Estimation of the Coastal Vulnerability Index Using Multi-Criteria Decision Making: The Coastal Social–Ecological System of Rachgoun, Western Algeria

Abstract

This research deals with spatial vulnerability in the coastal area of Rachgoun (Algeria), on the southern shore of the Mediterranean Sea. Over the past two decades, the coastal area of Rachgoun has been suffering from a large amount of pressure due to accelerated socioeconomic development, urbanization, tourism, fishing, and agriculture. The main objective of this study is to visualize the coastal vulnerability of Rachgoun using multi-criteria decision making (MCDM). A multidisciplinary approach that integrates geological, physical, and socioeconomic vulnerability was adopted. The selected parameters for the study include lithology, elevation, slope, shoreline change, significant wave height, population density, tourist density, land use/land cover (LULC), road network density, proximity to coast, distance from river, people’s awareness, and designated conservation areas and cultural heritage. Data from AlSat-2 Satellite imagery, aerial photography, topographic maps, and field surveys were processed. Spatial modeling was conducted through the MCDM approach and geographic information systems (GIS) to develop two sub-indices: the natural vulnerability index (CVI_N) and the socioeconomic vulnerability index (CVI_S). The combination of the two sub-indices allowed us to deduce the integrated coastal vulnerability index (ICVI). The outcomes present a coastal vulnerability map with a spatial resolution of 10 m of the identified problematic area. This map can guide decision-makers in implementing an effective integrated coastal zone management (ICZM) strategy.

1. Introduction

Coastal areas are specific regarding natural and anthropogenic components [1]. They are characterized by a high population density and accelerated socioeconomic growth [2,3,4,5]. Considered providers of ecosystem services on a global and a local scale [6], coastal areas host lot of vested interests in terms of financial gain and livelihood [7]. Approximately 40% of the world’s population lives within 100 km from the coastline [8]. Interactions between natural and human-made factors are the fundamental determinant of this area’s specificity and complexity. According to [9], Social–ecological systems are complex systems in which human and ecological components are in mutual interaction. The idea of a “socioecological” system emphasizes people’s interdependence with nature and rejects the artificial distinction between social and ecological systems [10,11,12]. In order to improve management policies and strategies, it is necessary to adopt tools that interlink social and ecological systems, and integrate local and scientific knowledge [13]. Several studies demonstrate that studying the coast in its complexity requires a new approach integrating ecological and socioeconomic (social) components [9,12,14]. Coastal zones are defined as coastal social–ecological systems (CSES) in which natural and human factors interact regularly [15,16,17].

Because of their proximity to the sea, coastal areas are vulnerable. However, anthropogenic activities accentuate coastal vulnerability. The fragility of this ecosystem, coupled with the high human dependency on coastal ecosystem services, requires an urgent study of their vulnerability. Due to the specificity of coastal zones, vulnerability must be understood in a holistic way [18]. Turner et al. and Adger et al. [19,20] suggested that human and biophysical vulnerability are interconnected and should be treated accordingly. Coastal vulnerability is frequently addressed to assess the risk of coastal regions due to physical, environmental, social, and economic vulnerabilities [18,21].

Different approaches have been used to assess coastal vulnerability [8,21]. The coastal vulnerability index (CVI) proposed by [22] is widely applied [8,23,24,25,26]. The original CVI considered geological and physical parameters, without incorporating human dimension. However, it argued that integrating population density as a social variable improves the CVI index [22]. In this study, the ICVI is estimated by integrating both natural (geologic, physical) and socioeconomic variables.

