sampled locations. Different techniques to apply species distribution models (SDM) exist and have been widely applied to the marine environment (e.g., Kaschner et al., 2001, Jones et al., 2012), including the Mediterranean Sea (e.g., Morfin et al., 2012, Pennino et al., 2013, Saraux et al., 2014). Despite their popularity, it is known that environmental drivers alone may not be sufficient to account for species distributions (Navarro et al., 2015). Other ecological processes, including trophic interactions (such as competition, predation, and facilitation), behavioural parameters, and population dynamics may affect the spatial arrangement of a species, in addition to human activities.

These processes can be considered in process-based oriented modelling, like in food web models such as Ecopath with Ecosim (EwE) approach (Christensen and Walters, 2004). EwE is a freely available software and approach that allows building food web models by describing the ecosystem by means of energy flows between functional groups, each representing a species, a sub-group of a species (e.g., juveniles and adults) or a group of species that have functional and ecological similarities. The functional groups can be set to represent consumers and primary producers, as well as non-living groups (e.g., detritus).

EwE consists of three main linked routines that have been developed over the last three decades: Ecopath, Ecosim and Ecospace. Ecopath is the mass-balance routine that allows building of spatially and time-averaged models of the trophic web (Polovina, 1984, Christensen and Pauly, 1992), and which serves to parameterize the dynamic modules of EwE; Ecosim is the time dynamic routine (Walters et al., 1997, Walters et al., 2000); and Ecospace is the spatial–temporal dynamic module that allows representing temporal and spatial 2D dynamics of trophic web components (Walters et al., 1999, Walters et al., 2010). Critical overviews of EwE approach and descriptions of the recent developments are available in the literature (Christensen and Walters, 2004, Steenbeek et al., 2014, Steenbeek et al., 2016, Ainsworth and Walters, 2015, Coll et al., 2015a, Heymans et al., 2016).

Ecospace has been widely applied to quantify the spatial impact of fisheries on marine species (Christensen et al., 2003, Christensen et al., 2014a, Christensen et al., 2015), to analyse the impact of management scenarios such as the establishment of marine protected areas, and to assess the correlation of spatial distributions of marine species and fishing effort (Walters, 2000, Martell et al., 2005, Fouzai et al., 2012). Ecospace can also be used to develop spatial optimization routines (Christensen et al., 2009) and to assess the impact of climate change on marine ecosystems by linking Ecospace with low trophic level models (Fulton, 2011) or by driving Ecospace directly with external spatial–temporal data (Steenbeek et al., 2013, Steenbeek et al., 2016).

Despite their capabilities, available tools offered limited capacity to model cumulative impacts of stressors and produce integrated assessments that take both food web dynamics and spatial–temporal environmental and human variability into account (Steenbeek et al., 2013). To overcome some limitations, the habitat foraging capacity model was recently added to Ecospace (Christensen et al., 2014b). This new development offers the ability to spatially drive foraging capacity of species from the cumulative effects of multiple physical, oceanographic, and environmental conditions (such as depth, bottom type, temperature, salinity, oxygen concentrations and primary production). The habitat foraging capacity model runs in Ecospace in conjunction with the food web and fisheries dynamics. This development, in combination with the spatial–temporal framework module (Steenbeek et al., 2013), bridged the gap between envelope environmental models and classic food web models (Christensen et al., 2014b, Christensen et al., 2015).

In this study we applied the new habitat foraging capacity model (Christensen et al., 2014b) to study the distribution of three commercially important fish species (European hake Merluccius merluccius, sardine Sardina pilchardus and anchovy Engraulis encrasicolus) of the Southern Catalan Sea (NW Mediterranean Sea, Fig. 1) and evaluate the combined effects of environmental drivers (primary production, salinity, temperature, depth and substrate), in addition to fishing, and food web structure in their dynamics.

