A GIS-based framework for addressing conflicting objectives in the context of an ecosystem approach to fisheries management—a case study of the Portuguese sardine fishery

A GIS-based framework for addressing conflicting objectives in the context of an ecosystem... Abstract An ecosystem approach to fisheries management (EAFM) is as a new paradigm in fisheries management. In this study, a combination of geographic information systems (GISs) and multi-criteria decision-making method (MCDM) is proposed as a framework supporting an ecosystem approach to European sardine (Sardina pilchardus, Clupeidae) fishery management in Portugal. This case study was chosen due to the recent severe decline shown by the species. To develop an EAFM for the sardine fishery, a set of criteria were defined based on literature review and expert knowledge. To address multiple conflicting objectives, namely conservation and fisheries, five scenarios were considered: (i) baseline; (ii) nurseries protection; (iii) spawning areas protection; (iv) fishery profit driven, and (v) safeguarding dependent fishing communities. Combination of GIS and MCDM methods highlighted important areas to implement spatial conservation measures for sardine. The analyses indicate that some areas are suitable for conservation in several scenarios, such as the area near Aveiro and the area near the Tejo estuary. However, conservation measures implemented in the area near Aveiro would imply higher economic trade-offs when compared with the actions applied in the region near the Tejo estuary. Results also suggested some of the conservation objectives, such as the protection of sardine eggs and juveniles, to not be compatible. The proposed framework is an important tool supporting EAFM by addressing conflicting objectives, trade-offs and identifying areas that could be considered as potential fishery closure sites or subjected to further analyses. Introduction Depletion of fish stocks worldwide (FAO, 2014) signal a failure of conventional fisheries management practices and ineffective governance (Ruckelshaus et al., 2008). Common Fisheries Policy (CFP) was settled with the aim to integrate environmental concerns and move towards a more holistic ecosystem approach to fisheries management (EAFM) (Ramírez-Monsalve et al., 2016a). Key characteristics of this approach include multispecies consideration, trophic interactions, socio-economic dimension, and impact on habitat and ecosystem (Garcia et al., 2003). There is a large body of literature presenting guidelines and frameworks aiming to facilitate EAFM operationalization (Fletcher, 2008; Link, 2010; Staples et al., 2014) and several methods have been explored. The most popular and frequently studied have been: (i) modelling approaches e.g. Ecosystem models (Polovina, 1984; Christensen and Pauly, 1992; Walters et al., 1997, 1999, Fulton et al., 2004a, 2004b); (ii) statistical methods that support Integrated Ecosystem Assessment (Link et al., 2002), (iii) spatial methods, such as the use of geographic information systems (GISs) (Olson, 2011), (iv) application of indicators (Link, 2005; Yeon et al., 2011), and finally (v) the use of socio-ecological frameworks (Leite and Gasalla, 2013). Even though, significant progress has been made towards a better operationalization of an EAFM in a recent CFP reform [regulation (EU) No. 1380/2013], several challenges, such as lack of transparent and legitimate framework to address and balance conflicting objectives and trade-offs, impede the implementation of an EAFM in EU fisheries (Ramírez-Monsalve et al., 2016b) The use of GIS, a versatile tool which can be applied at different stages of the fisheries management process, has been recommended by the Food and Agriculture Organization (FAO) as a tool for an EAFM (Carocci et al., 2009). GIS application in fisheries management has been already implemented in coastal regions around the world (Aswani and Lauer, 2006; Hall and Close, 2007; De Freitas and Tagliani, 2009). Despite its potential, the use of GIS in EAFM is still limited (Carocci et al., 2009; Meaden and Aguilar-Manjarrez, 2013). Multi-criteria decision making (MCDM) methods have the ability to assist in the management of renewable resources by taking into account conflicting objectives, therefore attaining interests within a wide range of management fields (Romero and Rehman, 1987). They have also been frequently applied to fisheries (Mardle and Pascoe, 1999; Chiou et al., 2005; Kjærsgaard, 2007; Innes and Pascoe, 2010). Nevertheless, spatial MCDM methods, that allow for the evaluation of multiple conflicting criteria to solve complex spatial problems (Malczewski, 2006), have had limited use in fisheries and coastal management studies (Hossain and Das, 2010; Dapueto et al., 2015) (Villa et al., 2002; Portman, 2007). In the EAFM context, this approach can be used to indicate suitable areas for biodiversity conservation, while minimizing the negative effect on fisheries. European sardine (Sardina pilchardus, Walbaum, 1792) represents an economically and socially significant fishery resource in Portugal and Spain (Silva et al., 2015). Sardine in the Iberian Peninsula Atlantic waters (ICES Divisions 8.c and 9.a) is considered a stock for management purposes (ICES, 2017a). In the last decade, its severe decline in the Iberian waters has been observed: biomass decreased around 80% and recruitment reached the lowest historical level in 40 years (ICES, 2017a). The majority of the Iberian sardine stock is distributed in Portuguese waters and its main recruitment and spawning areas are the northern Portugal and south of Spain (Gulf of Cadiz); (Bernal et al., 2007; Rodríguez-Climent et al., 2017). Clupeoid fishes, like sardine, are usually short-lived and the success of their recruitment is crucial to further increase the size of the population (Cole and Mcglade, 1998). These fish species usually show high variability in the recruitment process related to environmental conditions (Bakun, 1997). In this paper we describe a framework that demonstrates the utility of GIS to support fisheries management in the context of an ecosystem approach. It implements a GIS-based MCDM method combined with broad spatial analysis using the Iberian sardine as a case study. The presented framework has the capability to facilitate EAFM by identifying areas which (i) are of the highest importance for conservation, (ii) are the most crucial for fisheries, and (iii) imply trade-offs between conservation and fisheries objectives. Based on this information, the most appropriate choices for conservation were presented. Material and methods Study area Located on the western tip of Europe and with an extension of ∼52 000 km2, the Portuguese continental shelf (0–200 m; Figure 1) is in a biogeographic transition zone between temperate and subtropical waters (Briggs, 1974). Moreover, the western Portugal coast is located alongside the northernmost part of the Canary Current Upwelling System, one of the four major eastern boundary upwelling systems of the World Ocean classified as a Class I highly productive ecosystem (>300 g cm−2 year−1) (Carr, 2002). In this area, seasonal upwelling occurs during spring and summer as a result of steady northerly winds (Wooster et al., 1976; Fiúza et al., 1982). Figure 1. View largeDownload slide Map showing limits of the study area between 200 m isobath and Portuguese coastline, occurrence of sardine in Portuguese acoustic surveys between 2005 and 2010 (source: IPMA, PELAGO surveys); purse-seine fishing grounds between 2005 and 2010 (source: DGRM VMS-logbook data) and acoustic transect along which biological data were collected (IPMA, PELAGO surveys). Figure 1. View largeDownload slide Map showing limits of the study area between 200 m isobath and Portuguese coastline, occurrence of sardine in Portuguese acoustic surveys between 2005 and 2010 (source: IPMA, PELAGO surveys); purse-seine fishing grounds between 2005 and 2010 (source: DGRM VMS-logbook data) and acoustic transect along which biological data were collected (IPMA, PELAGO surveys). GIS-MCDM analysis A combination of GIS and MCDM methodology was used to identify the most suitable sites to preserve essential sardine habitats, while reducing the negative effect on fishing activities. A set of criteria representing a holistic ecosystem approach and considering sardine conservation and fishing industry, was defined. To select the most suitable option from available alternatives, MCDM technique (decision rule, algorithm), with the weighted linear combination (WLC) (weighted overlay) method, was used. Another widely used method, analytical hierarchy process (AHP) was employed to derive the weights associated with attribute map layers used in WLC method. Applied GIS-based MCDM procedure (Figure 2) consists of five steps: (i) identification of important criteria; (ii) data compilation; (iii) creation of criteria maps; (iv) production of suitability maps; and (v) assessment of suitability maps. Figure 2. View largeDownload slide Conceptual suitability model showing applied criteria and GIS-based MCDM procedure applied to select conservation areas that aim to protect sardine essential habitat and at the same time attempt to maintain the socio-economic efficiency of purse-seine fishery in coastal Portuguese waters. Figure 2. View largeDownload slide Conceptual suitability model showing applied criteria and GIS-based MCDM procedure applied to select conservation areas that aim to protect sardine essential habitat and at the same time attempt to maintain the socio-economic efficiency of purse-seine fishery in coastal Portuguese waters. Step 1—identification of criteria The first step of the suitability analysis was a selection of important criteria for sardine essential habitat and purse-seine fishery, and the development of a conceptual suitability model accordingly (Figure 2). The model consisted of 12 criteria representing three main dimensions of EAFM: biological, environmental and socio-economic (Tables 1–3). Criteria selection was based on the best available published literature and expert judgement. Table 1. Description of biological criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey a ESDU—elemental sample distance unit that corresponds to 1 nm. Table 1. Description of biological criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey a ESDU—elemental sample distance unit that corresponds to 1 nm. Table 2. Description of environmental criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Table 2. Description of environmental criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Table 3. Description of socio-economic criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Table 3. Description of socio-economic criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Step 2—data compilation The second step of the analysis was the compilation of spatial data for each criterion. The spatial data compiled covered a period of 6 years (2005–2010; with few exceptions listed in Tables 1–3). This time scale represents a period with contrasting sardine abundance, high numbers in the years 2005–2007 and low numbers in the years 2008–2010 (ICES, 2017a). Data regarding adult [Total length (TL) > 16 cm] and juvenile (TL ≤ 16 cm) sardine biomass, sardine eggs abundance, and zooplankton concentration were obtained from the acoustic surveys (PELAGO survey series) performed annually by the Portuguese Institute of Sea and Atmosphere (IPMA) in spring (April–May). Surveys were carried out onboard the R/V Noruega along predefined parallel transects perpendicular to the coastline covering the whole platform (0- to 200-m depth) with 8 nautical miles (nm) inter-transect distance (Figure 1). Acoustic fish data was obtained following the methodology described in ICES (2016) by performing echo integration along 1-nm Elemental sampling distance unit (ESDU), used as distance measure in acoustic surveys, and supported by fish samples collection through pelagic and bottom trawling. Estimates of sardine biomass were calculated following the procedures described in Simmonds and MacLennan (2005). For more details about acoustic surveys and sardine biomass estimates refer to publications by ICES (2016) and Rodríguez-Climent et al. (2017). Sardine eggs and zooplankton data were collected along the acoustic transects every 3 nm at the surface (∼3 m), using a Continuous Underway Fish Egg Sampler (CUFES, mesh 335 μm) (Checkley et al., 1997). Sardine eggs abundance (no./m3) was calculated based on the eggs data collected per unit of time and the pump flow rate. Zooplankton biomass expressed as volume of zooplankton (ml/10 m3) was estimated by volumetric method (displacement volume; Ikeda and Omori, 1992). Sardine competitors’ species abundance data: atlantic chub mackerel (Scomber colias, Gmelin 1789), bogue (Boops boops, Linnaeus 1758), anchovy (Engraulis encrasicolus, Linnaeus 1758), mediterranean horse mackerel (Trachurus mediterraneus, Steindachner, 1868) and blue jack mackerel (Trachurus picturatus, Bowdich, 1825), were assessed by acoustic data (Nautical Area Scattering Coefficient, NASC—SA m2/nm2) derived from the PELAGO surveys. The remaining competitors: horse mackerel (Trachurus trachurus, Linnaeus 1758) and atlantic mackerel (Scomber scombrus, Linnaeus 1758) and sardine predators, hake (Merluccius merluccius, Linnaeus 1758) and squid (Loligo spp. Lamarck, 1798) data were based on abundance data (kg/h) collected during demersal surveys (2005–2010). Those surveys were carried out in autumn (September–October) onboard the R/V Noruega at predefined stations in the Portuguese Exclusive Economic Zone at depths varying from 20- to 500-m depth. For details about abundance estimation based on demersal surveys refer to International bottom trawl surveys protocol published by ICES (2017b). Vessel monitoring system (VMS) and logbook data from purse-seine vessels ≥ 12 m (EU Council, 2009) for the years 2006–2010, obtained from the DGRM, were integrated to produce maps of fishing activity. Purse-seine catch distribution maps were developed based on analysing vessel speed patterns from VMS data and linking them to logbook data to identify fishing activity according to Katara and Silva (2017). We reported landings as the sum of tonnes of each species landed per year and per grid cell. VMS and logbook data for the year 2005 were not used because their spatial coverage was not sufficient and representative enough. Instead, the official landing data per main fishing port, registered and obtained from the DGRM were used for this year. These data were spatially attributed to fishing grounds previously defined according to spatial patterns recorded in 2006–2010 VMS-logbook data (Supplementary Material S1 and Supplementary Figure S1). For the criteria “other fisheries” the fleets using the following gear were considered: bottom trawl, longline, gill, trammel, and pots. Spatial effort (hours) and landings (tonnes) data (2005–2010) at ICES statistical rectangles level were obtained from the Scientific Technical and Economic Committee for Fisheries (STECF) (annex: WW; country: PRT; https://stecf.jrc.ec.europa.eu, 20 May 2017). Sea surface temperature (°C) and chlorophyll a concentration data (mg/m3) were obtained from MODIS aqua (L3; 4 μm) and SeaWIFS satellite sensor respectively (sources: www.opendap.jpl.nasa.gov/opendap/; www.oceancolour.org/portal/, 10 May 2017) and represented the period of the pelagic surveys: March–May Salinity was collected in situ during the IPMA acoustic surveys (PELAGO, 2005–2010). Depth (m) data were obtained from the Portuguese Hydrographic Institute website (www.hidrografico.pt/download-gratuito.php, 25 May 2017). Step 3—creation of criteria maps To create surface continuous layers, a similar approach to the one performed in Petitgas et al. (2014) allowing for a reasonable compromise between smoothing and maintaining details, was taken. Data collected along acoustic transects and at the demersal survey’s sampling stations were averaged on a common regular grid (8 × 8 km) using ArcMap neighbourhood point statistic function that allows the creation of cell-based rasters with a smoothing effect. Satellite data were converted to the 8 × 8 km grid to match the resolution of the remaining layers. All analyses were performed using ESRI ArcMap software (version: 10.4.1). Socio-economic, environmental, and some biological criteria—competitors and predators—required processing of initial data in order to be represented as meaningful maps useful for this study. Generation of socio-economic criteria The following socio-economic criteria calculated yearly were considered in this study (i) sardine catch per unit of effort (CPUE); (ii) sardine income; (iii) purse-seine fishery dependency from sardine; and (iv) other fisheries. Sardine CPUE was calculated by considering the effort to be the duration of a fishing trip times the number of sets per trip (where sardine was caught) except for the year 2005 where the effort was considered to be the number of fishing trips where sardine was caught. For the other fleets, effort was considered to be the number of fishing hours. Discards were negligible therefore we considered that landings per unit effort (LPUE) were equal to CPUE. Income was calculated by multiplying landings of the species of interest by its price (euro per kg). Prices used were those indicated in the Portuguese National Statistical Institute (INE) annual statistic report (INE 2006, 2007, 2008, 2009, 2010, 2011). Purse-seine fishery dependency from sardine was calculated by dividing sardine landings by total landings. The criterium ‘other fisheries’ is a combination of CPUE, landings and income for the respective fleet, i.e. bottom trawl, longline, gill, trammel, and pots. In order to combine all these variables, the spatial layers representing these variables, were first reclassified using the Jenks natural breaks method (Jenks, 1967) and then averaged using GIS raster algebra function. Jenks natural breaks method is the default classification method in ArcMap which identifies breakpoints between classes using a statistical formula (Jenks optimization) that minimizes the sum of the variance within each of the classes finding groups and patterns inherent to the data. Generation of suitable habitat criteria Maps representing suitable habitat for juvenile sardine, adult sardine and sardine eggs were produced using environmental variables that affect sardine’s distribution according to published literature. Environmental variables were reclassified in accordance with Rodríguez-Climent et al., (2017) for sardine juveniles, Zwolinski et al. (2010) for adults and Bernal et al. (2007) for eggs. The reclassified environmental maps were then combined by summing the reclassified layers, corresponding to each variable with raster algebra (raster calculator ArcGis function). Generation of biotic interactions criteria In this analysis, biotic interactions were represented by sardine competitors and predators. Areas with high number of competitors or predators were treated as a priority for conservation, as they constitute natural stress for sardine. Moreover, these areas were considered as priority because in these locations sardine seems to play a fundamental role to ecologically support other species and the whole marine ecosystem—as forage species (Pikitch et al., 2012). Species commonly caught along with sardine by the purse-seine fleet were considered as sardine competitors namely: S. colias, S. scombrus, T. trachurus, B. boops, E. encrasicolus, T. mediterraneus and