Algeria opens onto the southern Mediterranean Sea with a coastline of over 2148 km. The northern region of the country is considered as coastal area, with a width of 50 km to 100 km from east to west covering an area of 45,000 km², i.e., less than 2% of the territory, but hosting 36% of the population [27]. This area has a great interest related to socioeconomic needs: in particular, fishing and aquaculture activities, tourism, agriculture, industrial establishment, and infrastructures. On the other hand, Algerian coastal cities are affected by multiple threats related mainly to submersion, erosion, landslide, marine intrusion, and pollution [28,29,30,31,32,33,34,35,36]; therefore, the vulnerability is high. However, there are very few studies related to CVI in Algeria [37,38,39,40]. The existing studies are mostly based on physical criteria without any real integration of the socioeconomical component. This is probably due to difficulty accessing the necessary data [39,40]. Moreover, the interest in integration processes in Algeria is recent, officially starting with the implementation of the national integrated coastal zone management (ICZM) strategy in 2015 [41], and it is still difficult to adopt it on the ground. Therefore, to overcome these limitations, the present study aims to integrate natural and socioeconomic criteria and to develop a novel integrated coastal vulnerability index (ICVI) in the context of CSES. The three pillars of sustainability, namely social, economic, and environmental, are susceptible to negative impact in coastal areas. As a result, this study considers a large number of socioeconomic and natural aspects including local knowledge, which many previous studies lacked. Vulnerability assessment considers real interactions without depending on one process, i.e., it does not focus on flood and sea level rise hazards, but multihazard assessment was adopted. Coastal vulnerability is assessed on high spatial resolution (10 m) and based on the social–ecological system approach. It is also the first application in the western coast of Algeria (Oranie region). Based on an integrated approach, two sub-indices have been developed: the natural vulnerability sub-index (CVI_N), and the socioeconomic vulnerability sub-index (CVI_S). The CVI_N includes seven physical and geomorphological parameters (lithology, elevation, slope, significant wave height (Hs), shoreline change, proximity to coast, distance from river), and CVI_S includes seven socioeconomic variables (land use/land cover (LULC), road network density, population density, number of tourists, people’s awareness, and designated conservation areas and cultural heritage). The combination of these two sub-indices yields the global integrated coastal vulnerability index (ICVI). The indices were obtained through the multi-criteria decision making process (MCDM) [42], considering five levels of vulnerability ranging from very low to very high, in a geographic information system (GIS) environment.

The interpretation of the results obtained is mainly based on local geographical expertise, considered a basis for geographic information [43], combining spatial process information with geographic region knowledge. Updated field data, obtained during several field campaigns, surveys, and questionnaires conducted with various local stakeholders in the region between 2018 and 2022, were intersected with the mapped results in order to better reflect the vulnerability situation in the region.

In view of the fact that climate change, as well as unplanned development, increase coastal vulnerability [44], this study aims to develop an integrated coastal vulnerability index by including parameters form natural and socioeconomic criteria. The second objective is to apply MCDM as a new coastal management approach for coastal vulnerability assessment. Findings of this study therefore demonstrate the urgent need for an integrated approach to coastal vulnerability assessment, which serves as a tool to support decision making for Rachgoun CSES.

2. Materials and Methods

2.1. Study Area

The study area concerns a part of Northwest Algeria. On the international scale, it belongs to the Alboran Sea (Figure 1). It consists of a territorial unit comprising the lower valley of Tafna and extends to Rachgoun Island, defined in this document as a CSES. The area covers the two main coastal districts of Ain Temouchent Province: Beni Saf and Oulhaça El-Gheraba. In the east, Beni-Saf is considered as the most urbanized and populated coastal city with a population rate higher than the average of the Province (814 hab/km²), and an urbanization rate of 94%. Oulhaça El-Gheraba is a rural city and the second most populous coastal district (211 hab/km²), organized mostly in small villages (Budget Programming and Monitoring Department “BPMD”, 2020). Along a coastline of 27 km dominated by cliffs, there are four sandy beaches in the pocket mainly, from west to east: Rachgoun, Madrid, Marmite, and Wells Beach. Next to Wells beach (east of Beni Saf) is the port of Beni-Saf “Mersat Sidi Ahmed”, the main reason for the installation of Beni Saf city, by the first autochthonous population (fishermen). The administrative limit between the two municipalities Is defined by the Tafna Valley, which flows towards the sea by dividing Rachgoun beach into Rachgoun 1 and Rachgoun 2. At a distance of 2.5 km from the mouth is Rachgoun Island, or Layella in the local language, classified as a Ramsar site in 2011 (www.ramsar.org, accessed on 9 February 2022), and a future Marine Protected Area (MPA) (www.rac-spa.org, accessed on 18 February 2022). The region has acquired a certain celebrity among specialists in history and archeology [45,46,47,48,49,50] demonstrating the historical and archeological value of the region. Hence, an archeological complex is assigned. The main archeological and historical assets are:

• The prehistoric site of Rachgoun (7000 BC);
• Siga, Numidian kingdom of Syfax, also known as Siga-Takembrit by the local population (206 BC);
• The Mausoleum of Syph^ax (3rd century BC).