The NW Mediterranean Sea is one of the most important fishing grounds of the Mediterranean Sea, particularly the region surrounding the Ebro Delta area (Coll et al., 2012). Important environmental parameters drive the dynamics of commercial species in the area, such as small pelagic fish and important demersal species (Lloret et al., 2004, Palomera et al., 2005, Palomera et al., 2007, Martín et al., 2008, Ospina-Alvarez et al., 2015). Previous studies looking at the temporal dynamics of marine resources identified that both environmental factors, human activities and the structure of the food web were key elements to predict ecosystem dynamics (Coll et al., 2006, Coll et al., 2008, Mackinson et al., 2009).

Section snippets

Study area

Our study area was located on the Catalan Sea (within the Balearic Sea, NW Mediterranean Sea, Fig. 1). The NW Mediterranean Sea is an area of relatively high productivity due to a combined effect of the Northern current and the runoff of the Ebro and Rhone Rivers (Estrada, 1996, Bosc et al., 2004). In the northern area, the continental shelf is narrower, with the Northern current flowing south-westwards along the continental slope, towards the wider continental shelf surrounding the Ebro Delta

Comparison of Ecospace original with the capacity model

As a general trend, the overall biomasses and catches results from the implementation of the spatial dimension showed relevant decreases compared to the baseline Ecosim. Moreover the biomass and catch results for the Ecospace original model (Spatial_EO) were more optimistic than the ones resulting from the new habitat capacity model (Spatial_CM). Under Spatial_EO, the biomass of adult hake, anchovy and sardine showed a decrease of 77%, 38% and 33%, respectively, from 1978 to 2010 (Fig. 3a and

Comparison of Ecospace original with the capacity model

This study is one of the first applications of the new habitat capacity model in Ecospace (Christensen et al., 2014b) and illustrates its capabilities. It shows that the spatial–temporal modelling at the local level is possible if environmental, ecological and human activities are included with enough spatial detail, a challenge that it is present in most of the modelling initiatives to date (Barange et al., 2010).

Species distributions and abundances resulting from the new habitat capacity

Acknowledgements

MC was partially funded by the European Commission through the Marie Curie Career Integration Grant Fellowships – PCIG10-GA-2011-303534 – to the BIOWEB project and by the Spanish National Program Ramon y Cajal. This study is a contribution to the project ECOTRANS (CTM2011-26333, Ministerio de Economía y Competitividad, Spain).

References (92)

  • M. Coll et al.

    Trophic flows, ecosystem structure and fishing impacts in the South Catalan Sea, Northwestern Mediterranean

    J. Mar. Syst.

    (2006)
  • J.I. Díaz et al.

    Late Pleistocene and Holocene sedimentary facies on the Ebro continental shelf

    Mar. Geol.

    (1990)
  • N. Fouzai et al.

    Fishing management scenarios to rebuild exploited resources and ecosystems of the Northern-Central Adriatic (Mediterranean Sea)

    J. Mar. Syst.

    (2012)
  • E.A. Fulton

    Approaches to end-to-end ecosystem models

    J. Mar. Syst.

    (2010)
  • A. Guisan et al.

    Predictive habitat distribution models in ecology

    Ecol. Modell.

    (2000)
  • M.C. Jones et al.

    Modelling commercial fish distributions: prediction and assessment using different approaches

    Ecol. Modell.

    (2012)
  • D. Macias et al.

    Atmospheric-induced variability of hydrological and biogeochemical signatures in the NW Alboran Sea, Consequences for the spawning and nursery habitats of European anchovy

    Deep Sea Res. I

    (2011)
  • S. Mackinson et al.

    Which forcing factors fit? Using ecosystem models to investigate the relative influence of fishing and changes in primary productivity on the dynamics of marine ecosystems

    Ecol. Modell.

    (2009)
  • F. Maynou et al.

    Seasonal and spatial variability of hake (Merluccius merluccius L.) recruitment in the NW Mediterranean

    Fish. Res.

    (2003)
  • J. Navarro et al.

    The relative roles of the environment, human activities and spatial factors in the spatial distribution of marine biodiversity in the Western Mediterranean Sea

    Prog. Oceanogr.

    (2015)
  • M.P. Olivar et al.

    Spatial and temporal distribution and abundance of European hake, Merluccius merluccius, eggs and larvae in the Catalan coast (NW Mediterranean)

    Fish. Res.