T. picturatus. Species like M. merluccius and Loligo spp. were considered as common sardine predators (Coelho et al., 1997; Cabral and Murta, 2002; Mahe et al., 2007). Other important sardine predators, such as the common dolphin (Delphinus delphis L.1758) (Silva, 1999), were not considered in this study as a result of data availability limitations. Note that some of the species considered as sardine competitors (e.g. S. colias and B. boops) in this study are also predators for sardine eggs (Garrido et al., 2015). The overlap found between sardine and their competitors/predators, will be referred as spatial overlap intensity index (Oi) and was estimated using the Jaccard index of similarity (Jaccard, 1908) modified as follows: Oi=∑i=1n (PIL & SPiPIL | SPi × SPi )×Afi (1) where Oi is overlap between sardine and other species considered, PIL is sardine biomass; SPiis other species abundance; n is the number of other species considered and Afi is the other species abundance factor. Other species either refers to competitors in Competitors-sardine overlap or predators in the Predators-sardine overlap. The Jaccard index is one of the oldest and most widely used similarity indices for assessing compositional similarity of assemblages. It is based on the presence-absence records of species in paired assemblages (Chao et al., 2005). In order to reflect species abundance in the output of the binary index, the index was modified by adding multiplication by two variables: First, variable SPi that corresponds to species abundance in the used dataset and then by variable Af that is general species abundance factor. Available competitors’ species data had various units as they were obtained from various surveys. Data regarding sardine predators and competitors, atlantic mackerel and horse mackerel, were available from demersal surveys, while the remaining competitors’ data was obtained from the acoustic surveys. This fact made it mathematically impossible to apply Equation (1). Therefore in order to address different units in data, each species-sardine overlap raster was reclassified into 10 classes using Jenks natural breaks. Nevertheless, the outputs of the procedure of reclassification made it impossible to differentiate between very abundant species, such as horse mackerel, from low abundant species, such as bogue. Therefore, to address the importance of each species abundance relatively to each other, the reclassified rasters were multiplied by an abundance factor Af established for this purpose. This subjective abundance factor was attributed to each competitor species based on its 6-years abundance estimates from bottom Portuguese trawl (ICES data 2009–2014) and subjective classification established by author judgement, that aimed to highlight the differences between species abundance (see Table 4 for more details). After multiplications by SPi variable and Af factor, raster layers representing competitors and predators were summed. Table 4. Abundance factor assigned to each competitor species established based on its abundance in bottom Portuguese trawl data (source: ICES, 2009–2014). Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Table 4. Abundance factor assigned to each competitor species established based on its abundance in bottom Portuguese trawl data (source: ICES, 2009–2014). Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Step 4—production of suitability maps The final suitability maps for sardine conservation without compromising fisheries activities were produced using the WLC method. This method is one of the most widely used GIS-based decision-making methods for land suitability analysis (Hopkins, 1977; Malczewski, 2000). It was chosen because it is accessible, intuitively appealing to decision makers (Massam, 1988) and easy to implement in GIS (Tomlin, 1991). The WLC method involves the standardization of the attribute maps, assigning the weights of relative importance and combining the weighted criteria to obtain an overall suitability score. Therefore, all criterion maps were standardized, weighted and combined to produce the final suitability map as follows; Standardization Criteria were standardized to a uniform scale with values ranging from one to five (1: no suitability; 5 high suitability for conservation; Tables 1–3). The applied classification method was Jenks natural breaks (Jenks, 1967). Scenario definition The analysis was performed for five scenarios that varied depending on the priority objectives. The first scenario balanced conservation and fisheries objectives. The second and third scenario focussed on conservation objectives, while the fourth and fifth scenario focussed on fisheries objectives. Scenario 1, the baseline scenario, considered all criteria to have the same importance. Scenario 2, nurseries protection, aimed to protect sardine nursery areas. This scenario applied increased weight for the following criteria: juvenile biomass and suitable juvenile habitat. Scenario 3, spawning areas protection, aimed to protect sardine spawning grounds. It applied increased weight for criteria: sardine eggs abundance, adult sardine biomass (considered potential spawners) and suitable sardine eggs and adult sardine habitat. Biological criteria were obtained from observed values, while habitat criteria were output of modelling studies (for details see subsection Generation of suitable habitat criteria). Therefore, as a result of a higher uncertainty of the latter, they were assigned lower priority compared with the biological criteria. Scenario 4, fishery profit driven, aimed to protect sardine fishery revenue. Here increased weight was applied to sardine income criteria. Scenario 5, safeguarding dependent fishing communities, aimed to exclude from protection areas where fishers are highly dependent on sardine catches. It applied increased weight to areas were purse-seine fishery is highly dependent on sardine. Weight assignment The AHP method (Saaty, 1980) was used to assign weights to criteria, which differ depending on the applied scenario. This mathematical method is the most frequently used in spatial MCDM studies (Phua and Minowa, 2005; Bottero et al., 2013; Rahman et al., 2013) and it establishes weights among criteria by making a series of judgements based on pairwise comparisons of the criteria. The relative importance of each criterion is measured according to a numerical scale from one to nine (Table 5). In this study, weights were chosen to represent potential differences in the probable stakeholders’ objectives. Even though, stakeholders’ weren’t directly engaged, as it was out of the scope of the study, the framework design and a choice of methods ensure their inclusion when necessary. Once a pairwise comparison matrix was built, each criterion weight was calculated by computing the eigenvector of each column. Second, the matrix components were normalized by averaging the values across the rows (Table 6). Table 5. Scale for AHP pairwise comparisons and description how scales were attributed to criteria in each scenario. Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Table 5. Scale for AHP pairwise comparisons and description how scales were attributed to criteria in each scenario. Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Table 6. Weight values for each criterion in each scenario derived by the AHP method. Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 The bold text indicates an increased weight applied to the criterion that was considered the most important in the scenario. Table 6. Weight values for each criterion in each scenario derived by the AHP method. Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 The bold text indicates an increased weight applied to the criterion that was considered the most important in the scenario. Integration of criteria maps Each criterion map (Figure 3) is composed of i pixels. Each i pixel is characterized by attributes which describe location (coordinate data) and value xij, representing standardized score of ith pixel with respect to the jth criterion (attribute value is associated with the location). The suitability of each ith pixel is evaluated using the following equation: Si=∑j=1nwjxij (2) where S is suitability of the ith pixel in the final map; n is a number of criteria (n = 12); wj is a weight of jth criterion and xij is the standarized score for the ith pixel with respect to the jth criterion. Figure 3. View largeDownload slide Example of the WLC method of criteria raster maps (redrawn from Pérez et al., 2005; Dapueto et al., 2015). The suitability map is a result of the linear combination of j criteria (in the figure j refers to the A and B criteria maps, with weights 0.65 and 0.35, respectively). Each pixel i on criteria map j is represented by a standardized value xij. Figure 3. View largeDownload slide Example of the WLC method of criteria raster maps (redrawn from Pérez et al., 2005; Dapueto et al., 2015). The suitability map is a result of the linear combination of j criteria (in the figure j refers to the A and B criteria maps, with weights 0.65 and 0.35, respectively). Each pixel i on criteria map j is represented by a standardized value xij. Firstly, in order to identify the conflicting areas, biological and environmental criteria and socio-economic criteria were combined separately. Then combination of all criteria was performed for five different scenarios. Model verification—sensitivity analysis Sensitivity analysis (SA) with a one-at-a-time method (Daniel, 1973, Chen et al. 2010) was performed to evaluate how weights influenced criteria and the spatial patterns on the suitability maps. This method introduces changes in one factor (weight) at a time, while all the other factors remain fixed, and detects how this change might influence the output. A range of percent change of ±100% and increment of percent range of ±1% were applied to the complete set of criteria used in this study. Therefore, the SA consisted of 200 simulation runs for each criterion. It was performed by using Model Builder module of ArcMap that enables semi-automatic processing of a large data set and options. Step 5—assessment of suitability maps For each scenario, areas with high and very high suitability conservation score (4 and 5, respectively) were selected (Supplementary Figure S3) in order to be analysed in terms of costs to fishing industry and benefits for sardine conservation (Table 7). Among them, only areas with at least two grid cells (minimum of 128 km2) were considered. As costs the following aspects were considered: (i) loss in sardine fishery profit and (ii) loss in profit from the other fisheries. As benefits the protection of sardine was considered for: (i) juvenile; (ii) spawning areas; and (iii) adults. Table 7. Summary of potential costs and benefits estimated for each selected area in the five tested scenarios. Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Data are shown as yearly summary in terms of yearly loss in fisheries income and amount of protected sardine biomass and eggs. a Average refers to eggs abundance column. Table 7. Summary of potential costs and benefits estimated for each selected area in the five tested scenarios. Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Data are shown as yearly summary in terms of yearly loss in fisheries income and amount of protected sardine biomass and eggs. a Average refers to eggs abundance column. Results Sensitivity analysis The SA showed that all criteria have a relatively similar sensitivity to change in weights. Effects were first observed after ∼ ±20% change from the baseline. Further changes were observed after each additional ±10 or ±5% changes in the input. Changes higher than ±80% from a base did not cause any further changes in the output. This pattern was similar for most of the criteria (For details refer to Supplementary Figure S2). Sardine conservation and socio-economic related criteria Areas with the highest scores for the biological and environmental criteria occurred in the following areas: south of Aveiro (Area 1), near Figueira de Foz (Area 2), Ericeira (Area 3), near the Tejo estuary, between Cabo Raso and Cabo Espichel (Area 4) and near Setúbal (Area 5) (Figure 4a). Area 1 had high scores for juvenile, adult sardine biomass and sardine eggs abundance (Figure 4b–d). Area 2, had a high level of juvenile biomass whereas Area 3, had medium levels of adult biomass, predators, and competitors (Figure 4b, c, e, and f). Also, Area 4 had high levels of competitors and predators but here high, not adult, but juvenile sardine biomass obtained higher scores (Figure 4b, c, and e). Finally, Area 5 had high competitors levels and moderate eggs abundance (Figure 4d and e). All areas had medium to high levels of suitable juvenile, adult and eggs habitat (Figure 4g–i) with the exception of Area 5, which showed low levels of suitable adult sardine habitat (Figure 4h). Figure 4. View largeDownload slide Map showing all biological and environmental criteria combined. Areas that obtained their highest scores were highlighted (a); Maps showing means of each biological and environmental criterion separately (b–i). Figure 4. View largeDownload slide Map showing all biological and environmental criteria combined. Areas that obtained their highest scores were highlighted (a); Maps showing means of each biological and environmental criterion separately (b–i). On the other hand, if we considered the socio-economic criteria (Figure 5a), areas 1–3 had high levels of socio-economic features (Figure 5a). Therefore, those areas were identified as conflicting areas. On the other hand, areas 4 and 5 (Figure 4a) had a low level of almost all socio-economic criteria (Figure 5a–d) except for other fisheries feature that showed medium levels (Figure 5e). Figure 5. View largeDownload slide Map showing all socio-economic criteria combined. Areas that obtained the highest scores were highlighted (a); Maps showing means of each socio-economic criterion separately (b–h). Figure 5. View largeDownload slide Map showing all socio-economic criteria combined. Areas that obtained the highest scores were highlighted (a); Maps showing means of each socio-economic criterion separately (b–h). Overlay of all criteria for various scenarios The suitability for sardine conservation varied across scenarios (Figure 6). Figure 6. View largeDownload slide Suitability classification for the Iberian sardine stock conservation areas for the different scenarios (a–e). Figure 6. View largeDownload slide Suitability classification for the Iberian sardine stock conservation areas for the different scenarios (a–e). The baseline scenario had the highest number of areas (min 128 km2—two pixels) with moderate suitability for conservation (Figure 6a—yellow grid cells) and small and fragmented areas of high or very high suitability for conservation (Figure 6a—orange and red grid cells, Table 7). Those areas were situated nearby Aveiro and Nazaré and the larger ones near Ericeira, Tejo, and Setúbal. In the south of Portugal, there were also a few regions with high suitability for conservation (Figure 6a). In the remaining scenarios, the high and very highly suitable areas were more aggregated and constrained to specific regions. In the nurseries protection scenario, the main suitable areas were situated near Aveiro, Ericeira, and Tejo (Figure 6b). Similarly, in the spawning areas protection scenario, these areas were also situated near Aveiro and Ericeira and additionally near Setúbal (Figure 6c). In the scenarios that focussed on socio-economic features, the suitable areas for conservation were mainly situated in the area near Tejo, Setúbal for fishery profit driven scenario and near Ericeria, Tejo and Setúbal for safeguarding dependent fishing communities’ scenario. Moreover, for both scenarios, the areas situated in the south of Portugal were identified as suitable (Figure 6d and e). Assessment of selected areas—cost-benefit analysis Cost and benefits in scenarios varied depending on selected areas and prioritized objectives (Table 7). If considered scenarios and areas they encompass as total, it is difficult to select the scenario that best balanced conflicting objectives. For example, nurseries protection scenario provided protection for the highest quantity of juvenile 85% and adult sardine 29%. However, this conservation benefits were associated with the high cost in fisheries income that was estimated for 16 and 9% of total income for purse seine and other fisheries. On the other hand, baseline scenario provided protection for the second highest quantity of juvenile (51%) and adult (28%) sardine. However, in this scenario, the impact on the fisheries income was comparable to the impact observed in nurseries protection scenario and was estimated for 15 and 10% for purse seine and other fisheries, respectively. The analysis of the main areas within scenarios individually makes the interpretation of the compromise between fisheries and conservation easier. For example, areas situated in the north were the most suitable for juvenile and adult sardine protection. Their highest numbers were observed in: nurseries protection, baseline, and spawning areas protection scenarios. However, protection of these areas would be associated with the high cost for fishery that vary from 6% in baseline scenario and about 12% of affected income, in nurseries protection and spawning areas protection scenarios. Moreover, protection of north areas would not benefit spawning areas that are mainly situated in the area near Setúbal. On the other hand, protection of the area near Setúbal, would benefit considerably protection of sardine eggs (areas with high eggs density—180% higher than the observed mean), but would have negligible effect on juvenile (<1% of protected juvenile biomass) and adult (∼ 4% of protected adult biomass) sardine protection. This area would also have minimal impact on fisheries, affecting only 1% of purse seine and <0.5% of other fisheries income. On the other hand, the area near Tejo, excluding spawning areas protection scenario, was the area that protected the second highest biomass of juvenile sardine (between 12 and 17% of total biomass) and between 3 and 5% of total adult sardine. It also had relatively low impact on the fisheries income (1–2% on purse-seine fishery and 2–4% on other fisheries). Therefore, when comparing cost and benefits of protecting the Tejo area to the North area, it can be stated that with the first a higher percentage of juvenile sardine would be protected, with a lower economic impact on the purse-seine fishery. Discussion This study describes a framework that combines GIS and MCDM methodology to advance EAFM. The most important aspect of this approach is its ability to address multiple conflicting objectives and at the same time integrate and synthesize large amounts of information relevant to EAFM into a single analysis. This is essential in EAFM (Garcia and Cochrane, 2005) and has been highlighted to have the potential to significantly facilitate the management process (Leslie and McLeod, 2007). Nowadays, GIS and spatial analysis are frequently applied in fisheries management (Witt and Godley, 2007; Maxwell et al., 2009; Moore et al., 2009) but there is still a lack of a more direct, integrative application of GIS in EAFM, as the one presented in this study. Carocci et al. (2009) stressed the importance of GIS application in EAFM and suggest that it can be especially useful as a first step of a decision support framework for an EAFM, namely in issue identification. Therefore, GIS can be