This region has been the subject of several sectorial and/or localized studies, without there being a systemic synthesis of man/nature interactions [50,51,52,53,54,55,56,57,58,59,60], which are mostly based on a classic and fragmentary approach. The coastal vulnerability issue has never been the subject of a study in the Northwest region of Algeria, the reason for which this study presents a new multidisciplinary approach based on an integrated approach for the study of coastal vulnerability.

Figure 1. Study area location.

Figure 1. Study area location.

2.2. Methodology

The complexity of the coastal area requires a multidisciplinary approach, which integrates variables focused on physical, geological, and socioeconomic dimensions. Based on the social ecological system approach (SES) [11,12,61,62], a large amount of data (maps, statistics, geospatial data, among others) were procured from Algerian government departments, relevant websites, various published reports and articles, and field surveys. To accomplish this study, satellite imagery (AlSat-2 (Algeria Satellite-2)), aerial photography, and topographic and geological maps were processed in a GIS environment (through ArcGIS software 10.7).

The vulnerability assessment consists of three distinct phases: (i) parameter selection; (ii) vulnerability ranking and relevant analytical technique; (iii) integrated spatial modeling and development of the CVI for the study area (Figure 2).

2.2.1. Parameter Selection and Ranking

Fourteen variables were considered relevant in reflecting coastal vulnerability in the study area (

Table 1. Data description and parameters.

Variable	Data Source	Spatial Resolution (m)/Scale	Coordinate System
Lithology	Geological map from Cartotheque University of Oran 2 (https://fstu.univ-oran2.dz/)	1/50,000	WGS_1984_UTM_Zone_30N
Elevation and slope	Topographic map from Cartotheque University of Oran 2 (https://fstu.univ-oran2.dz/)	1/25,000
Significant wave height (Hs)	Database from the ‘’Assessment of Wave energy along the Algerian coast over the 39 years’’ study (0) [28]	Mean of 39 years 1979–2017
Population density	Budget Programming and Monitoring Department (BPMD) (www.asjp.cerist.dz)	2019
Tourist density	Civil Protection Department (Ain Temouchent) (www.protectioncivile.dz)	2017
LULC	Alsat2 image from Algerian Space Agency (ASAL) (https://asal.dz/)	10 m (2019)
Road	Cartography Quickmap services plugins (QGIS 3.8 software) (www.qgis.org, accessed on 12 August 2020 )	2020
Proximity to coast and distance from river	Cartography from Alsat2 image from Algerian Space Agency (ASAL) (https://asal.dz/)	10 m (2019)
Shoreline change	Cartography from aerial photography from National Institute of Cartography and Remote Sensing (INCT) (http://www.inct.mdn.dz/)	2 m resolution (1980) and 1 m resolution (2019)
People’s awareness	Field data and questionnaire	Total of 358 samples (2018 to 2022)
Designated conservation areas and cultural heritage	Rachgoun wetland from RAMSAR Website (www.ramsar.org, accessed on 9 February 2022 ) Cartography in field using archeological data from National Office for the Management and Exploitation of Protected Cultural Assets (OGBEC), Ain Temouchent Province, with consultation of archeological experts of the region, Department of the Environment and Official Journal of the Algerian Republic (OJAR) (www.joradp.dz, accessed on 5 July 2020)	2020 (fieldwork)

). The selected parameters were aggregated in two main categories: natural parameters, including geomorphological and physical variables linked to natural characteristics and physical processes related to coastal environment dynamics (lithology, slope, elevation, distance from river, proximity to coast, shoreline change, and significant wave height); and socioeconomic parameters related to human-made variables (LULC, road network density, population density, number of tourists, people’s awareness, designated conservation areas, and cultural heritage). These parameters were ranked into 5 levels of vulnerability: Very Low, Low, Moderate, High, and Very High (

Table 2. Vulnerability ranking assigned for parameters.