    (2003)
  • I. Palomera et al.

    Small pelagic fish in the NW Mediterranean Sea: an ecological review

    Prog. Oceanogr.

    (2007)
  • M.G. Pennino et al.

    Modeling sensitive elasmobranch habitats

    J. Sea Res.

    (2013)
  • A. Shchepetkin et al.

    The Regional Ocean Modeling System (ROMS): a split-explicit, free-surface, topography-following coordinates ocean model

    Ocean Modell.

    (2005)
  • J. Steenbeek et al.

    Ecopath with Ecosim as a model-building toolbox: source code capabilities, extensions, and variations

    Ecol. Modell.

    (2016)
  • J. Steenbeek et al.

    Bridging the gap between ecosystem modeling tools and geographic information systems: driving a food web model with external spatial–temporal data

    Ecol. Modell.

    (2013)
  • R. Abad et al.

    Acoustic estimation of abundance and distribution of anchovy in the NW Mediterranean

    Sci. Mar.

    (1998)
  • M. Adani et al.

    Quality assessment of a 1985–2007 Mediterranean Sea reanalysis

    J. Atmos. Ocean. Technol.

    (2011)
  • R.N.M. Ahrens et al.

    Foraging arena theory

    Fish Fish.

    (2012)
  • C.H. Ainsworth et al.

    Potential impacts of climate change on Northeast Pacific marine foodwebs and fisheries

    ICES J. Mar. Sci.

    (2011)
  • E. Bosc et al.

    Seasonal and interannual variability in algal biomass and primary production in the Mediterranean Sea, as derived from 4 years of SeaWiFS observations

    Global Biogeochem. Cycles

    (2004)
  • E. Calvo et al.

    Effects of climate change on Mediterranean marine ecosystems: the case of the Catalan Sea

    Clim. Res.

    (2012)
  • V. Christensen et al.

    The global ocean is an ecosystem: simulating marine life and fisheries

    Global Ecol. Biogeogr.

    (2015)
  • V. Christensen et al.

    Fish biomass in the world ocean: a century of decline

    Mar. Ecol. Prog. Ser.

    (2014)
  • V. Christensen et al.

    Representing variable habitat quality in a spatial food web model

    Ecosystems

    (2014)
  • V. Christensen et al.

    Hundred-year decline of North Atlantic predatory fishes

    Fish Fish.

    (2003)
  • V. Christensen et al.

    Ecopath with Ecosim: A user's guide

    (2005)
  • V. Christensen et al.

    Ecopath with Ecosim Version 6. User Guide – November 2008. Lenfest Ocean Futures Project 2008

    (2008)
  • M. Coll et al.

    Modelling dynamic ecosystems: venturing beyond boundaries with the Ecopath approach

    Rev. Fish Biol. Fish.

    (2015)
  • M. Coll et al.

    Assessing fishing and marine biodiversity changes using fishers’ perceptions: the Spanish Mediterranean and Gulf of Cadiz case study

    PLOS ONE

    (2014)
  • M. Coll et al.

    The scientific strategy needed to promote a regional ecosystem-based approach to fisheries in the Mediterranean and Black Seas

    Rev. Fish Biol. Fish.

    (2013)
  • M. Coll et al.

    The Mediterranean Sea under siege: spatial overlap between marine biodiversity, cumulative threats and marine reserves

    Global Ecol. Biogeogr.

    (2012)
  • M. Coll et al.

    Ranking the ecological relative status of exploited marine ecosystems

    ICES J. Mar. Sci.

    (2010)
  • M. Coll et al.

    “Low hanging fruits” for conservation of marine vertebrate species at risk in the Mediterranean Sea

    Global Ecol. Biogeogr.

    (2015)
  • F. Colloca et al.

    Rebuilding Mediterranean fisheries: a new paradigm for ecological sustainability

    Fish Fish.

    (2013)
  • C.M. Crain et al.

    Interactive and cumulative effects of multiple human stressors in marine systems

    Ecol. Lett.

    (2008)
  • Cited by (60)

    View all citing articles on Scopus
    View full text