useful in identification of areas which: (i) require conservation, (ii) are particularly important to fishing industry, and/or (iii) are in conflict depending on stakeholders’ objectives. Moreover, the adoption and visualization of alternative scenarios and assessment of their trade-offs is particularly important in the management processes of any complex system (Costanza, 1998). The approach described in this study deal with both these tasks. Adoption of five exploratory scenarios enabled us to identify (i) important areas for specific objectives; (ii) conflicting areas between fisheries and conservation objectives; and (iii) suitable areas for sardine conservation. The area near Aveiro was identified as suitable for both nurseries and spawning protection scenarios but was also identified as one of the most conflicting due to its crucial role in fisheries. In this area, the income from sardine fishery is high as well as the dependence of purse-seine fishery on sardine. Thus, a permanent closure of fishing in this area seems unlikely, as would not be economically and socially accepted, even though the conservation benefits might be high. To rebuild the sardine stock, temporary closures during spawning season and in areas with high juvenile abundance might be an important management solution. In this study, the spatial analysis repeatedly highlighted the area near Tejo estuary to be appropriate for conservation. This area is an important spot for sardine recruitment, but not as important for the fishing sector. Its closure has the potential to protect between 12 and 17% of total mean juvenile biomass (observed between 2005 and 2010) barely compromising economic objectives, as the affected purse-seine income was estimated to be 1% of the total income. The design of marine reserves is a process in which some economic losses are inevitable therefore the analysis and discussion of trade-offs is very important (Stewart and Possingham, 2005). Exploratory scenarios such as the ones presented in this study, can make a valuable contribution to the marine reserve design process through the indication of the magnitude of trade-offs assumed. In our case study, the most important trade-offs were between protecting juvenile sardine and minimizing the negative economic impacts. The analysis also revealed the two conservation objectives, eggs abundance, and juvenile biomass, to be not totally compatible. Thus, the protection of one of them does not imply the protection of the other, suggesting the difficulty of setting one area able to protect sardine in all its vital life-stages (egg, juvenile, and adult). The application of spatial closures as a conservation measure in fisheries is difficult (Jones, 2006) and if not properly performed might end in frustration and failure (Fiske, 1992). The effectiveness of such closures have not always been successful (Abbott and Haynie, 2012). There are some examples where the implementation of an MPA resulted in higher aggregation of fishing vessels around the protected areas, a phenomenon called “fishing the line” (Kellner et al., 2007; Stelzenmüller et al., 2008). On the other hand, there are also numerous success stories of spatial closures applied as management measure (Halpern and Warner, 2002). The potential benefits include increased species richness, biomass and size of organisms within the reserve (Lester et al., 2009). Another important enquiry when considering spatial closures, is the extent to which they should or can be applied in practice. The duration of the protection and the size of the no-take areas have showed to play an important role in MPA effectiveness (Vandeperre et al., 2011). Should the spatial closures encompass only the fishery that targets sardine or be a total closure of fishing activities in the area? What about other sectors such as tourism or recreational fishing? And how do they affect other species and trophic interactions? These are examples of some questions emerging when talking about spatial closures in EAFM (Halpern et al., 2010). In this study, an attempt was made to deal with those questions by including criteria such as “other fisheries” and biotic interactions (sardine competitors and predators). As shown in this study closure of all fishing activities might considerably affect the fishing industry. Biotic interactions are very difficult to represent in this kind of analysis and their involvement was limited to mapping their distribution and indicating the areas overlapping with sardine. To obtain information about how spatial closures might affect trophic interactions and the ecosystem, the application of more sophisticated methods, such as ecosystem models (Polovina, 1984; Christensen and Pauly, 1992; Walters et al., 1997, 1999; Fulton et al., 2004a, 2004b) might be required. For example, Colléter et al. (2014), used food web models of Mediterranean and Senegalese ecosystems and found out the benefits of an MPA, such as increased biomass and catch yields, to be dependent on the MPA size. One of the limitations of the performed analysis was data availability. Lack of long time-series impeded the inclusion of statistical predictive modelling which have the capability to reduce uncertainty related to variability in tested variables. GIS environment support integration of spatial analysis with other tools in order to represent the temporal domain (Sahin and Mohamed, 2013; Chen et al., 2015), but these approaches are extremely data demanding, and in our case study, spatio-temporal data, especially related to VMS fishery data, were not available. In our study, 6 years of data were averaged and subjected to the analysis to reduce the uncertainty in the spatial distribution. The period considered is a good representation of our case study because it covers a time when both, high, and low sardine biomass, was observed (ICES, 2017a). Another important data limitation is the lack of seasonality. Variables included in the analysis, such as productivity patterns, zooplankton, sardine eggs and sardine distribution, tend to vary with season (Bernal et al., 2007; Moita et al., 2010; Zwolinski et al., 2010; Sobrinho-Gonçalves et al., 2013). However, Portuguese surveys do not collect seasonal data for these variables. This limitation might increase the uncertainty of the results and consequently impede the implementation of temporal and permanent conservation areas, thus more studies should be done to analyse the temporal variation of those variables and their effect on sardine biomass. One of the study’s strengths was the application of the AHP process to select weights. Although, it is a common practice in MCDM studies (Phua and Minowa, 2005; Chen et al., 2010; Gdoura et al., 2015) sometimes it is criticized (Dyer, 1990). Nevertheless, this process is appropriate for this type of analysis because it enables the inclusion of multiple stakeholders (Phua and Minowa, 2005; De Feo and De Gisi, 2010). Stakeholder involvement in EAFM decision support frameworks is essential and a highly recommended practice (Varjopuro et al., 2008; Kincaid et al., 2017). Even though, stakeholders were not included in the weights selection in this study, the design and methodology applied easily enables their inclusion in the future. In spatial analysis the uncertainty is introduced at its various stages (Refsgaard et al., 2007; Mosadeghi et al., 2013). In our case, uncertainty can arise from satellite data resolution conversion, creation of raster maps from point observations, identification of criteria and weights selection. In the case of criteria weights estimation, uncertainty was addressed through the performance of a SA which revelled that majority of variables are quite stable to the changes introduced by the weights (Supplementary Figure S1). Uncertainty was not quantitatively assessed in other stages, but its presence is acknowledged and therefore results should be interpreted with caution. Furthermore, it should be kept in mind that an array of other available statistical approaches, which could substitute the methods used in the present work (e.g. WLC method, AHP, Jenks natural breaks reclassification method), exist and their selection might considerably influence the results. The biggest difficulty in the application of this type of methodologies is the integration of the various data, disciplines, and institutions together (Leslie and McLeod, 2007). The united effort, collaboration and data sharing between scientific institutions, governments, fisheries institutions and non-governmental organizations is extremely important to a successful development of this and other data-hungry, integrative methods. In this study, scientists specialized in distinctive fields were involved. Moreover, various type of data, e.g. fisheries, satellite, landings, and biological were compiled and processed. Conclusion The proposed framework presents a useful tool to support an EAFM implementation. The analysis performed with Portuguese sardine fishery as a study case, proved that a combination of GIS and MCDM methodology could synthesize existing information in a fishery system, aid in the visualization of spatial existing problems and pinpoint areas which require special attention. This tool stands out as an efficient first step in an EAFM decision support framework. Further development of the method should start by performing seasonal analysis that will reflect annual variation in species distribution. Other aspects that should be addressed in the future are the involvement of stakeholders in the weights selection process and quantitative assessment of uncertainty in the other stages of the analysis, especially surface maps production and method selection. Supplementary data Supplementary material is available at the ICESJMS online version of the manuscript. Acknowledgements The work of D. Szalaj was supported by an FCT doctoral grant, Ref: PD/BD/114540/2016. Acoustic and bottom trawl surveys were financed by the EU/DG Fisheries’ Data Collection Framework (DCF). Logbook and VMS data were provided by the Portuguese Direcção-Geral de Recursos Marinhos Segurança e Serviços Marinhos. The authors thank to everyone who was involved in data collection and processing. Also, we would like to acknowledge editor and two anonymous reviewers for their constructive criticism that considerably improved this manuscript. References Abbott J. K. , Haynie A. C. 2012 . What are we protecting? Fisher behavior and the unintended consequences of spatial closures as a fishery management tool . Ecological Applications , 22 : 762 – 777 . Google Scholar CrossRef Search ADS PubMed Aswani S. , Lauer M. 2006 . Incorporating fishers’ local knowledge and behavior into geographical information systems (GIS) for designing marine protected areas in oceania . Human Organization , 65 : 81 – 102 . Google Scholar CrossRef Search ADS Bakun A. 1997 . Patterns in the Ocean: Ocean Processes and Marine Population Dynamics, Vol. 17, pp. 1945 – 1946 . California Sea Grant, La Jolla, CA, and Centro de Investigaciones Biologicas del Noroeste, La Paz, BCS, Mexico. Bernal M. , Stratoudakis Y. , Coombs S. , Angelico M. M. , de Lanzós A. L. , Porteiro C. , Sagarminaga Y. 2007 . Sardine spawning off the European Atlantic coast: characterization of and spatio-temporal variability in spawning habitat . Progress in Oceanography , 74 : 210 – 227 . Google Scholar CrossRef Search ADS Bottero M. , Comino E. , Duriavig M. , Ferretti V. , Pomarico S. 2013 . The application of a Multicriteria Spatial Decision Support System (MCSDSS) for the assessment of biodiversity conservation in the Province of Varese (Italy) . Land Use Policy , 30 : 730 – 738 . Google Scholar CrossRef Search ADS Briggs J. C. 1974 . Marine Zoogeography. McGraw Hill Book Company , New York . 475 pp. Cabral H. N. , Murta A. G. 2002 . The diet of blue whiting, hake, horse mackerel and mackerel off Portugal . Journal of Applied Ichthyology , 18 : 14 – 23 . Google Scholar CrossRef Search ADS Carocci F. , Bianchi G. , Eastwood P. , Meaden G. 2009 . Geographic Information Systems to support the ecosystem approach to fisheries: status, opportunities and challenges . Aquaculture , 101 . FAO Fisheries and Aquaculture Technical Paper, 532. Rome, FAO; 101 pp. Carr M. E. 2001 . Estimation of potential productivity in Eastern Boundary Currents using remote sensing . Deep-Sea Research Part II: Topical Studies in Oceanography , 49 : 59 – 80 . Google Scholar CrossRef Search ADS Chao A. , Chazdon R. L. , Colwell R. K. , Shen T. 2005 . A new statistical approach for assessing similarity of species composition with incidence and abundance data. Ecology Letters, 8: 148–159. Checkley D. M. , Ortner P. B. , Settle L. R. , Cummings S. R. 1997 . A continuous, underway fish egg sampler . Fisheries Oceanography , 6 : 58 – 73 . Google Scholar CrossRef Search ADS Chen Y. , Yu J. , Khan S. 2010 . Spatial sensitivity analysis of multi-criteria weights in GIS-based land suitability evaluation . Environmental Modelling and Software , 25 : 1582 – 1591 . Google Scholar CrossRef Search ADS Chen Y. , Liu R. , Barrett D. , Gao L. , Zhou M. , Renzullo L. , Emelyanova I. 2015 . A spatial assessment framework for evaluating flood risk under extreme climates . Science of the Total Environment , 538 : 512 – 523 . Google Scholar CrossRef Search ADS PubMed Chiou H. K. , Tzeng G. H. , Cheng D. C. 2005 . Evaluating sustainable fishing development strategies using fuzzy MCDM approach . Omega , 33 : 223 – 234 . Google Scholar CrossRef Search ADS Christensen V. , Pauly D. 1992 . ECOPATH II - a software for balancing steady-state ecosystem models and calculating network characteristics . Ecological Modelling , 61 : 169 – 185 . Google Scholar CrossRef Search ADS Coelho M. , Domingues P. , Balguerias E. , Fernandez M. , Andrade J. P. 1997 . A comparative study of the diet of Loligo vulgaris (Lamarck, 1799) (Mollusca: cephalopoda) from the south coast of Portugal and the Saharan Bank (Central-East Atlantic) . Fisheries Research , 29 : 245 – 255 . Google Scholar CrossRef Search ADS Cole J. , Mcglade J. 1998 . Clupeoid population variability, the environment and satellite imagery in coastal upwelling systems . Reviews in Fish Biology and Fisheries , 8 : 445 – 471 . Google Scholar CrossRef Search ADS Colléter M. , Gascuel D. , Albouy C. , Francour P. , Tito de Morais L. , Valls A. , Le Loc'h F. 2014 . Fishing inside or outside? A case studies analysis of potential spillover effect from marine protected areas, using food web models . Journal of Marine Systems , 139 : 383 – 395 . Google Scholar CrossRef Search ADS Costanza R. 1998 . Principles for sustainable governance of the oceans . Science , 281 : 198 – 199 . Google Scholar CrossRef Search ADS PubMed Daniel C. 1973 . One-at-a-time plans . Journal of the American Statistical Association , 68 : 353 – 360 . Google Scholar CrossRef Search ADS Dapueto G. , Massa F. , Costa S. , Cimoli L. , Olivari E. , Chiantore M. , Federici B. et al. 2015 . A spatial multi-criteria evaluation for site selection of offshore marine fish farm in the Ligurian Sea, Italy . Ocean and Coastal Management , 116 : 64 – 77 . Google Scholar CrossRef Search ADS De Feo G. , De Gisi S. 2010 . Using an innovative criteria weighting tool for stakeholders involvement to rank MSW facility sites with the AHP . Waste Management , 30 : 2370 – 2382 . Google Scholar CrossRef Search ADS PubMed De Freitas D. M. , Tagliani P. R. A. 2009 . The use of GIS for the integration of traditional and scientific knowledge in supporting artisanal fisheries management in southern Brazil . Journal of Environmental Management , 90 : 2071 – 2080 . Google Scholar CrossRef Search ADS PubMed Dyer J. S. 1990 . Remarks on the Analytic Hierarchy Process . Management Science , 36 : 249 – 258 . Google Scholar CrossRef Search ADS Council E. U. , 2009 . Official Journal of the European Union L 300/1, 2009: 1 – 33 . FAO . 2014 . The State of World Fisheries and Aquaculture 2014. Rome. 223 pp. Fiske S. J. 1992 . Sociocultural aspects of establishing marine protected areas . Ocean and Coastal Management , 17 : 25 – 46 . Google Scholar CrossRef Search ADS Fiúza A. F. D. , de Macedo M. E. , Guerreiro M. R. 1982 . Climatological space and time variation of the Portuguese coastal upwelling . Oceanologica Acta , 5 : 31 – 40 . Fletcher W. J. 2008 . Implementing an ecosystem approach to fisheries management: lessons learned from applying a practical EAFM framework in Australia and the Pacific. In The Ecosystem Approach to Fisheries , pp. 112 – 124 . Ed. by Bianchi G. , Skjodal H. R. . Food and Agriculture Organization of the United Nations (FAO ), Rome, Italy. Google Scholar CrossRef Search ADS Fulton E. A. , Smith A. D. M. , Johnson C. R. 2004a . Effects of spatial resolution on the performance and interpretation of marine ecosystem models . Ecological Modelling , 176 : 27 – 42 . Google Scholar CrossRef Search ADS Fulton E. A. , Parslow J. S. , Smith A. D. M. , Johnson C. R. 2004b . Biogeochemical marine ecosystem models II: the effect of physiological detail on model performance . Ecological Modelling , 173 : 371 – 406 . Google Scholar CrossRef Search ADS Garcia S. , Zerbi A. , Aliaume C. , Do Chi T. , Lasserre G. 2003 . The ecosystem approach to fisheries. Issues, terminology, principles, institutional foundations, implementation and outlook . FAO Fisheries Technical Paper , 443 : 71 . Garcia S. M. , Cochrane K. L. 2005 . Ecosystem approach to fisheries: a review of implementation guidelines . ICES Journal of Marine Science , 62 : 311 – 318 . Google Scholar CrossRef Search ADS Garrido S. , Silva A. , Pastor J. , Dominguez R. , Silva A. V. , Santos A. M. 2015 . Trophic ecology of pelagic fish species off the Iberian coast: diet overlap, cannibalism and intraguild predation . Marine Ecology Progress Series , 539 : 271 – 286 . Google Scholar CrossRef Search ADS Gdoura K. , Anane M. , Jellali S. 2015 . Geospatial and AHP-multicriteria analyses to locate and rank suitable sites for groundwater recharge with reclaimed water. Resources, Conservation and Recycling, 104: 19–30. Hall G. B. , Close C. H. 2007 . Local knowledge assessment for a small-scale fishery using geographic information systems . Fisheries Research , 83 : 11 – 22 . Google Scholar CrossRef Search ADS Halpern B. S. , Warner R. R. 2002 . Marine reserves have rapid and long lasting effects . Ecology Letters , 5 : 361 – 366 . Google Scholar CrossRef Search ADS Halpern B. S. , Lester S. E. , McLeod K. L. 2010 . Placing marine protected areas onto the ecosystem-based management seascape . Proceedings of the National Academy of Sciences of the United States of America , 107 : 18312 – 18317 . Google Scholar CrossRef Search ADS PubMed Hopkins L. D. 1977 . Methods for generating land suitability maps: a comparative evaluation . Journal of the American Planning Association , 43 : 386 – 400 . Hossain M. S. , Das N. G. 2010 . GIS-based multi-criteria evaluation to land suitability modelling for giant prawn (Macrobrachium rosenbergii) farming in Companigonj Upazila of Noakhali, Bangladesh . Computers and Electronics in Agriculture , 70 : 172 – 186 . Google Scholar CrossRef Search ADS ICES . 2016 . Second Interim report of the Working Group on Acoustic and Egg Surveys for Sardine and Anchovy in ICES Areas VII, VIII and IX (WGACEGG), 16–26 November 2015, Lowestoft, UK. ICES CM: 2015/SSGIEOM: 31. 398 pp. ICES . 2017a . Report of the Working Group on Southern Horse Mackerel, Anchovy and Sardine (WGHANSA), 24–29 June 2017, Bilbao, Spain. ICES CM: 2017/ACOM: 17. 547 pp. ICES . 