	Variables	Ranking	Very Low (1)	Low (2)	Moderate (3)	High (4)	Very High (5)
Natural Parameters	Lithology		Volcanic	Metamorphic	Sedimentary rock	Consolidate dune	Sand plains
	Slope (degree)		<2	2–5	5–10	10–15	>15
	Elevation (m)		≤2	2–5	5–10	10–20	>20
	Significant wave height (m)		<0.35	0.35–0.70	0.70–1.05	1.05–1.70	>1.70
	Proximity to coast (m)		100	300	8000	3000	>3000
	Distance from river (m)		25	200	700	2500	>2500
	Shoreline change		>0	−0.5_0	−1.5_−0.5	−2_−1.5	≤−2
Socioeconomic Parameters	Land cover		Wet areas	Natural coastal zones	Forests/dense vegetation	Agricultural areas	Urban and/or industrial areas
	Road network density		0–2	2–5	5–15	10–15	>15
	Population density (person/Km2)		<200	200–350	350–450	450–600	>600
	Number of tourists		<150	150–250	250–350	350–500	>500
	People’s awareness (%)		<20	20–40	40–60	60–80	>80
	Designated conservation area status and cultural heritage		No status	Beni Ghenen (Syphax)	Siga	Rachgoun TEZ	Rachgoun MPA (environmentally sensitive areas)

). The data used were projected on the World Geodetic System (WGS) 1984_Universal Transverse Mercator (UTM) Zone 30N coordinate system (EPSG: 32630).

Natural Criteria 

Natural criteria includes two main groups: geomorphological parameters and physical parameters

i

Geomorphological Parameters

Geomorphological parameters refer to the coastal characteristics of the study area. Five parameters were selected as geomorphological components: lithology, slope, elevation, distance from river, and proximity to coast.

1.

Lithology

Lithology describes types of landforms (rock type), being a crucial parameter in studying the vulnerabilities [63,64,65]. This parameter is considered an important indicator of the erodibility (erosive activity) of different landform types [22,65,66,67].

In this study, the lithology data were acquired by digitizing the geological map of Algeria, Beni Saf, published in 1995 at a scale of 1:50.000 using ArcGIS software (Figure 3).

2.

Elevation and slope

Elevation and slope are indicators of inundation risk [22,67,68,69]. Elevation is a key factor in determining coastal vulnerability [63,65]. Other studies considered coastal slopes [66,70]. However, by combining both variables—elevation and slope—in obtaining CVI, pertinent results were produced [68,71,72]. In addition to information about susceptibility of flooding events, the coastal slope allows one to determine the interaction between seawater and groundwater [8,64,66,73].

This study considers that the lowest elevation areas are the most exposed to inundation [22,25,74,75] and nearshore areas with gentle slope are considered as highly vulnerable [67,68,73].

Elevation was estimated from a topographic map of Algeria, Beni Saf, on a 1:25.000 scale with a 10 m contour line. Elevation contours of the study area were generated using ArcGIS software in order to prepare an elevation map, from which the digital elevation model (DEM) was generated. The slope was estimated from the DEM using inverse distance weighting (IDW) interpolation in ArcGIS software (Figure 4).

Concerning regional elevation, this littoral band is part of the coastal mountain range of Beni Saf whose average altitude is 200 m; hence, elevation is marked as very low. The slope values within the study area, which range between 0–57.2%, are classified into five categories (Table 2). The study area is characterized mainly by steep slopes, which are rated as low vulnerability. A gentle slope is recorded mostly along the Tafna River and the small beaches of Rachgoun and Wells with a slope range of 0–2%, indicating very high vulnerability.

3.

Proximity to coast and distance from river

Proximity to coast and distance from river are mostly used in flood and submersion risk analysis [8,76,77,78]. They have an important role in vulnerability assessment. The proximity to the coast characterizes the risk of exposure to the energetic forces of the sea [8,76,79] and the impact of saltwater intrusion to groundwater resources [34,76]. In addition, areas near rivers and coastlines are potentially affected by floods [76,80]. The areas close to the mouth (embouchure) have also a potential risk related to pollution, in particular a foul smell problem [81,82]. Due to the mismatch between urbanization and wastewater treatment in developing countries, the accumulation of anthropogenic pollutants (both organic and inorganic) from waste streams generated along the banks leads to odorization [81]. This phenomenon affects the aquatic ecosystem and humans [82]. Hence, it is considered that vulnerability increases with the proximity to the shoreline and the river (Table 2).

This study considers the delimitation of coastal bands defined by coastal law 02-02 (Official Journal of the Algerian Republic (OJAR)), defined by the Algerian authorities for the protection and valorization of coastal resources, which defines four main intervals: 100 m, 300 m, 800 m, and 3 km [83]. The vulnerability classes are ranked successively as extremely high, high, moderate, low, and very low vulnerability. For distance from the river, five classes are established [84]: 25 m, 200 m, 700 m, 2000 m, and above 2000 m with a decreased vulnerability score. The classification of the proximity to coast and distance from river was conducted by creating a buffer distance from the coastline and the river, respectively (Figure 5).

Coastal law 02-02 of 5 February 2002 on the protection and enhancement of the coastline is defined in the OJAR: www.joradp.dz, accessed on 19 March 2022.

ii

Physical Parameters

The physical parameters determine natural forcing processes occurring in the study area. Two parameters were selected as physical variables: shoreline change and significant wave height (Hs).

1.

Shoreline change

Shoreline change is a relevant input for coastal hazard monitoring [42,85,86,87]. As sandy shorelines are highly subject to changes due to coastal processes, this study considers shoreline change on a low coast, which is at risk of flooding [88]. Moreover, this low-lying area is characterized by an important human settlement. Thus, shoreline change was estimated for the sandy beaches of Rachgoun, Madrid, Marmite, and Wells, between 1980 and 2017, using aerial photos with 2 m and 0.5 m of spatial resolution, respectively (Table 1). The extraction of information about shoreline change was performed with the Digital Shoreline Analysis System (DSAS) [71] (Figure 6).

The main agglomerations were developed and are still growing relatively close to the coast; in particular, near sandy beaches (Rachgoun, Madrid, and Wells). As a result, along the low coast, socioeconomic implications are very high. Given the importance of the coastal dynamic and the socioeconomical stakes, shoreline change parameters were calculated using DSAS for the study area’s low coast. Rocha et al [8] defines erosional susceptibility according to the rock hardness. Hard rocks have very low susceptibility to erosion. Therefore, cliffs composed mainly of volcanic and metamorphic rocks, as shown in the lithological chart, are of extremely high and high hardness and are highly resistant to erosion. In addition, they have very low socioeconomic stakes. Based on the literature [8,89,90,91], expert knowledge, and field survey observations, shoreline change is considered in the low vulnerability class for cliff types (Table 2).

2.

Significant wave height (Hs)

Significant wave height (Hs) is considered an indicator of vulnerability due to the sea level rise [22,63], and as variable of potential erosion [70]. Information about this parameter could be widely helpful for the implementation of development policies that support coastal protection. For this parameter, data were obtained from the previously published work about wave energy resource assessment along the Algerian coast, from 1979 to 2017 [28]. The mean significant wave height of Ain Temouchent coast reported for 39 years is 0.946 m (Figure 7); therefore, it is assigned as moderate vulnerability (Table 2).

Socioeconomic Variables 

Socioeconomic variables refer to human-influenced factors. It can multiply the vulnerability level of a given coastal area [25]. For this study, seven parameters were considered relevant: LULC, road network density, population density, number of tourists, people’s awareness, designated conservation areas, and cultural heritage.

1.

Land use/land cover (LULC)

Land use/land cover (LULC) is important for understanding coastal areas. Several studies used this parameter for studying coastal vulnerability [64,68,79,92]. Remote sensing has become widely applied in coastal areas [64,93,94,95]. This parameter was obtained by applying a supervised classification to the Alsat2 image of 2019 using a maximum likelihood (ML) algorithm in ENVI 4.5 software with 10 m of spatial resolution (Figure 8). A field survey conducted between August 2018 and April 2022, as well as Google Earth aerial photos, strongly supported the classification of AlSat2 Image. Ground truth observations and Google Earth Pro were used to validate the classification’s result. The overall accuracy obtained was 95.03% and the kappa coefficient was 0.8875.