2017b . Manual of the IBTS North Eastern Atlantic Surveys . Series of ICES Survey Protocols SISP , 15 : 92 . Ikeda T. , Omori M. 1992 . Methods in Marine Zooplankton Ecology. Krieger Plubishing Company , Florida . 329 pp. INE–Instituto Nacional de Estatística . 2006 . Estatísticas da Pesca 2005. INE, Lisboa, 79 pp. INE–Instituto Nacional de Estatística . 2007 . Estatísticas da Pesca 2006. INE, Lisboa, 97 pp. INE–Instituto Nacional de Estatística . 2008 . Estatísticas da Pesca 2007. INE, Lisboa, 97 pp. INE–Instituto Nacional de Estatística . 2009 . Estatísticas da Pesca 2008. INE, Lisboa, 98 pp. INE–Instituto Nacional de Estatística . 2010 . Estatísticas da Pesca 2009. INE, Lisboa, 101 pp. INE–Instituto Nacional de Estatística . 2011 . Estatísticas da Pesca 2010. INE, Lisboa, 104 pp. Innes J. P. , Pascoe S. 2010 . A multi-criteria assessment of fishing gear impacts in demersal fisheries . Journal of Environmental Management , 91 : 932 – 939 . Elsevier Ltd. Google Scholar CrossRef Search ADS PubMed Jaccard P. 1908 . Nouvelles recherches sur la distribution florale . Bulletin de la Société Vaudoise Des Sciences Naturelles , 44 : 223 – 270 . Jenks G. F. 1967 . The data model concept in statistical mapping . International Yearbook of Cartography , 7 : 186 – 190 . Jones P. J. S. 2006 . Collective action problems posed by no-take zones . Marine Policy , 30 : 143 – 156 . Google Scholar CrossRef Search ADS Katara I. , Silva A. 2017 . Mismatch between VMS data temporal resolution and fishing activity time scales . Fisheries Research , 188 : 1 – 5 . Google Scholar CrossRef Search ADS Kellner J. B. , Tetreault I. , Gaines S. D. , Nisbet R. M. 2007 . Fishing the line near marine reserves in single and multispecies fisheries. Ecological Applications, 17: 1039–1054. Kincaid K. , Rose G. , Devillers R. 2017 . How fisher-influenced marine closed areas contribute to ecosystem-based management: a review and performance indicator scorecard . Fish and Fisheries , 18 : 860 – 876 . Google Scholar CrossRef Search ADS Kjærsgaard J. 2007 . Multiple criteria decision making in fisheries management: experiences and implications . Food Economics - Acta Agriculturae Scandinavica, Section C , 4 : 49 – 64 . Google Scholar CrossRef Search ADS Leite M. C. F. , Gasalla M. A. 2013 . A method for assessing fishers’ ecological knowledge as a practical tool for ecosystem-based fisheries management: seeking consensus in Southeastern Brazil . Fisheries Research , 145 : 43 – 53 . Google Scholar CrossRef Search ADS Leslie H. M. , McLeod K. L. 2007 . Confronting the challenges of implementing marine ecosystem-based management . Frontiers in Ecology and the Environment , 5 : 540 – 548 . Google Scholar CrossRef Search ADS Lester S. E. , Halpern B. S. , Grorud-Colvert K. , Lubchenco J. , Ruttenberg B. I. , Gaines S. D. , Airamé S. et al. 2009 . Biological effects within no-take marine reserves: a global synthesis . Marine Ecology Progress Series , 384 : 33 – 46 . Google Scholar CrossRef Search ADS Link J. S. , Brodziak J. K. T. , Edwards S. F. , Overholtz W. J. , Mountain D. , Jossi J. W. , Smith T. D. , et al. 2002 . Marine ecosystem assessment in a fisheries management context . Canadian Journal of Fisheries and Aquatic Sciences , 59 : 1429 – 1440 . Google Scholar CrossRef Search ADS Link J. S. 2005 . Translating ecosystem indicators into decision criteria . ICES Journal of Marine Science , 62 : 569 – 576 . Google Scholar CrossRef Search ADS Link J. S. 2010 . Ecosystem-based fisheries management: Confronting tradeoffs. Cambridge University Press, New York. 221 pp. Mahe K. , Amara R. , Bryckaert T. , Kacher M. , Brylinski J. M. 2007 . Ontogenetic and spatial variation in the diet of hake (Merluccius merluccius) in the Bay of Biscay and the Celtic Sea . Journal of Marine Science , 64 : 1210 – 1219 . Malczewski J. 2000 . On the use of weighted linear combination method in GIS: common and best practice approaches . Transactions in GIS , 4 : 5 – 22 . Google Scholar CrossRef Search ADS Malczewski J. 2006 . GIS-based multicriteria decision analysis: a survey of the literature . International Journal of Geographical Information Science , 20 : 703 – 726 . Google Scholar CrossRef Search ADS Mardle S. , Pascoe S. 1999 . A review of applications of multiple-criteria decision-making techniques to fisheries . Marine Resource Economics , 14 : 41 – 63 . Google Scholar CrossRef Search ADS Massam B. H. 1988 . Multi-Criteria Decision Making (MCDM) techniques in planning . Progress in Planning , 30 : 1 – 84 . Google Scholar CrossRef Search ADS Maxwell D. L. , Stelzenmüller V. , Eastwood P. D. , Rogers S. I. 2009 . Modelling the spatial distribution of plaice (Pleuronectes platessa), sole (Solea solea) and thornback ray (Raja clavata) in UK waters for marine management and planning . Journal of Sea Research , 61 : 258 – 267 . Google Scholar CrossRef Search ADS Meaden G. , Aguilar-Manjarrez J. 2013 . Advances in geographic information systems and remote sensing for fisheries and aquaculture, FAO Fisheries And Aquaculture Technical Paper 552. Food and Agriculture Organization of the United Nations (FAO) Rome, 2013. 425 pp. Moita M. T. , Silva A. , Palma S. , Vilarinho M. G. 2010 . The coccolithophore summer-autumn assemblage in the upwelling waters of Portugal: patterns of mesoscale distribution (1985-2005) . Estuarine, Coastal and Shelf Science , 87 : 411 – 419 . Google Scholar CrossRef Search ADS Moore C. H. , Harvey E. S. , Van Niel K. P. 2009 . Spatial prediction of demersal fish distributions: enhancing our understanding of species-environment relationships . In ICES Journal of Marine Science , 66 : 2068 – 2075 . Google Scholar CrossRef Search ADS Mosadeghi R. , Warnken J. , Tomlinson R. , Mirfenderesk H. 2013 . Uncertainty analysis in the application of multi-criteria decision-making methods in Australian strategic environmental decisions . Journal of Environmental Planning and Management , 56 : 1097 – 1124 . Google Scholar CrossRef Search ADS Olson J. 2011 . Producing nature and enacting difference in ecosystem-based fisheries management: an example from the Northeastern US . Marine Policy , 35 : 528 – 535 . Google Scholar CrossRef Search ADS Pérez O. M. , Telfer T. C. , Ross L. G. 2005 . Geographical information systems-based models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary Islands . Aquaculture Research , 36 : 946 – 961 . Google Scholar CrossRef Search ADS Petitgas P. , Doray M. , Huret M. , Massé J. , Woillez M. 2014 . Modelling the variability in fish spatial distributions over time with empirical orthogonal functions: anchovy in the Bay of Biscay . ICES Journal of Marine Science , 71 : 2379 – 2389 . Google Scholar CrossRef Search ADS Phua M. H. , Minowa M. 2005 . A GIS-based multi-criteria decision making approach to forest conservation planning at a landscape scale: a case study in the Kinabalu Area, Sabah, Malaysia . Landscape and Urban Planning , 71 : 207 – 222 . Google Scholar CrossRef Search ADS Pikitch E. S. , Boersma P. D. , Boyd I. L. , Conover D. O. , Cury P. , Essington T. , Heppell S. S. et al. 2012 . Little Fish, Big Impact, pp.. Lenfest Ocean Program, Washington, DC. Polovina J. J. 1984 . Model of a coral reef ecosystem - I. The ECOPATH model and its application to French Frigate Shoals . Coral Reefs , 3 : 1 – 11 . Google Scholar CrossRef Search ADS Portman M. E. 2007 . Zoning design for cross-border marine protected areas: the Red Sea Marine Peace Park case study . Ocean and Coastal Management , 50 : 499 – 522 . Google Scholar CrossRef Search ADS Rahman M. A. , Rusteberg B. , Uddin M. S. , Lutz A. , Saada M. A. , Sauter M. 2013 . An integrated study of spatial multicriteria analysis and mathematical modelling for managed aquifer recharge site suitability mapping and site ranking at Northern Gaza coastal aquifer . Journal of Environmental Management , 124 : 25 – 39 . Google Scholar CrossRef Search ADS PubMed Ramírez-Monsalve P. , Raakjær J. , Nielsen K. N. , Santiago J. L. , Ballesteros M. , Laksá U. , Degnbol P. 2016a . Ecosystem Approach to Fisheries Management (EAFM) in the EU - Current science-policy-society interfaces and emerging requirements . Marine Policy , 66 : 83 – 92 . Google Scholar CrossRef Search ADS Ramírez-Monsalve P. , Raakjær J. , Nielsen K. N. , Laksá U. , Danielsen R. , Degnbol D. , Ballesteros M. et al. 2016b . Institutional challenges for policy-making and fisheries advice to move to a full EAFM approach within the current governance structures for marine policies . Marine Policy , 69 : 1 – 12 . Google Scholar CrossRef Search ADS Refsgaard J. C. , van der Sluijs J. P. , Højberg A. L. , Vanrolleghem P. A. 2007 . Uncertainty in the environmental modelling process - A framework and guidance . Environmental Modelling and Software , 22 : 1543 – 1556 . Google Scholar CrossRef Search ADS Rodríguez-Climent S. , Angélico M. M. , Marques V. , Oliveira P. , Wise L. , Silva A. 2017 . Essential habitat for sardine juveniles in Iberian waters . Scientia Marina , 81 : 351–000. Romero C. , Rehman T. 1987 . Natural resource management and the use of multiple criteria decision-making techniques: a review . European Review of Agricultural Economics , 14 : 61 – 89 . Google Scholar CrossRef Search ADS Ruckelshaus M. , Klinger T. , Knowlton N. , DeMaster D. P. 2008 . Marine ecosystem-based management in practice : scientific and governance challenges . BioScience, 58: 53–63. Saaty T. L. 1980 . The Analytic Hierarchy Process. McGraw-Hill, New York. 287 pp. Sahin O. , Mohamed S. 2013 . A spatial temporal decision framework for adaptation to sea level rise . Environmental Modelling and Software , 46 : 129 – 141 . Google Scholar CrossRef Search ADS Silva A. , Moreno A. , Riveiro I. , Santos B. , Pita C. , Rodrigues J. G. , Villasante S. , et al. 2015 . Research for PECH committee- Sardine fisheries: resource assessment and social and economic situation. Study, European Parliament's Committee on Fisheries. IP/B/PECH/IC/2015_133, 68 pp. Silva M. A. 1999 . Diet of common dolphins, Delphinus delphis, off the Portuguese continental coast . Journal of the Marine Biological Association of the United Kingdom , 79 : 531 – 540 . Google Scholar CrossRef Search ADS Simmonds J. , MacLennan D. 2005 . Biological Acoustics. Fisheries Acoustics. Theory and Practice: Fisheries Acoustics, 2nd edn. Blackwell Science, Oxford, 437 pp. Sobrinho-Gonçalves L. , Moita M. T. , Garrido S. , Cunha M. E. 2013 . Environmental forcing on the interactions of plankton communities across a continental shelf in the Eastern Atlantic upwelling system . Hydrobiologia , 713 : 167 – 182 . Google Scholar CrossRef Search ADS Staples D. , Brainard R. , Capezzuoli S. , Funge-smith S. , Grose C. , Heenan A. , Hermes R. et al. 2014 . Essential EAFM. Ecosystem Approach to Fisheries Management Training Course. 319 pp. Stelzenmüller V. , Maynou F. , Bernard G. , Cadiou G. , Camilleri M. , Crec’hriou R. , Criquet G. et al. 2008 . Spatial assessment of fishing effort around European marine reserves: implications for successful fisheries management . Marine Pollution Bulletin , 56 : 2018 – 2026 . Google Scholar CrossRef Search ADS PubMed Stewart R. R. , Possingham H. P. 2005 . Efficiency, costs and trade-offs in marine reserve system design . Environmental Modeling and Assessment , 10 : 203 – 213 . Google Scholar CrossRef Search ADS Tomlin C. D. 1991 . Cartographic Modelling . Geographic Information Systems , 1 : 361 – 374 . Vandeperre F. , Higgins R. M. , Sánchez-Meca J. , Maynou F. , Goñi R. , Martín-Sosa P. , Pérez-Ruzafa A. , et al. 2011 . Effects of no-take area size and age of marine protected areas on fisheries yields: a meta-analytical approach . Fish and Fisheries , 12 : 412 – 426 . Google Scholar CrossRef Search ADS Varjopuro R. , Gray T. , Hatchard J. , Rauschmayer F. , Wittmer H. 2008 . Introduction: interaction between environment and fisheries-The role of stakeholder participation . Marine Policy , 32 : 147 – 157 . Google Scholar CrossRef Search ADS Villa F. , Tunesi L. , Agardy T. 2002 . Zoning marine protected areas through spatial multiple-criteria analysis: the case of the Asinara Island National Marine Reserve of Italy . Conservation Biology , 16 : 515 – 526 . Google Scholar CrossRef Search ADS Walters C. , Christensen V. , Pauly D. 1997 . Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments . Reviews in Fish Biology and Fisheries , 7 : 139 – 172 . Google Scholar CrossRef Search ADS Walters C. , Pauly D. , Christensen V. 1999 . Ecospace: prediction of mesoscale spation patterns in trophi relatiohsips of exploited ecosystem, with empahsis on the impacts of marine protected areas . Ecosystems , 2 : 539 – 554 . Google Scholar CrossRef Search ADS Witt M. J. , Godley B. J. 2007 . A step towards seascape scale conservation: using vessel monitoring systems (VMS) to map fishing activity . PLoS One , 2 : e1111. Google Scholar CrossRef Search ADS PubMed Wooster W. S. , Bakun A. , McLain R. M. 1976 . Seasonal upwelling cycle along the Eastern boundary of the North Atlantic . Journal of Marine Research , 34 : 131 – 141 . Yeon I. , Song M. Y. , Sohn M. H. , Hwang H. J. , Im Y. J. , Kim D. H. , Park H. W. et al. 2011 . An ecosystem-based assessment of the Korean blue crab trammel net fishery in the Yellow Sea and management implications . Fisheries Research , 112 : 179 – 188 . Google Scholar CrossRef Search ADS Zwolinski J. P. , Oliveira P. B. , Quintino V. , Stratoudakis Y. 2010 . Sardine potential habitat and environmental forcing off western Portugal . ICES Journal of Marine Science , 67 : 1553 – 1564 . Google Scholar CrossRef Search ADS © International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png ICES Journal of Marine Science Oxford University Press

A GIS-based framework for addressing conflicting objectives in the context of an ecosystem approach to fisheries management—a case study of the Portuguese sardine fishery

ICES Journal of Marine Science , Volume Advance Article – Jul 31, 2018

Loading next page...
 
/lp/oxford-university-press/a-gis-based-framework-for-addressing-conflicting-objectives-in-the-3VQ5PieXmD
Publisher
ICES/CIEM
Copyright
© International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1054-3139
eISSN
1095-9289
D.O.I.
10.1093/icesjms/fsy094
Publisher site
See Article on Publisher Site

Abstract

Abstract An ecosystem approach to fisheries management (EAFM) is as a new paradigm in fisheries management. In this study, a combination of geographic information systems (GISs) and multi-criteria decision-making method (MCDM) is proposed as a framework supporting an ecosystem approach to European sardine (Sardina pilchardus, Clupeidae) fishery management in Portugal. This case study was chosen due to the recent severe decline shown by the species. To develop an EAFM for the sardine fishery, a set of criteria were defined based on literature review and expert knowledge. To address multiple conflicting objectives, namely conservation and fisheries, five scenarios were considered: (i) baseline; (ii) nurseries protection; (iii) spawning areas protection; (iv) fishery profit driven, and (v) safeguarding dependent fishing communities. Combination of GIS and MCDM methods highlighted important areas to implement spatial conservation measures for sardine. The analyses indicate that some areas are suitable for conservation in several scenarios, such as the area near Aveiro and the area near the Tejo estuary. However, conservation measures implemented in the area near Aveiro would imply higher economic trade-offs when compared with the actions applied in the region near the Tejo estuary. Results also suggested some of the conservation objectives, such as the protection of sardine eggs and juveniles, to not be compatible. The proposed framework is an important tool supporting EAFM by addressing conflicting objectives, trade-offs and identifying areas that could be considered as potential fishery closure sites or subjected to further analyses. Introduction Depletion of fish stocks worldwide (FAO, 2014) signal a failure of conventional fisheries management practices and ineffective governance (Ruckelshaus et al., 2008). Common Fisheries Policy (CFP) was settled with the aim to integrate environmental concerns and move towards a more holistic ecosystem approach to fisheries management (EAFM) (Ramírez-Monsalve et al., 2016a). Key characteristics of this approach include multispecies consideration, trophic interactions, socio-economic dimension, and impact on habitat and ecosystem (Garcia et al., 2003). There is a large body of literature presenting guidelines and frameworks aiming to facilitate EAFM operationalization (Fletcher, 2008; Link, 2010; Staples et al., 2014) and several methods have been explored. The most popular and frequently studied have been: (i) modelling approaches e.g. Ecosystem models (Polovina, 1984; Christensen and Pauly, 1992; Walters et al., 1997, 1999, Fulton et al., 2004a, 2004b); (ii) statistical methods that support Integrated Ecosystem Assessment (Link et al., 2002), (iii) spatial methods, such as the use of geographic information systems (GISs) (Olson, 2011), (iv) application of indicators (Link, 2005; Yeon et al., 2011), and finally (v) the use of socio-ecological frameworks (Leite and Gasalla, 2013). Even though, significant progress has been made towards a better operationalization of an EAFM in a recent CFP reform [regulation (EU) No. 1380/2013], several challenges, such as lack of transparent and legitimate framework to address and balance conflicting objectives and trade-offs, impede the implementation of an EAFM in EU fisheries (Ramírez-Monsalve et al., 2016b) The use of GIS, a versatile tool which can be applied at different stages of the fisheries management process, has been recommended by the Food and Agriculture Organization (FAO) as a tool for an EAFM (Carocci et al., 2009). GIS application in fisheries management has been already implemented in coastal regions around the world (Aswani and Lauer, 2006; Hall and Close, 2007; De Freitas and Tagliani, 2009). Despite its potential, the use of GIS in EAFM is still limited (Carocci et al., 2009; Meaden and Aguilar-Manjarrez, 2013). Multi-criteria decision making (MCDM) methods have the ability to assist in the management of renewable resources by taking into account conflicting objectives, therefore attaining interests within a wide range of management fields (Romero and Rehman, 1987). They have also been frequently applied to fisheries (Mardle and Pascoe, 1999; Chiou et al., 2005; Kjærsgaard, 2007; Innes and Pascoe, 2010). Nevertheless, spatial MCDM methods, that allow for the evaluation of multiple conflicting criteria to solve complex spatial problems (Malczewski, 2006), have had limited use in fisheries and coastal management studies (Hossain and Das, 2010; Dapueto et al., 2015) (Villa et al., 2002; Portman, 2007). In the EAFM context, this approach can be used to indicate suitable areas for biodiversity conservation, while minimizing the negative effect on fisheries. European sardine (Sardina pilchardus, Walbaum, 1792) represents an economically and socially significant fishery resource in Portugal and Spain (Silva et al., 2015). Sardine in the Iberian Peninsula Atlantic waters (ICES Divisions 8.c and 9.a) is considered a stock for management purposes (ICES, 2017a). In the last decade, its severe decline in the Iberian waters has been observed: biomass decreased around 80% and recruitment reached the lowest historical level in 40 years (ICES, 2017a). The majority of the Iberian sardine stock is distributed in Portuguese waters and its main recruitment and spawning areas are the northern Portugal and south of Spain (Gulf of Cadiz); (Bernal et al., 2007; Rodríguez-Climent et al., 2017). Clupeoid fishes, like sardine, are usually short-lived and the success of their recruitment is crucial to further increase the size of the population (Cole and Mcglade, 1998). These fish species usually show high variability in the recruitment process related to environmental conditions (Bakun, 1997). In this paper we describe a framework that demonstrates the utility of GIS to support fisheries management in the context of an ecosystem approach. It implements a GIS-based MCDM method combined with broad spatial analysis using the Iberian sardine as a case study. The presented framework has the capability to facilitate EAFM by identifying areas which (i) are of the highest importance for conservation, (ii) are the most crucial for fisheries, and (iii) imply trade-offs between conservation and fisheries objectives. Based on this information, the most appropriate choices for conservation were presented. Material and methods Study area Located on the western tip of Europe and with an extension of ∼52 000 km2, the Portuguese continental shelf (0–200 m; Figure 1) is in a biogeographic transition zone between temperate and subtropical waters (Briggs, 1974). Moreover, the western Portugal coast is located alongside the northernmost part of the Canary Current Upwelling System, one of the four major eastern boundary upwelling systems of the World Ocean classified as a Class I highly productive ecosystem (>300 g cm−2 year−1) (Carr, 2002). In this area, seasonal upwelling occurs during spring and summer as a result of steady northerly winds (Wooster et al., 1976; Fiúza et al., 1982). Figure 1. View largeDownload slide Map showing limits of the study area between 200 m isobath and Portuguese coastline, occurrence of sardine in Portuguese acoustic surveys between 2005 and 2010 (source: IPMA, PELAGO surveys); purse-seine fishing grounds between 2005 and 2010 (source: DGRM VMS-logbook data) and acoustic transect along which biological data were collected (IPMA, PELAGO surveys). Figure 1. View largeDownload slide Map showing limits of the study area between 200 m isobath and Portuguese coastline, occurrence of sardine in Portuguese acoustic surveys between 2005 and 2010 (source: IPMA, PELAGO surveys); purse-seine fishing grounds between 2005 and 2010 (source: DGRM VMS-logbook data) and acoustic transect along which biological data were collected (IPMA, PELAGO surveys). GIS-MCDM analysis A combination of GIS and MCDM methodology was used to identify the most suitable sites to preserve essential sardine habitats, while reducing the negative effect on fishing activities. A set of criteria representing a holistic ecosystem approach and considering sardine conservation and fishing industry, was defined. To select the most suitable option from available alternatives, MCDM technique (decision rule, algorithm), with the weighted linear combination (WLC) (weighted overlay) method, was used. Another widely used method, analytical hierarchy process (AHP) was employed to derive the weights associated with attribute map layers used in WLC method. Applied GIS-based MCDM procedure (Figure 2) consists of five steps: (i) identification of important criteria; (ii) data compilation; (iii) creation of criteria maps; (iv) production of suitability maps; and (v) assessment of suitability maps. Figure 2. View largeDownload slide Conceptual suitability model showing applied criteria and GIS-based MCDM procedure applied to select conservation areas that aim to protect sardine essential habitat and at the same time attempt to maintain the socio-economic efficiency of purse-seine fishery in coastal Portuguese waters. Figure 2. View largeDownload slide Conceptual suitability model showing applied criteria and GIS-based MCDM procedure applied to select conservation areas that aim to protect sardine essential habitat and at the same time attempt to maintain the socio-economic efficiency of purse-seine fishery in coastal Portuguese waters. Step 1—identification of criteria The first step of the suitability analysis was a selection of important criteria for sardine essential habitat and purse-seine fishery, and the development of a conceptual suitability model accordingly (Figure 2). The model consisted of 12 criteria representing three main dimensions of EAFM: biological, environmental and socio-economic (Tables 1–3). Criteria selection was based on the best available published literature and expert judgement. Table 1. Description of biological criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey a ESDU—elemental sample distance unit that corresponds to 1 nm. Table 1. Description of biological criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey Criteria Reason to include Assignment of priority values Type of objective Years Data source Juvenile sardine (TL ≤ 16 cm) biomass (tonnes/ESDUa) Sardine recruits (juveniles < 1 year) are the proxy for nursery areas which conservation is important for maintenance of healthy stock. High juvenile sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Adult sardine (TL > 16 cm) biomass (tonnes/ESDUa) High biomass of adult sardine is important for conservation because adult sardine has the maturity to spawn and might be considered important for spawning areas conservation. High adult sardine biomass—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey Sardine eggs abundance (no. /m3) Sardine eggs abundance is an indicator of sardine spawning grounds that are of high importance for conservation. High sardine eggs abundance—high conservation priority Conservation 2005–2010 IPMA, PELAGO survey; CUFES Sardine competitors Atlantic chub mackerel Atlantic mackerel Horse mackerel Mediterranean horse mackerel Blue jack mackerel Bogue European anchovy In the study, sardine competitors are considered species that are caught by the same fleet as a sardine. A higher number of competitors is an indicator of extra stress for sardine as they compete for space, food etc. Therefore, the areas that have a high number of competitors have a priority for conservation. High sardine competitors’ abundance—high conservation priority Conservation 2007–2010 2005–2010 2005–2010 2007–2010 2007–2010 2007–2010 2005–2010 IPMA, PELAGO survey IPMA, Demersal survey IPMA, PELAGO survey Sardine predators European hake Squid Sardine predators are indicators of natural mortality of sardine. The higher number of predators means higher stress (natural withdrawal of sardine from the environment) and therefore, intensified necessity for conservation. Moreover, these areas indicate locations important to protect in order to ensure ecological support for other species and whole marine ecosystem (Pikitch et al., 2012). High sardine predators’ abundance—high conservation priority Conservation 2005–2010 2005–2010 IPMA, Demersal survey a ESDU—elemental sample distance unit that corresponds to 1 nm. Table 2. Description of environmental criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Table 2. Description of environmental criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Criteria Reason to include Assignment of priority values Type of objective Years Data source Suitable juvenile sardine (TL ≤ 16 cm) habitat Areas that have a high probability of juvenile sardine occurrence are of high priority for conservation as they are indicators of sardine nurseries areas. Conservation West (37–42°N; 9–10°W) Temperature (°C) Juvenile sardine abundance showed a doomed shaped response with temperature, displaying a peak at around 14°C (Rodríguez-Climent et al., 2017). 13.5–14.5°C—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS aqua satellite sensor Salinity (PSU) Juvenile sardine abundance showed a bimodal shape response with salinity, displaying one peak at ∼35.3 PSU and another one at ∼33.7 PSU (Rodríguez-Climent et al., 2017). 33.5–34 PSU and 35–35.5 PSU—high probability of juvenile occurrence—high conservation priority. 2005–2010 IPMA, in situ; PELAGO survey; Sensor associated to CUFES Latitude Juvenile sardine abundance showed a bimodal shape response with latitude, displaying one peak at ∼38.6°N and another one at ∼40.3°N (Rodríguez-Climent et al., 2017). 38.5–39°N and 40.0–40.6°N—high probability of juvenile occurrence—high conservation priority. Depth (m) Juvenile sardine depth tolerance is between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. 2005–2010 Portuguese Hydrographic Institute South (35–37°N; 7–9°W) Temperature (°C) Juvenile sardine abundance is positively correlated with temperature (Rodríguez-Climent et al., 2017). High temperature—high probability of juvenile occurrence—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) The highest abundance of juvenile sardine occurs in the areas with average chlorophyll concentration (Rodríguez-Climent et al., 2017). Average chlorophyll—high probability of juvenile occurrence—high conservation priority. 2005–2010 SeaWIFS satellite sensor Zooplankton (ml/10m3) Juvenile sardine abundance is inversely correlated with zooplankton abundance (Rodríguez-Climent et al., 2017). Low zooplankton concentration – high probability of juvenile occurrence –high conservation priority 2005–2010 IPMA, PELAGO survey; CUFES Depth (m) Juvenile sardine occurs with the highest probability between the coastline and a depth of 100 m. <100 m—high probability of juvenile occurrence—high conservation priority. Portuguese Hydrographic Institute Suitable adult sardine (TL > 16 cm) habitat Areas that have a high probability of adult sardine occurrence are of high importance for conservation as adult sardine has the maturity to spawn and might be considered as important for spawning areas conservation. Conservation Temperature (°C) Adult sardine abundance is inversely correlated with temperature values (Zwolinski et al., 2010). Low temperature-high adult sardine biomass—high conservation priority. 2005–2010 MODIS satellite sensor Chlorophyll a (mg/m3) Adult sardine abundance is positively correlated with chlorophyll a concentration (Zwolinski et al., 2010). High chlorophyll—high adult sardine biomass—high conservation priority. 2005–2010 SeaWIFS satellite sensor Salinity (PSU) Adult sardine abundance is inversely correlated with salinity values (Zwolinski et al., 2010) that indicate regions of fresh water influence. Low salinity—high adult sardine biomass—high conservation priority. 2005–2010 IPMA PELAGO survey, sensor associated to CUFES Depth (m) Adult sardine occurs with the highest probability between the coastline and a depth of 100 m (Zwolinski et al., 2010). < 100m – high probability of sardine occurrence – high conservation priority. Portuguese Hydrographic Institute Suitable sardine eggs habitat Areas that have a high probability of sardine eggs occurrence are of high importance for conservation as they are the areas where sardine spawn. Conservation Temperature (°C) Sardine spawns in the temperature range between 12 and 17°C, while the temperature preference range is between 13.5 and 15°C (Bernal et al., 2007). 13.5–15°C—high sardine eggs occurrence probability—high conservation priority; <12 and >17°C—low probability of eggs occurrence—low conservation priority. 2005–2010 MODIS satellite sensor Depth (m) Sardine depth tolerance is between the coastline and a depth of 200 m, near the shelf edge, while preferences are found between depths around 10 m to around 150 m (Bernal et al., 2007). 10–150 m—high sardine eggs occurrence probability—high conservation priority; <10 and > 150 m—low probability of eggs occurrence—low conservation priority. Portuguese Hydrographic Institute Table 3. Description of socio-economic criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Table 3. Description of socio-economic criteria used in the GIS multi-criteria decision making method. Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Criteria Reason to include Assignment of priority values Type of objective Years Data source Sardine CPUE High sardine LPUE are attributed to the high socio-economic importance; therefore, they have low suitability to conservation. High sardine LPUE—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Sardine income (euro/kg) High sardine income has big socio-economic importance; because it secures fishers well-being. High sardine income—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Purse-seine fishery dependency from sardine High sardine catch relative to total catch is an indicator of dependency of purse-seine industry from sardine. High ratio of sardine catches to total catch—high fishery priority Fishery 2005 2006–2010 DGRM, landings data DGRM, VMS logbook data Other fisheries Fleets other than purse seine (e.g. trawl) were included in the analysis as they are of importance for socio-economic efficiency of fishing industry and might be affected by conservation measures. High value for other fisheries criteria—high fishery priority Fishery 2005–2010 STECF Step 2—data compilation The second step of the analysis was the compilation of spatial data for each criterion. The spatial data compiled covered a period of 6 years (2005–2010; with few exceptions listed in Tables 1–3). This time scale represents a period with contrasting sardine abundance, high numbers in the years 2005–2007 and low numbers in the years 2008–2010 (ICES, 2017a). Data regarding adult [Total length (TL) > 16 cm] and juvenile (TL ≤ 16 cm) sardine biomass, sardine eggs abundance, and zooplankton concentration were obtained from the acoustic surveys (PELAGO survey series) performed annually by the Portuguese Institute of Sea and Atmosphere (IPMA) in spring (April–May). Surveys were carried out onboard the R/V Noruega along predefined parallel transects perpendicular to the coastline covering the whole platform (0- to 200-m depth) with 8 nautical miles (nm) inter-transect distance (Figure 1). Acoustic fish data was obtained following the methodology described in ICES (2016) by performing echo integration along 1-nm Elemental sampling distance unit (ESDU), used as distance measure in acoustic surveys, and supported by fish samples collection through pelagic and bottom trawling. Estimates of sardine biomass were calculated following the procedures described in Simmonds and MacLennan (2005). For more details about acoustic surveys and sardine biomass estimates refer to publications by ICES (2016) and Rodríguez-Climent et al. (2017). Sardine eggs and zooplankton data were collected along the acoustic transects every 3 nm at the surface (∼3 m), using a Continuous Underway Fish Egg Sampler (CUFES, mesh 335 μm) (Checkley et al., 1997). Sardine eggs abundance (no./m3) was calculated based on the eggs data collected per unit of time and the pump flow rate. Zooplankton biomass expressed as volume of zooplankton (ml/10 m3) was estimated by volumetric method (displacement volume; Ikeda and Omori, 1992). Sardine competitors’ species abundance data: atlantic chub mackerel (Scomber colias, Gmelin 1789), bogue (Boops boops, Linnaeus 1758), anchovy (Engraulis encrasicolus, Linnaeus 1758), mediterranean horse mackerel (Trachurus mediterraneus, Steindachner, 1868) and blue jack mackerel (Trachurus picturatus, Bowdich, 1825), were assessed by acoustic data (Nautical Area Scattering Coefficient, NASC—SA m2/nm2) derived from the PELAGO surveys. The remaining competitors: horse mackerel (Trachurus trachurus, Linnaeus 1758) and atlantic mackerel (Scomber scombrus, Linnaeus 1758) and sardine predators, hake (Merluccius merluccius, Linnaeus 1758) and squid (Loligo spp. Lamarck, 1798) data were based on abundance data (kg/h) collected during demersal surveys (2005–2010). Those surveys were carried out in autumn (September–October) onboard the R/V Noruega at predefined stations in the Portuguese Exclusive Economic Zone at depths varying from 20- to 500-m depth. For details about abundance estimation based on demersal surveys refer to International bottom trawl surveys protocol published by ICES (2017b). Vessel monitoring system (VMS) and logbook data from purse-seine vessels ≥ 12 m (EU Council, 2009) for the years 2006–2010, obtained from the DGRM, were integrated to produce maps of fishing activity. Purse-seine catch distribution maps were developed based on analysing vessel speed patterns from VMS data and linking them to logbook data to identify fishing activity according to Katara and Silva (2017). We reported landings as the sum of tonnes of each species landed per year and per grid cell. VMS and logbook data for the year 2005 were not used because their spatial coverage was not sufficient and representative enough. Instead, the official landing data per main fishing port, registered and obtained from the DGRM were used for this year. These data were spatially attributed to fishing grounds previously defined according to spatial patterns recorded in 2006–2010 VMS-logbook data (Supplementary Material S1 and Supplementary Figure S1). For the criteria “other fisheries” the fleets using the following gear were considered: bottom trawl, longline, gill, trammel, and pots. Spatial effort (hours) and landings (tonnes) data (2005–2010) at ICES statistical rectangles level were obtained from the Scientific Technical and Economic Committee for Fisheries (STECF) (annex: WW; country: PRT; https://stecf.jrc.ec.europa.eu, 20 May 2017). Sea surface temperature (°C) and chlorophyll a concentration data (mg/m3) were obtained from MODIS aqua (L3; 4 μm) and SeaWIFS satellite sensor respectively (sources: www.opendap.jpl.nasa.gov/opendap/; www.oceancolour.org/portal/, 10 May 2017) and represented the period of the pelagic surveys: March–May Salinity was collected in situ during the IPMA acoustic surveys (PELAGO, 2005–2010). Depth (m) data were obtained from the Portuguese Hydrographic Institute website (www.hidrografico.pt/download-gratuito.php, 25 May 2017). Step 3—creation of criteria maps To create surface continuous layers, a similar approach to the one performed in Petitgas et al. (2014) allowing for a reasonable compromise between smoothing and maintaining details, was taken. Data collected along acoustic transects and at the demersal survey’s sampling stations were averaged on a common regular grid (8 × 8 km) using ArcMap neighbourhood point statistic function that allows the creation of cell-based rasters with a smoothing effect. Satellite data were converted to the 8 × 8 km grid to match the resolution of the remaining layers. All analyses were performed using ESRI ArcMap software (version: 10.4.1). Socio-economic, environmental, and some biological criteria—competitors and predators—required processing of initial data in order to be represented as meaningful maps useful for this study. Generation of socio-economic criteria The following socio-economic criteria calculated yearly were considered in this study (i) sardine catch per unit of effort (CPUE); (ii) sardine income; (iii) purse-seine fishery dependency from sardine; and (iv) other fisheries. Sardine CPUE was calculated by considering the effort to be the duration of a fishing trip times the number of sets per trip (where sardine was caught) except for the year 2005 where the effort was considered to be the number of fishing trips where sardine was caught. For the other fleets, effort was considered to be the number of fishing hours. Discards were negligible therefore we considered that landings per unit effort (LPUE) were equal to CPUE. Income was calculated by multiplying landings of the species of interest by its price (euro per kg). Prices used were those indicated in the Portuguese National Statistical Institute (INE) annual statistic report (INE 2006, 2007, 2008, 2009, 2010, 2011). Purse-seine fishery dependency from sardine was calculated by dividing sardine landings by total landings. The criterium ‘other fisheries’ is a combination of CPUE, landings and income for the respective fleet, i.e. bottom trawl, longline, gill, trammel, and pots. In order to combine all these variables, the spatial layers representing these variables, were first reclassified using the Jenks natural breaks method (Jenks, 1967) and then averaged using GIS raster algebra function. Jenks natural breaks method is the default classification method in ArcMap which identifies breakpoints between classes using a statistical formula (Jenks optimization) that minimizes the sum of the variance within each of the classes finding groups and patterns inherent to the data. Generation of suitable habitat criteria Maps representing suitable habitat for juvenile sardine, adult sardine and sardine eggs were produced using environmental variables that affect sardine’s distribution according to published literature. Environmental variables were reclassified in accordance with Rodríguez-Climent et al., (2017) for sardine juveniles, Zwolinski et al. (2010) for adults and Bernal et al. (2007) for eggs. The reclassified environmental maps were then combined by summing the reclassified layers, corresponding to each variable with raster algebra (raster calculator ArcGis function). Generation of biotic interactions criteria In this analysis, biotic interactions were represented by sardine competitors and predators. Areas with high number of competitors or predators were treated as a priority for conservation, as they constitute natural stress for sardine. Moreover, these areas were considered as priority because in these locations sardine seems to play a fundamental role to ecologically support other species and the whole marine ecosystem—as forage species (Pikitch et al., 2012). Species commonly caught along with sardine by the purse-seine fleet were considered as sardine competitors namely: S. colias, S. scombrus, T. trachurus, B. boops, E. encrasicolus, T. mediterraneus and T. picturatus. Species like M. merluccius and Loligo spp. were considered as common sardine predators (Coelho et al., 1997; Cabral and Murta, 2002; Mahe et al., 2007). Other important sardine predators, such as the common dolphin (Delphinus delphis L.1758) (Silva, 1999), were not considered in this study as a result of data availability limitations. Note that some of the species considered as sardine competitors (e.g. S. colias and B. boops) in this study are also predators for sardine eggs (Garrido et al., 2015). The overlap found between sardine and their competitors/predators, will be referred as spatial overlap intensity index (Oi) and was estimated using the Jaccard index of similarity (Jaccard, 1908) modified as follows: Oi=∑i=1n (PIL & SPiPIL | SPi × SPi )×Afi (1) where Oi is overlap between sardine and other species considered, PIL is sardine biomass; SPiis other species abundance; n is the number of other species considered and Afi is the other species abundance factor. Other species either refers to competitors in Competitors-sardine overlap or predators in the Predators-sardine overlap. The Jaccard index is one of the oldest and most widely used similarity indices for assessing compositional similarity of assemblages. It is based on the presence-absence records of species in paired assemblages (Chao et al., 2005). In order to reflect species abundance in the output of the binary index, the index was modified by adding multiplication by two variables: First, variable SPi that corresponds to species abundance in the used dataset and then by variable Af that is general species abundance factor. Available competitors’ species data had various units as they were obtained from various surveys. Data regarding sardine predators and competitors, atlantic mackerel and horse mackerel, were available from demersal surveys, while the remaining competitors’ data was obtained from the acoustic surveys. This fact made it mathematically impossible to apply Equation (1). Therefore in order to address different units in data, each species-sardine overlap raster was reclassified into 10 classes using Jenks natural breaks. Nevertheless, the outputs of the procedure of reclassification made it impossible to differentiate between very abundant species, such as horse mackerel, from low abundant species, such as bogue. Therefore, to address the importance of each species abundance relatively to each other, the reclassified rasters were multiplied by an abundance factor Af established for this purpose. This subjective abundance factor was attributed to each competitor species based on its 6-years abundance estimates from bottom Portuguese trawl (ICES data 2009–2014) and subjective classification established by author judgement, that aimed to highlight the differences between species abundance (see Table 4 for more details). After multiplications by SPi variable and Af factor, raster layers representing competitors and predators were summed. Table 4. Abundance factor assigned to each competitor species established based on its abundance in bottom Portuguese trawl data (source: ICES, 2009–2014). Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Table 4. Abundance factor assigned to each competitor species established based on its abundance in bottom Portuguese trawl data (source: ICES, 2009–2014). Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Species Total abundance in 2009–2014 (×1000 no. of species in trawl) Abundance factor T. trachurus 396.50 10 S. scombrus 168.90 8 T. picturatus 111.45 6 E. encrasicolus 65.91 4 S. colias 19.39 2 B. boops 15.14 2 T. mediterraneus 0.12 1 Step 4—production of suitability maps The final suitability maps for sardine conservation without compromising fisheries activities were produced using the WLC method. This method is one of the most widely used GIS-based decision-making methods for land suitability analysis (Hopkins, 1977; Malczewski, 2000). It was chosen because it is accessible, intuitively appealing to decision makers (Massam, 1988) and easy to implement in GIS (Tomlin, 1991). The WLC method involves the standardization of the attribute maps, assigning the weights of relative importance and combining the weighted criteria to obtain an overall suitability score. Therefore, all criterion maps were standardized, weighted and combined to produce the final suitability map as follows; Standardization Criteria were standardized to a uniform scale with values ranging from one to five (1: no suitability; 5 high suitability for conservation; Tables 1–3). The applied classification method was Jenks natural breaks (Jenks, 1967). Scenario definition The analysis was performed for five scenarios that varied depending on the priority objectives. The first scenario balanced conservation and fisheries objectives. The second and third scenario focussed on conservation objectives, while the fourth and fifth scenario focussed on fisheries objectives. Scenario 1, the baseline scenario, considered all criteria to have the same importance. Scenario 2, nurseries protection, aimed to protect sardine nursery areas. This scenario applied increased weight for the following criteria: juvenile biomass and suitable juvenile habitat. Scenario 3, spawning areas protection, aimed to protect sardine spawning grounds. It applied increased weight for criteria: sardine eggs abundance, adult sardine biomass (considered potential spawners) and suitable sardine eggs and adult sardine habitat. Biological criteria were obtained from observed values, while habitat criteria were output of modelling studies (for details see subsection Generation of suitable habitat criteria). Therefore, as a result of a higher uncertainty of the latter, they were assigned lower priority compared with the biological criteria. Scenario 4, fishery profit driven, aimed to protect sardine fishery revenue. Here increased weight was applied to sardine income criteria. Scenario 5, safeguarding dependent fishing communities, aimed to exclude from protection areas where fishers are highly dependent on sardine catches. It applied increased weight to areas were purse-seine fishery is highly dependent on sardine. Weight assignment The AHP method (Saaty, 1980) was used to assign weights to criteria, which differ depending on the applied scenario. This mathematical method is the most frequently used in spatial MCDM studies (Phua and Minowa, 2005; Bottero et al., 2013; Rahman et al., 2013) and it establishes weights among criteria by making a series of judgements based on pairwise comparisons of the criteria. The relative importance of each criterion is measured according to a numerical scale from one to nine (Table 5). In this study, weights were chosen to represent potential differences in the probable stakeholders’ objectives. Even though, stakeholders’ weren’t directly engaged, as it was out of the scope of the study, the framework design and a choice of methods ensure their inclusion when necessary. Once a pairwise comparison matrix was built, each criterion weight was calculated by computing the eigenvector of each column. Second, the matrix components were normalized by averaging the values across the rows (Table 6). Table 5. Scale for AHP pairwise comparisons and description how scales were attributed to criteria in each scenario. Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Table 5. Scale for AHP pairwise comparisons and description how scales were attributed to criteria in each scenario. Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Intensity of importance Description Scenarios Importance Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities 1 Equal All criteria Remaining criteria Remaining criteria Remaining criteria Remaining criteria 3 Moderate Suitable juvenile sardine habitat Suitable eggs and adult sardine habitat 5 Essential Juvenile sardine biomass Adult sardine biomass and sardine eggs abundance Sardine income Purse-seine fishery dependency from sardine Table 6. Weight values for each criterion in each scenario derived by the AHP method. Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 The bold text indicates an increased weight applied to the criterion that was considered the most important in the scenario. Table 6. Weight values for each criterion in each scenario derived by the AHP method. Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 Criteria Scenarios Baseline Nurseries protection Spawning areas protection Fishery profit driven Safeguarding dependent fishing communities Biological Juvenile sardine biomass 0.083 0.401 0.027 0.05 0.05 Adult sardine biomass 0.083 0.041 0.279 0.05 0.05 Sardine eggs abundance 0.083 0.041 0.279 0.05 0.05 Sardine competitors 0.083 0.041 0.027 0.05 0.05 Sardine predators 0.083 0.041 0.027 0.05 0.05 Habitat Suitable juvenile sardine habitat 0.083 0.191 0.027 0.05 0.05 Suitable adult sardine habitat 0.083 0.041 0.115 0.05 0.05 Suitable sardine eggs habitat 0.083 0.041 0.115 0.05 0.05 Socio-economic Sardine CPUE 0.083 0.041 0.027 0.05 0.05 Sardine income 0.083 0.041 0.027 0.45 0.05 Purse-seine fishery dependency from sardine 0.083 0.041 0.027 0.05 0.45 Other fisheries 0.083 0.041 0.027 0.05 0.05 The bold text indicates an increased weight applied to the criterion that was considered the most important in the scenario. Integration of criteria maps Each criterion map (Figure 3) is composed of i pixels. Each i pixel is characterized by attributes which describe location (coordinate data) and value xij, representing standardized score of ith pixel with respect to the jth criterion (attribute value is associated with the location). The suitability of each ith pixel is evaluated using the following equation: Si=∑j=1nwjxij (2) where S is suitability of the ith pixel in the final map; n is a number of criteria (n = 12); wj is a weight of jth criterion and xij is the standarized score for the ith pixel with respect to the jth criterion. Figure 3. View largeDownload slide Example of the WLC method of criteria raster maps (redrawn from Pérez et al., 2005; Dapueto et al., 2015). The suitability map is a result of the linear combination of j criteria (in the figure j refers to the A and B criteria maps, with weights 0.65 and 0.35, respectively). Each pixel i on criteria map j is represented by a standardized value xij. Figure 3. View largeDownload slide Example of the WLC method of criteria raster maps (redrawn from Pérez et al., 2005; Dapueto et al., 2015). The suitability map is a result of the linear combination of j criteria (in the figure j refers to the A and B criteria maps, with weights 0.65 and 0.35, respectively). Each pixel i on criteria map j is represented by a standardized value xij. Firstly, in order to identify the conflicting areas, biological and environmental criteria and socio-economic criteria were combined separately. Then combination of all criteria was performed for five different scenarios. Model verification—sensitivity analysis Sensitivity analysis (SA) with a one-at-a-time method (Daniel, 1973, Chen et al. 2010) was performed to evaluate how weights influenced criteria and the spatial patterns on the suitability maps. This method introduces changes in one factor (weight) at a time, while all the other factors remain fixed, and detects how this change might influence the output. A range of percent change of ±100% and increment of percent range of ±1% were applied to the complete set of criteria used in this study. Therefore, the SA consisted of 200 simulation runs for each criterion. It was performed by using Model Builder module of ArcMap that enables semi-automatic processing of a large data set and options. Step 5—assessment of suitability maps For each scenario, areas with high and very high suitability conservation score (4 and 5, respectively) were selected (Supplementary Figure S3) in order to be analysed in terms of costs to fishing industry and benefits for sardine conservation (Table 7). Among them, only areas with at least two grid cells (minimum of 128 km2) were considered. As costs the following aspects were considered: (i) loss in sardine fishery profit and (ii) loss in profit from the other fisheries. As benefits the protection of sardine was considered for: (i) juvenile; (ii) spawning areas; and (iii) adults. Table 7. Summary of potential costs and benefits estimated for each selected area in the five tested scenarios. Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Data are shown as yearly summary in terms of yearly loss in fisheries income and amount of protected sardine biomass and eggs. a Average refers to eggs abundance column. Table 7. Summary of potential costs and benefits estimated for each selected area in the five tested scenarios. Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Potential costs for industry Potential benefits for sardine stock Affected income Protected No. area Area name Area Purse seine Other fisheries Juvenile sardine biomass Sardine eggs abundance Adult sardine biomass Scenario (km2) (×1000 euro/year) % (×1000 euro/year) % (tonnes) % (No. /m3/ grid cell) Deviation from average (%) (tonnes) % Baseline 1 North 768 65.4 6.1 3637 1.0 4456 31.5 13 −31 9719 27.8 2 Nazare 128 5.5 0.51 2666 0.74 184 1.30 8.4 −54.6 839 2.40 3 Ericeira 448 35.3 3.32 8380 2.32 757 5.35 18.0 −2.8 1571 4.50 4 Tejo 448 14.2 1.33 9702 2.69 1674 11.8 41.5 124.6 1563 4.48 5 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 6 South 1088 26.5 2.49 10321 2.86 47 0.34 31.1 67.9 2440 7.0 Sum/Averagea 3200 156.8 15 35781 10 7239 51 27 47 17500 50 Nurseries protection 1 North 1664 127.3 12 14785 4 8732 61.7 10 −44.7 7619 21.8 2 Nazare 192 7.5 0.70 4000 1.11 487 3.44 3.57 −80.7 745 2.13 3 Ericeira 320 25.0 2.34 5608 1.55 455 3.22 22.6 22.0 1331 3.81 4 Tejo 320 11.3 1.07 6930 1.92 2362 16.7 12.7 −31.2 982 2.81 Sum/Averagea 2496 171.1 16 31323 9 12036 85 12 -34 10676 31 Spawning areas protection 1 Aveiro 1600 120.2 11.3 10477 2.90 4334 30.6 26 43 10022 28.7 2 Ericeira 960 62.0 5.83 16495 4.57 766 5.41 33 81 6109 17.5 3 Tejo 128 3.0 0.28 2772 0.77 209 1.48 13.8 −25.4 1094 3.13 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1347 3.86 5 South 384 8.5 0.80 1464.01 0.41 18.09 0.13 18.2 98.4 1037 3.0 Sum/Averagea 3392 203.6 19.1 32282 8.9 5448.8 38.5 28.6 74.6 19608 56.1 Fishery profit driven 1 North 320 10.9 1.02 1054 0.29 102 0.72 1.85 −90.0 4677 13.39 2 Nazare 192 4.8 0.45 4000 1.11 561 3.96 6.1 −66.9 312 0.89 3 Tejo 576 14.0 1.32 11389 3.15 1674 11.8 32 76 1600 4.58 4 Setubal 320 10.0 0.94 1074 0.30 121 0.85 51.2 176.9 1367 3.91 5 South 3328 74.7 7.0 41789 11.6 100.5 0.7 28.0 51 6620 19.0 Sum/Averagea 4736 114.4 11 59306 16 2559 18 24 29 14576 42 Safeguarding dependent fishing communities 1 Nazare 128 14.0 1.31 7999 2.21 568 4.01 5.5 −70.4 312 0.89 2 Ericeira 704 56.1 5.27 16100 4.46 810.65 5.73 21.3 15.1 2643 7.6 3 Tejo 576 17.9 1.69 12775 3.54 1674 11.8 28.4 53.6 1600 4.58 4 Setubal 384 12.8 1.20 1289 0.36 121 0.85 41.5 124.6 1367 3.91 5 South 3328 85.1 8.00 33253 9.21 66.03 0.47 27 43 6006 17.2 Sum/Averagea 5120 185.9 17 71416 20 3240 22.9 25 33.3 11 929 34.2 Data are shown as yearly summary in terms of yearly loss in fisheries income and amount of protected sardine biomass and eggs. a Average refers to eggs abundance column. Results Sensitivity analysis The SA showed that all criteria have a relatively similar sensitivity to change in weights. Effects were first observed after ∼ ±20% change from the baseline. Further changes were observed after each additional ±10 or ±5% changes in the input. Changes higher than ±80% from a base did not cause any further changes in the output. This pattern was similar for most of the criteria (For details refer to Supplementary Figure S2). Sardine conservation and socio-economic related criteria Areas with the highest scores for the biological and environmental criteria occurred in the following areas: south of Aveiro (Area 1), near Figueira de Foz (Area 2), Ericeira (Area 3), near the Tejo estuary, between Cabo Raso and Cabo Espichel (Area 4) and near Setúbal (Area 5) (Figure 4a). Area 1 had high scores for juvenile, adult sardine biomass and sardine eggs abundance (Figure 4b–d). Area 2, had a high level of juvenile biomass whereas Area 3, had medium levels of adult biomass, predators, and competitors (Figure 4b, c, e, and f). Also, Area 4 had high levels of competitors and predators but here high, not adult, but juvenile sardine biomass obtained higher scores (Figure 4b, c, and e). Finally, Area 5 had high competitors levels and moderate eggs abundance (Figure 4d and e). All areas had medium to high levels of suitable juvenile, adult and eggs habitat (Figure 4g–i) with the exception of Area 5, which showed low levels of suitable adult sardine habitat (Figure 4h). Figure 4. View largeDownload slide Map showing all biological and environmental criteria combined. Areas that obtained their highest scores were highlighted (a); Maps showing means of each biological and environmental criterion separately (b–i). Figure 4. View largeDownload slide Map showing all biological and environmental criteria combined. Areas that obtained their highest scores were highlighted (a); Maps showing means of each biological and environmental criterion separately (b–i). On the other hand, if we considered the socio-economic criteria (Figure 5a), areas 1–3 had high levels of socio-economic features (Figure 5a). Therefore, those areas were identified as conflicting areas. On the other hand, areas 4 and 5 (Figure 4a) had a low level of almost all socio-economic criteria (Figure 5a–d) except for other fisheries feature that showed medium levels (Figure 5e). Figure 5. View largeDownload slide Map showing all socio-economic criteria combined. Areas that obtained the highest scores were highlighted (a); Maps showing means of each socio-economic criterion separately (b–h). Figure 5. View largeDownload slide Map showing all socio-economic criteria combined. Areas that obtained the highest scores were highlighted (a); Maps showing means of each socio-economic criterion separately (b–h). Overlay of all criteria for various scenarios The suitability for sardine conservation varied across scenarios (Figure 6). Figure 6. View largeDownload slide Suitability classification for the Iberian sardine stock conservation areas for the different scenarios (a–e). Figure 6. View largeDownload slide Suitability classification for the Iberian sardine stock conservation areas for the different scenarios (a–e). The baseline scenario had the highest number of areas (min 128 km2—two pixels) with moderate suitability for conservation (Figure 6a—yellow grid cells) and small and fragmented areas of high or very high suitability for conservation (Figure 6a—orange and red grid cells, Table 7). Those areas were situated nearby Aveiro and Nazaré and the larger ones near Ericeira, Tejo, and Setúbal. In the south of Portugal, there were also a few regions with high suitability for conservation (Figure 6a). In the remaining scenarios, the high and very highly suitable areas were more aggregated and constrained to specific regions. In the nurseries protection scenario, the main suitable areas were situated near Aveiro, Ericeira, and Tejo (Figure 6b). Similarly, in the spawning areas protection scenario, these areas were also situated near Aveiro and Ericeira and additionally near Setúbal (Figure 6c). In the scenarios that focussed on socio-economic features, the suitable areas for conservation were mainly situated in the area near Tejo, Setúbal for fishery profit driven scenario and near Ericeria, Tejo and Setúbal for safeguarding dependent fishing communities’ scenario. Moreover, for both scenarios, the areas situated in the south of Portugal were identified as suitable (Figure 6d and e). Assessment of selected areas—cost-benefit analysis Cost and benefits in scenarios varied depending on selected areas and prioritized objectives (Table 7). If considered scenarios and areas they encompass as total, it is difficult to select the scenario that best balanced conflicting objectives. For example, nurseries protection scenario provided protection for the highest quantity of juvenile 85% and adult sardine 29%. However, this conservation benefits were associated with the high cost in fisheries income that was estimated for 16 and 9% of total income for purse seine and other fisheries. On the other hand, baseline scenario provided protection for the second highest quantity of juvenile (51%) and adult (28%) sardine. However, in this scenario, the impact on the fisheries income was comparable to the impact observed in nurseries protection scenario and was estimated for 15 and 10% for purse seine and other fisheries, respectively. The analysis of the main areas within scenarios individually makes the interpretation of the compromise between fisheries and conservation easier. For example, areas situated in the north were the most suitable for juvenile and adult sardine protection. Their highest numbers were observed in: nurseries protection, baseline, and spawning areas protection scenarios. However, protection of these areas would be associated with the high cost for fishery that vary from 6% in baseline scenario and about 12% of affected income, in nurseries protection and spawning areas protection scenarios. Moreover, protection of north areas would not benefit spawning areas that are mainly situated in the area near Setúbal. On the other hand, protection of the area near Setúbal, would benefit considerably protection of sardine eggs (areas with high eggs density—180% higher than the observed mean), but would have negligible effect on juvenile (<1% of protected juvenile biomass) and adult (∼ 4% of protected adult biomass) sardine protection. This area would also have minimal impact on fisheries, affecting only 1% of purse seine and <0.5% of other fisheries income. On the other hand, the area near Tejo, excluding spawning areas protection scenario, was the area that protected the second highest biomass of juvenile sardine (between 12 and 17% of total biomass) and between 3 and 5% of total adult sardine. It also had relatively low impact on the fisheries income (1–2% on purse-seine fishery and 2–4% on other fisheries). Therefore, when comparing cost and benefits of protecting the Tejo area to the North area, it can be stated that with the first a higher percentage of juvenile sardine would be protected, with a lower economic impact on the purse-seine fishery. Discussion This study describes a framework that combines GIS and MCDM methodology to advance EAFM. The most important aspect of this approach is its ability to address multiple conflicting objectives and at the same time integrate and synthesize large amounts of information relevant to EAFM into a single analysis. This is essential in EAFM (Garcia and Cochrane, 2005) and has been highlighted to have the potential to significantly facilitate the management process (Leslie and McLeod, 2007). Nowadays, GIS and spatial analysis are frequently applied in fisheries management (Witt and Godley, 2007; Maxwell et al., 2009; Moore et al., 2009) but there is still a lack of a more direct, integrative application of GIS in EAFM, as the one presented in this study. Carocci et al. (2009) stressed the importance of GIS application in EAFM and suggest that it can be especially useful as a first step of a decision support framework for an EAFM, namely in issue identification. Therefore, GIS can be useful in identification of areas which: (i) require conservation, (ii) are particularly important to fishing industry, and/or (iii) are in conflict depending on stakeholders’ objectives. Moreover, the adoption and visualization of alternative scenarios and assessment of their trade-offs is particularly important in the management processes of any complex system (Costanza, 1998). The approach described in this study deal with both these tasks. Adoption of five exploratory scenarios enabled us to identify (i) important areas for specific objectives; (ii) conflicting areas between fisheries and conservation objectives; and (iii) suitable areas for sardine conservation. The area near Aveiro was identified as suitable for both nurseries and spawning protection scenarios but was also identified as one of the most conflicting due to its crucial role in fisheries. In this area, the income from sardine fishery is high as well as the dependence of purse-seine fishery on sardine. Thus, a permanent closure of fishing in this area seems unlikely, as would not be economically and socially accepted, even though the conservation benefits might be high. To rebuild the sardine stock, temporary closures during spawning season and in areas with high juvenile abundance might be an important management solution. In this study, the spatial analysis repeatedly highlighted the area near Tejo estuary to be appropriate for conservation. This area is an important spot for sardine recruitment, but not as important for the fishing sector. Its closure has the potential to protect between 12 and 17% of total mean juvenile biomass (observed between 2005 and 2010) barely compromising economic objectives, as the affected purse-seine income was estimated to be 1% of the total income. The design of marine reserves is a process in which some economic losses are inevitable therefore the analysis and discussion of trade-offs is very important (Stewart and Possingham, 2005). Exploratory scenarios such as the ones presented in this study, can make a valuable contribution to the marine reserve design process through the indication of the magnitude of trade-offs assumed. In our case study, the most important trade-offs were between protecting juvenile sardine and minimizing the negative economic impacts. The analysis also revealed the two conservation objectives, eggs abundance, and juvenile biomass, to be not totally compatible. Thus, the protection of one of them does not imply the protection of the other, suggesting the difficulty of setting one area able to protect sardine in all its vital life-stages (egg, juvenile, and adult). The application of spatial closures as a conservation measure in fisheries is difficult (Jones, 2006) and if not properly performed might end in frustration and failure (Fiske, 1992). The effectiveness of such closures have not always been successful (Abbott and Haynie, 2012). There are some examples where the implementation of an MPA resulted in higher aggregation of fishing vessels around the protected areas, a phenomenon called “fishing the line” (Kellner et al., 2007; Stelzenmüller et al., 2008). On the other hand, there are also numerous success stories of spatial closures applied as management measure (Halpern and Warner, 2002). The potential benefits include increased species richness, biomass and size of organisms within the reserve (Lester et al., 2009). Another important enquiry when considering spatial closures, is the extent to which they should or can be applied in practice. The duration of the protection and the size of the no-take areas have showed to play an important role in MPA effectiveness (Vandeperre et al., 2011). Should the spatial closures encompass only the fishery that targets sardine or be a total closure of fishing activities in the area? What about other sectors such as tourism or recreational fishing? And how do they affect other species and trophic interactions? These are examples of some questions emerging when talking about spatial closures in EAFM (Halpern et al., 2010). In this study, an attempt was made to deal with those questions by including criteria such as “other fisheries” and biotic interactions (sardine competitors and predators). As shown in this study closure of all fishing activities might considerably affect the fishing industry. Biotic interactions are very difficult to represent in this kind of analysis and their involvement was limited to mapping their distribution and indicating the areas overlapping with sardine. To obtain information about how spatial closures might affect trophic interactions and the ecosystem, the application of more sophisticated methods, such as ecosystem models (Polovina, 1984; Christensen and Pauly, 1992; Walters et al., 1997, 1999; Fulton et al., 2004a, 2004b) might be required. For example, Colléter et al. (2014), used food web models of Mediterranean and Senegalese ecosystems and found out the benefits of an MPA, such as increased biomass and catch yields, to be dependent on the MPA size. One of the limitations of the performed analysis was data availability. Lack of long time-series impeded the inclusion of statistical predictive modelling which have the capability to reduce uncertainty related to variability in tested variables. GIS environment support integration of spatial analysis with other tools in order to represent the temporal domain (Sahin and Mohamed, 2013; Chen et al., 2015), but these approaches are extremely data demanding, and in our case study, spatio-temporal data, especially related to VMS fishery data, were not available. In our study, 6 years of data were averaged and subjected to the analysis to reduce the uncertainty in the spatial distribution. The period considered is a good representation of our case study because it covers a time when both, high, and low sardine biomass, was observed (ICES, 2017a). Another important data limitation is the lack of seasonality. Variables included in the analysis, such as productivity patterns, zooplankton, sardine eggs and sardine distribution, tend to vary with season (Bernal et al., 2007; Moita et al., 2010; Zwolinski et al., 2010; Sobrinho-Gonçalves et al., 2013). However, Portuguese surveys do not collect seasonal data for these variables. This limitation might increase the uncertainty of the results and consequently impede the implementation of temporal and permanent conservation areas, thus more studies should be done to analyse the temporal variation of those variables and their effect on sardine biomass. One of the study’s strengths was the application of the AHP process to select weights. Although, it is a common practice in MCDM studies (Phua and Minowa, 2005; Chen et al., 2010; Gdoura et al., 2015) sometimes it is criticized (Dyer, 1990). Nevertheless, this process is appropriate for this type of analysis because it enables the inclusion of multiple stakeholders (Phua and Minowa, 2005; De Feo and De Gisi, 2010). Stakeholder involvement in EAFM decision support frameworks is essential and a highly recommended practice (Varjopuro et al., 2008; Kincaid et al., 2017). Even though, stakeholders were not included in the weights selection in this study, the design and methodology applied easily enables their inclusion in the future. In spatial analysis the uncertainty is introduced at its various stages (Refsgaard et al., 2007; Mosadeghi et al., 2013). In our case, uncertainty can arise from satellite data resolution conversion, creation of raster maps from point observations, identification of criteria and weights selection. In the case of criteria weights estimation, uncertainty was addressed through the performance of a SA which revelled that majority of variables are quite stable to the changes introduced by the weights (Supplementary Figure S1). Uncertainty was not quantitatively assessed in other stages, but its presence is acknowledged and therefore results should be interpreted with caution. Furthermore, it should be kept in mind that an array of other available statistical approaches, which could substitute the methods used in the present work (e.g. WLC method, AHP, Jenks natural breaks reclassification method), exist and their selection might considerably influence the results. The biggest difficulty in the application of this type of methodologies is the integration of the various data, disciplines, and institutions together (Leslie and McLeod, 2007). The united effort, collaboration and data sharing between scientific institutions, governments, fisheries institutions and non-governmental organizations is extremely important to a successful development of this and other data-hungry, integrative methods. In this study, scientists specialized in distinctive fields were involved. Moreover, various type of data, e.g. fisheries, satellite, landings, and biological were compiled and processed. Conclusion The proposed framework presents a useful tool to support an EAFM implementation. The analysis performed with Portuguese sardine fishery as a study case, proved that a combination of GIS and MCDM methodology could synthesize existing information in a fishery system, aid in the visualization of spatial existing problems and pinpoint areas which require special attention. This tool stands out as an efficient first step in an EAFM decision support framework. Further development of the method should start by performing seasonal analysis that will reflect annual variation in species distribution. Other aspects that should be addressed in the future are the involvement of stakeholders in the weights selection process and quantitative assessment of uncertainty in the other stages of the analysis, especially surface maps production and method selection. Supplementary data Supplementary material is available at the ICESJMS online version of the manuscript. Acknowledgements The work of D. Szalaj was supported by an FCT doctoral grant, Ref: PD/BD/114540/2016. Acoustic and bottom trawl surveys were financed by the EU/DG Fisheries’ Data Collection Framework (DCF). Logbook and VMS data were provided by the Portuguese Direcção-Geral de Recursos Marinhos Segurança e Serviços Marinhos. The authors thank to everyone who was involved in data collection and processing. Also, we would like to acknowledge editor and two anonymous reviewers for their constructive criticism that considerably improved this manuscript. References Abbott J. K. , Haynie A. C. 2012 . What are we protecting? Fisher behavior and the unintended consequences of spatial closures as a fishery management tool . Ecological Applications , 22 : 762 – 777 . Google Scholar CrossRef Search ADS PubMed Aswani S. , Lauer M. 2006 . Incorporating fishers’ local knowledge and behavior into geographical information systems (GIS) for designing marine protected areas in oceania . Human Organization , 65 : 81 – 102 . Google Scholar CrossRef Search ADS Bakun A. 1997 . Patterns in the Ocean: Ocean Processes and Marine Population Dynamics, Vol. 17, pp. 1945 – 1946 . California Sea Grant, La Jolla, CA, and Centro de Investigaciones Biologicas del Noroeste, La Paz, BCS, Mexico. Bernal M. , Stratoudakis Y. , Coombs S. , Angelico M. M. , de Lanzós A. L. , Porteiro C. , Sagarminaga Y. 2007 . Sardine spawning off the European Atlantic coast: characterization of and spatio-temporal variability in spawning habitat . Progress in Oceanography , 74 : 210 – 227 . Google Scholar CrossRef Search ADS Bottero M. , Comino E. , Duriavig M. , Ferretti V. , Pomarico S. 2013 . The application of a Multicriteria Spatial Decision Support System (MCSDSS) for the assessment of biodiversity conservation in the Province of Varese (Italy) . Land Use Policy , 30 : 730 – 738 . Google Scholar CrossRef Search ADS Briggs J. C. 1974 . Marine Zoogeography. McGraw Hill Book Company , New York . 475 pp. Cabral H. N. , Murta A. G. 2002 . The diet of blue whiting, hake, horse mackerel and mackerel off Portugal . Journal of Applied Ichthyology , 18 : 14 – 23 . Google Scholar CrossRef Search ADS Carocci F. , Bianchi G. , Eastwood P. , Meaden G. 2009 . Geographic Information Systems to support the ecosystem approach to fisheries: status, opportunities and challenges . Aquaculture , 101 . FAO Fisheries and Aquaculture Technical Paper, 532. Rome, FAO; 101 pp. Carr M. E. 2001 . Estimation of potential productivity in Eastern Boundary Currents using remote sensing . Deep-Sea Research Part II: Topical Studies in Oceanography , 49 : 59 – 80 . Google Scholar CrossRef Search ADS Chao A. , Chazdon R. L. , Colwell R. K. , Shen T. 2005 . A new statistical approach for assessing similarity of species composition with incidence and abundance data. Ecology Letters, 8: 148–159. Checkley D. M. , Ortner P. B. , Settle L. R. , Cummings S. R. 1997 . A continuous, underway fish egg sampler . Fisheries Oceanography , 6 : 58 – 73 . Google Scholar CrossRef Search ADS Chen Y. , Yu J. , Khan S. 2010 . Spatial sensitivity analysis of multi-criteria weights in GIS-based land suitability evaluation . Environmental Modelling and Software , 25 : 1582 – 1591 . Google Scholar CrossRef Search ADS Chen Y. , Liu R. , Barrett D. , Gao L. , Zhou M. , Renzullo L. , Emelyanova I. 2015 . A spatial assessment framework for evaluating flood risk under extreme climates . Science of the Total Environment , 538 : 512 – 523 . Google Scholar CrossRef Search ADS PubMed Chiou H. K. , Tzeng G. H. , Cheng D. C. 2005 . Evaluating sustainable fishing development strategies using fuzzy MCDM approach . Omega , 33 : 223 – 234 . Google Scholar CrossRef Search ADS Christensen V. , Pauly D. 1992 . ECOPATH II - a software for balancing steady-state ecosystem models and calculating network characteristics . Ecological Modelling , 61 : 169 – 185 . Google Scholar CrossRef Search ADS Coelho M. , Domingues P. , Balguerias E. , Fernandez M. , Andrade J. P. 1997 . A comparative study of the diet of Loligo vulgaris (Lamarck, 1799) (Mollusca: cephalopoda) from the south coast of Portugal and the Saharan Bank (Central-East Atlantic) . Fisheries Research , 29 : 245 – 255 . Google Scholar CrossRef Search ADS Cole J. , Mcglade J. 1998 . Clupeoid population variability, the environment and satellite imagery in coastal upwelling systems . Reviews in Fish Biology and Fisheries , 8 : 445 – 471 . Google Scholar CrossRef Search ADS Colléter M. , Gascuel D. , Albouy C. , Francour P. , Tito de Morais L. , Valls A. , Le Loc'h F. 2014 . Fishing inside or outside? A case studies analysis of potential spillover effect from marine protected areas, using food web models . Journal of Marine Systems , 139 : 383 – 395 . Google Scholar CrossRef Search ADS Costanza R. 1998 . Principles for sustainable governance of the oceans . Science , 281 : 198 – 199 . Google Scholar CrossRef Search ADS PubMed Daniel C. 1973 . One-at-a-time plans . Journal of the American Statistical Association , 68 : 353 – 360 . Google Scholar CrossRef Search ADS Dapueto G. , Massa F. , Costa S. , Cimoli L. , Olivari E. , Chiantore M. , Federici B. et al. 2015 . A spatial multi-criteria evaluation for site selection of offshore marine fish farm in the Ligurian Sea, Italy . Ocean and Coastal Management , 116 : 64 – 77 . Google Scholar CrossRef Search ADS De Feo G. , De Gisi S. 2010 . Using an innovative criteria weighting tool for stakeholders involvement to rank MSW facility sites with the AHP . Waste Management , 30 : 2370 – 2382 . Google Scholar CrossRef Search ADS PubMed De Freitas D. M. , Tagliani P. R. A. 2009 . The use of GIS for the integration of traditional and scientific knowledge in supporting artisanal fisheries management in southern Brazil . Journal of Environmental Management , 90 : 2071 – 2080 . Google Scholar CrossRef Search ADS PubMed Dyer J. S. 1990 . Remarks on the Analytic Hierarchy Process . Management Science , 36 : 249 – 258 . Google Scholar CrossRef Search ADS Council E. U. , 2009 . Official Journal of the European Union L 300/1, 2009: 1 – 33 . FAO . 2014 . The State of World Fisheries and Aquaculture 2014. Rome. 223 pp. Fiske S. J. 1992 . Sociocultural aspects of establishing marine protected areas . Ocean and Coastal Management , 17 : 25 – 46 . Google Scholar CrossRef Search ADS Fiúza A. F. D. , de Macedo M. E. , Guerreiro M. R. 1982 . Climatological space and time variation of the Portuguese coastal upwelling . Oceanologica Acta , 5 : 31 – 40 . Fletcher W. J. 2008 . Implementing an ecosystem approach to fisheries management: lessons learned from applying a practical EAFM framework in Australia and the Pacific. In The Ecosystem Approach to Fisheries , pp. 112 – 124 . Ed. by Bianchi G. , Skjodal H. R. . Food and Agriculture Organization of the United Nations (FAO ), Rome, Italy. Google Scholar CrossRef Search ADS Fulton E. A. , Smith A. D. M. , Johnson C. R. 2004a . Effects of spatial resolution on the performance and interpretation of marine ecosystem models . Ecological Modelling , 176 : 27 – 42 . Google Scholar CrossRef Search ADS Fulton E. A. , Parslow J. S. , Smith A. D. M. , Johnson C. R. 2004b . Biogeochemical marine ecosystem models II: the effect of physiological detail on model performance . Ecological Modelling , 173 : 371 – 406 . Google Scholar CrossRef Search ADS Garcia S. , Zerbi A. , Aliaume C. , Do Chi T. , Lasserre G. 2003 . The ecosystem approach to fisheries. Issues, terminology, principles, institutional foundations, implementation and outlook . FAO Fisheries Technical Paper , 443 : 71 . Garcia S. M. , Cochrane K. L. 2005 . Ecosystem approach to fisheries: a review of implementation guidelines . ICES Journal of Marine Science , 62 : 311 – 318 . Google Scholar CrossRef Search ADS Garrido S. , Silva A. , Pastor J. , Dominguez R. , Silva A. V. , Santos A. M. 2015 . Trophic ecology of pelagic fish species off the Iberian coast: diet overlap, cannibalism and intraguild predation . Marine Ecology Progress Series , 539 : 271 – 286 . Google Scholar CrossRef Search ADS Gdoura K. , Anane M. , Jellali S. 2015 . Geospatial and AHP-multicriteria analyses to locate and rank suitable sites for groundwater recharge with reclaimed water. Resources, Conservation and Recycling, 104: 19–30. Hall G. B. , Close C. H. 2007 . Local knowledge assessment for a small-scale fishery using geographic information systems . Fisheries Research , 83 : 11 – 22 . Google Scholar CrossRef Search ADS Halpern B. S. , Warner R. R. 2002 . Marine reserves have rapid and long lasting effects . Ecology Letters , 5 : 361 – 366 . Google Scholar CrossRef Search ADS Halpern B. S. , Lester S. E. , McLeod K. L. 2010 . Placing marine protected areas onto the ecosystem-based management seascape . Proceedings of the National Academy of Sciences of the United States of America , 107 : 18312 – 18317 . Google Scholar CrossRef Search ADS PubMed Hopkins L. D. 1977 . Methods for generating land suitability maps: a comparative evaluation . Journal of the American Planning Association , 43 : 386 – 400 . Hossain M. S. , Das N. G. 2010 . GIS-based multi-criteria evaluation to land suitability modelling for giant prawn (Macrobrachium rosenbergii) farming in Companigonj Upazila of Noakhali, Bangladesh . Computers and Electronics in Agriculture , 70 : 172 – 186 . Google Scholar CrossRef Search ADS ICES . 2016 . Second Interim report of the Working Group on Acoustic and Egg Surveys for Sardine and Anchovy in ICES Areas VII, VIII and IX (WGACEGG), 16–26 November 2015, Lowestoft, UK. ICES CM: 2015/SSGIEOM: 31. 398 pp. ICES . 2017a . Report of the Working Group on Southern Horse Mackerel, Anchovy and Sardine (WGHANSA), 24–29 June 2017, Bilbao, Spain. ICES CM: 2017/ACOM: 17. 547 pp. ICES . 2017b . Manual of the IBTS North Eastern Atlantic Surveys . Series of ICES Survey Protocols SISP , 15 : 92 . Ikeda T. , Omori M. 1992 . Methods in Marine Zooplankton Ecology. Krieger Plubishing Company , Florida . 329 pp. INE–Instituto Nacional de Estatística . 2006 . Estatísticas da Pesca 2005. INE, Lisboa, 79 pp. INE–Instituto Nacional de Estatística . 2007 . Estatísticas da Pesca 2006. INE, Lisboa, 97 pp. INE–Instituto Nacional de Estatística . 2008 . Estatísticas da Pesca 2007. INE, Lisboa, 97 pp. INE–Instituto Nacional de Estatística . 2009 . Estatísticas da Pesca 2008. INE, Lisboa, 98 pp. INE–Instituto Nacional de Estatística . 2010 . Estatísticas da Pesca 2009. INE, Lisboa, 101 pp. INE–Instituto Nacional de Estatística . 2011 . Estatísticas da Pesca 2010. INE, Lisboa, 104 pp. Innes J. P. , Pascoe S. 2010 . A multi-criteria assessment of fishing gear impacts in demersal fisheries . Journal of Environmental Management , 91 : 932 – 939 . Elsevier Ltd. Google Scholar CrossRef Search ADS PubMed Jaccard P. 1908 . Nouvelles recherches sur la distribution florale . Bulletin de la Société Vaudoise Des Sciences Naturelles , 44 : 223 – 270 . Jenks G. F. 1967 . The data model concept in statistical mapping . International Yearbook of Cartography , 7 : 186 – 190 . Jones P. J. S. 2006 . Collective action problems posed by no-take zones . Marine Policy , 30 : 143 – 156 . Google Scholar CrossRef Search ADS Katara I. , Silva A. 2017 . Mismatch between VMS data temporal resolution and fishing activity time scales . Fisheries Research , 188 : 1 – 5 . Google Scholar CrossRef Search ADS Kellner J. B. , Tetreault I. , Gaines S. D. , Nisbet R. M. 2007 . Fishing the line near marine reserves in single and multispecies fisheries. Ecological Applications, 17: 1039–1054. Kincaid K. , Rose G. , Devillers R. 2017 . How fisher-influenced marine closed areas contribute to ecosystem-based management: a review and performance indicator scorecard . Fish and Fisheries , 18 : 860 – 876 . Google Scholar CrossRef Search ADS Kjærsgaard J. 2007 . Multiple criteria decision making in fisheries management: experiences and implications . Food Economics - Acta Agriculturae Scandinavica, Section C , 4 : 49 – 64 . Google Scholar CrossRef Search ADS Leite M. C. F. , Gasalla M. A. 2013 . A method for assessing fishers’ ecological knowledge as a practical tool for ecosystem-based fisheries management: seeking consensus in Southeastern Brazil . Fisheries Research , 145 : 43 – 53 . Google Scholar CrossRef Search ADS Leslie H. M. , McLeod K. L. 2007 . Confronting the challenges of implementing marine ecosystem-based management . Frontiers in Ecology and the Environment , 5 : 540 – 548 . Google Scholar CrossRef Search ADS Lester S. E. , Halpern B. S. , Grorud-Colvert K. , Lubchenco J. , Ruttenberg B. I. , Gaines S. D. , Airamé S. et al. 2009 . Biological effects within no-take marine reserves: a global synthesis . Marine Ecology Progress Series , 384 : 33 – 46 . Google Scholar CrossRef Search ADS Link J. S. , Brodziak J. K. T. , Edwards S. F. , Overholtz W. J. , Mountain D. , Jossi J. W. , Smith T. D. , et al. 2002 . Marine ecosystem assessment in a fisheries management context . Canadian Journal of Fisheries and Aquatic Sciences , 59 : 1429 – 1440 . Google Scholar CrossRef Search ADS Link J. S. 2005 . Translating ecosystem indicators into decision criteria . ICES Journal of Marine Science , 62 : 569 – 576 . Google Scholar CrossRef Search ADS Link J. S. 2010 . Ecosystem-based fisheries management: Confronting tradeoffs. Cambridge University Press, New York. 221 pp. Mahe K. , Amara R. , Bryckaert T. , Kacher M. , Brylinski J. M. 2007 . Ontogenetic and spatial variation in the diet of hake (Merluccius merluccius) in the Bay of Biscay and the Celtic Sea . Journal of Marine Science , 64 : 1210 – 1219 . Malczewski J. 2000 . On the use of weighted linear combination method in GIS: common and best practice approaches . Transactions in GIS , 4 : 5 – 22 . Google Scholar CrossRef Search ADS Malczewski J. 2006 . GIS-based multicriteria decision analysis: a survey of the literature . International Journal of Geographical Information Science , 20 : 703 – 726 . Google Scholar CrossRef Search ADS Mardle S. , Pascoe S. 1999 . A review of applications of multiple-criteria decision-making techniques to fisheries . Marine Resource Economics , 14 : 41 – 63 . Google Scholar CrossRef Search ADS Massam B. H. 1988 . Multi-Criteria Decision Making (MCDM) techniques in planning . Progress in Planning , 30 : 1 – 84 . Google Scholar CrossRef Search ADS Maxwell D. L. , Stelzenmüller V. , Eastwood P. D. , Rogers S. I. 2009 . Modelling the spatial distribution of plaice (Pleuronectes platessa), sole (Solea solea) and thornback ray (Raja clavata) in UK waters for marine management and planning . Journal of Sea Research , 61 : 258 – 267 . Google Scholar CrossRef Search ADS Meaden G. , Aguilar-Manjarrez J. 2013 . Advances in geographic information systems and remote sensing for fisheries and aquaculture, FAO Fisheries And Aquaculture Technical Paper 552. Food and Agriculture Organization of the United Nations (FAO) Rome, 2013. 425 pp. Moita M. T. , Silva A. , Palma S. , Vilarinho M. G. 2010 . The coccolithophore summer-autumn assemblage in the upwelling waters of Portugal: patterns of mesoscale distribution (1985-2005) . Estuarine, Coastal and Shelf Science , 87 : 411 – 419 . Google Scholar CrossRef Search ADS Moore C. H. , Harvey E. S. , Van Niel K. P. 2009 . Spatial prediction of demersal fish distributions: enhancing our understanding of species-environment relationships . In ICES Journal of Marine Science , 66 : 2068 – 2075 . Google Scholar CrossRef Search ADS Mosadeghi R. , Warnken J. , Tomlinson R. , Mirfenderesk H. 2013 . Uncertainty analysis in the application of multi-criteria decision-making methods in Australian strategic environmental decisions . Journal of Environmental Planning and Management , 56 : 1097 – 1124 . Google Scholar CrossRef Search ADS Olson J. 2011 . Producing nature and enacting difference in ecosystem-based fisheries management: an example from the Northeastern US . Marine Policy , 35 : 528 – 535 . Google Scholar CrossRef Search ADS Pérez O. M. , Telfer T. C. , Ross L. G. 2005 . Geographical information systems-based models for offshore floating marine fish cage aquaculture site selection in Tenerife, Canary Islands . Aquaculture Research , 36 : 946 – 961 . Google Scholar CrossRef Search ADS Petitgas P. , Doray M. , Huret M. , Massé J. , Woillez M. 2014 . Modelling the variability in fish spatial distributions over time with empirical orthogonal functions: anchovy in the Bay of Biscay . ICES Journal of Marine Science , 71 : 2379 – 2389 . Google Scholar CrossRef Search ADS Phua M. H. , Minowa M. 2005 . A GIS-based multi-criteria decision making approach to forest conservation planning at a landscape scale: a case study in the Kinabalu Area, Sabah, Malaysia . Landscape and Urban Planning , 71 : 207 – 222 . Google Scholar CrossRef Search ADS Pikitch E. S. , Boersma P. D. , Boyd I. L. , Conover D. O. , Cury P. , Essington T. , Heppell S. S. et al. 2012 . Little Fish, Big Impact, pp.. Lenfest Ocean Program, Washington, DC. Polovina J. J. 1984 . Model of a coral reef ecosystem - I. The ECOPATH model and its application to French Frigate Shoals . Coral Reefs , 3 : 1 – 11 . Google Scholar CrossRef Search ADS Portman M. E. 2007 . Zoning design for cross-border marine protected areas: the Red Sea Marine Peace Park case study . Ocean and Coastal Management , 50 : 499 – 522 . Google Scholar CrossRef Search ADS Rahman M. A. , Rusteberg B. , Uddin M. S. , Lutz A. , Saada M. A. , Sauter M. 2013 . An integrated study of spatial multicriteria analysis and mathematical modelling for managed aquifer recharge site suitability mapping and site ranking at Northern Gaza coastal aquifer . Journal of Environmental Management , 124 : 25 – 39 . Google Scholar CrossRef Search ADS PubMed Ramírez-Monsalve P. , Raakjær J. , Nielsen K. N. , Santiago J. L. , Ballesteros M. , Laksá U. , Degnbol P. 2016a . Ecosystem Approach to Fisheries Management (EAFM) in the EU - Current science-policy-society interfaces and emerging requirements . Marine Policy , 66 : 83 – 92 . Google Scholar CrossRef Search ADS Ramírez-Monsalve P. , Raakjær J. , Nielsen K. N. , Laksá U. , Danielsen R. , Degnbol D. , Ballesteros M. et al. 2016b . Institutional challenges for policy-making and fisheries advice to move to a full EAFM approach within the current governance structures for marine policies . Marine Policy , 69 : 1 – 12 . Google Scholar CrossRef Search ADS Refsgaard J. C. , van der Sluijs J. P. , Højberg A. L. , Vanrolleghem P. A. 2007 . Uncertainty in the environmental modelling process - A framework and guidance . Environmental Modelling and Software , 22 : 1543 – 1556 . Google Scholar CrossRef Search ADS Rodríguez-Climent S. , Angélico M. M. , Marques V. , Oliveira P. , Wise L. , Silva A. 2017 . Essential habitat for sardine juveniles in Iberian waters . Scientia Marina , 81 : 351–000. Romero C. , Rehman T. 1987 . Natural resource management and the use of multiple criteria decision-making techniques: a review . European Review of Agricultural Economics , 14 : 61 – 89 . Google Scholar CrossRef Search ADS Ruckelshaus M. , Klinger T. , Knowlton N. , DeMaster D. P. 2008 . Marine ecosystem-based management in practice : scientific and governance challenges . BioScience, 58: 53–63. Saaty T. L. 1980 . The Analytic Hierarchy Process. McGraw-Hill, New York. 287 pp. Sahin O. , Mohamed S. 2013 . A spatial temporal decision framework for adaptation to sea level rise . Environmental Modelling and Software , 46 : 129 – 141 . Google Scholar CrossRef Search ADS Silva A. , Moreno A. , Riveiro I. , Santos B. , Pita C. , Rodrigues J. G. , Villasante S. , et al. 2015 . Research for PECH committee- Sardine fisheries: resource assessment and social and economic situation. Study, European Parliament's Committee on Fisheries. IP/B/PECH/IC/2015_133, 68 pp. Silva M. A. 1999 . Diet of common dolphins, Delphinus delphis, off the Portuguese continental coast . Journal of the Marine Biological Association of the United Kingdom , 79 : 531 – 540 . Google Scholar CrossRef Search ADS Simmonds J. , MacLennan D. 2005 . Biological Acoustics. Fisheries Acoustics. Theory and Practice: Fisheries Acoustics, 2nd edn. Blackwell Science, Oxford, 437 pp. Sobrinho-Gonçalves L. , Moita M. T. , Garrido S. , Cunha M. E. 2013 . Environmental forcing on the interactions of plankton communities across a continental shelf in the Eastern Atlantic upwelling system . Hydrobiologia , 713 : 167 – 182 . Google Scholar CrossRef Search ADS Staples D. , Brainard R. , Capezzuoli S. , Funge-smith S. , Grose C. , Heenan A. , Hermes R. et al. 2014 . Essential EAFM. Ecosystem Approach to Fisheries Management Training Course. 319 pp. Stelzenmüller V. , Maynou F. , Bernard G. , Cadiou G. , Camilleri M. , Crec’hriou R. , Criquet G. et al. 2008 . Spatial assessment of fishing effort around European marine reserves: implications for successful fisheries management . Marine Pollution Bulletin , 56 : 2018 – 2026 . Google Scholar CrossRef Search ADS PubMed Stewart R. R. , Possingham H. P. 2005 . Efficiency, costs and trade-offs in marine reserve system design . Environmental Modeling and Assessment , 10 : 203 – 213 . Google Scholar CrossRef Search ADS Tomlin C. D. 1991 . Cartographic Modelling . Geographic Information Systems , 1 : 361 – 374 . Vandeperre F. , Higgins R. M. , Sánchez-Meca J. , Maynou F. , Goñi R. , Martín-Sosa P. , Pérez-Ruzafa A. , et al. 2011 . Effects of no-take area size and age of marine protected areas on fisheries yields: a meta-analytical approach . Fish and Fisheries , 12 : 412 – 426 . Google Scholar CrossRef Search ADS Varjopuro R. , Gray T. , Hatchard J. , Rauschmayer F. , Wittmer H. 2008 . Introduction: interaction between environment and fisheries-The role of stakeholder participation . Marine Policy , 32 : 147 – 157 . Google Scholar CrossRef Search ADS Villa F. , Tunesi L. , Agardy T. 2002 . Zoning marine protected areas through spatial multiple-criteria analysis: the case of the Asinara Island National Marine Reserve of Italy . Conservation Biology , 16 : 515 – 526 . Google Scholar CrossRef Search ADS Walters C. , Christensen V. , Pauly D. 1997 . Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments . Reviews in Fish Biology and Fisheries , 7 : 139 – 172 . Google Scholar CrossRef Search ADS Walters C. , Pauly D. , Christensen V. 1999 . Ecospace: prediction of mesoscale spation patterns in trophi relatiohsips of exploited ecosystem, with empahsis on the impacts of marine protected areas . Ecosystems , 2 : 539 – 554 . Google Scholar CrossRef Search ADS Witt M. J. , Godley B. J. 2007 . A step towards seascape scale conservation: using vessel monitoring systems (VMS) to map fishing activity . PLoS One , 2 : e1111. Google Scholar CrossRef Search ADS PubMed Wooster W. S. , Bakun A. , McLain R. M. 1976 . Seasonal upwelling cycle along the Eastern boundary of the North Atlantic . Journal of Marine Research , 34 : 131 – 141 . Yeon I. , Song M. Y. , Sohn M. H. , Hwang H. J. , Im Y. J. , Kim D. H. , Park H. W. et al. 2011 . An ecosystem-based assessment of the Korean blue crab trammel net fishery in the Yellow Sea and management implications . Fisheries Research , 112 : 179 – 188 . Google Scholar CrossRef Search ADS Zwolinski J. P. , Oliveira P. B. , Quintino V. , Stratoudakis Y. 2010 . Sardine potential habitat and environmental forcing off western Portugal . ICES Journal of Marine Science , 67 : 1553 – 1564 . Google Scholar CrossRef Search ADS © International Council for the Exploration of the Sea 2018. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

ICES Journal of Marine ScienceOxford University Press

Published: Jul 31, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off