Eight classes of land cover were detected in the study area (Figure 8): built-up land, agricultural land, dense vegetation (forest), rough terrain (moderately vegetated), sands and sediments, bare soil, cliffs, and water bodies. Vulnerability values were assigned according to Table 2.

From the LULC obtained, it is possible to conclude that the study area contains a variety of land use patterns. The highest urban density is observed on the eastern part of the Tafna River (Beni Saf district). The chief town of Beni Saf is organized around the fishing port, which is the main factor creating this city by the fishermen community, whereas the western part of Tafna (Oulhaça El Gheraba district) is an area of high relief characterized by weakly urbanized villages.

Due to the high relief characterizing the area, small and medium-sized settlements are organized mostly according to the network of agricultural roads. Agriculture is marked, as it is a primitive sector in this region. Rachgoun, the most important secondary agglomeration expanding at the downstream of Tafna and around the coastline in the flatland (the Tafna embouchure), is subject to sea-related impacts (erosion, submersion, etc.) as well as valley-related influences (flooding of Oued Tafna; pollution and its repercussions). The coastline is mostly dominated by cliffs, with a few pocket beaches mostly made of sand. Urbanization interacting with economic development and population growth may increase the natural vulnerability of the region.

2.

Road network density

Road network density is considered a relevant socioeconomic factor to assess CVI [39,74,75,79,96]. The road network data were vectorized from maps available on the QuickMapServices plugin in QGIS 3.8, and the road density was estimated (Figure 9).

The road density was estimated using ArcGIS software. For this parameter, vulnerability is considered to increase in areas with a high road density, unlike areas with few roads (low density), which are considered to be in the low vulnerability category [24] (Table 2).

3.

Designated conservation areas and cultural heritage

Designated conservation areas and cultural heritage are used as pertinent socioeconomic indicators to assess coastal vulnerability [24,75,97]. Cultural heritage is defined as the number of historical buildings, museums, etc. [98], whereas conservation area status is a way to define the ecological and biological value of land [79,96].

Conservation areas and cultural heritage were considered by digitizing important ecological, archeological, and historical sites and other places using ArcGIS software (Figure 10). Scores assigned to the cultural heritage and designated conservation areas parameter are categorical (Table 2).

For cultural heritage, two classes were prepared based on the criteria of whether human pressure is high or low in this site; the ranking is given as 4 for high human pressure and 2 for low human frequentation. Based on on-site observations and local expert knowledge, the historical site of Siga is ranked as highly vulnerable due to agricultural activity, which conflicts with the preservation of the archeological site (on-site observations), whereas the Syphax site or (Beni Rhenane in the local language) is ranked as weakly vulnerable (low human frequentation). Cultural heritage sites were delimited based on national reports from the National Office for the Management and Exploitation of Protected Cultural Assets, Ain Temouchent Province (NOMEPCA), and fieldwork with archeological experts.

For conservation area status, the study area contains two main sites: Expansion Zone and Tourist Site (ZEST Rachgoun), considered a form of conservation area status declared by the Algerian government in order to limit abusive exploitation of coastal resources. Rachgoun Island was classified as a RAMSAR site and MPA (Figure 10). Conservation area status and cultural heritage parameters are described in

Table 3. Designated conservation area status and cultural heritage description.

	Site Name	Designated Status	Area (ha)	Management Body	Year of Establishment
National designation	Rachgoun Expansion Zone Tourist Site (ZEST)	Tourist Expansion Zone	50	Tourism Department of Ain Temouchent Province www.mta.gov.dz, accessed on 27 Macrch 2022	9 October 2014
	Siga “Numidian kingdom of Syfax”	Archeological and Historical Site	/	Department of Culture of Ain Temouchent Province www.m-culture.gov.dz/, accessed on 27 March 2022	26 September 2007
	Mausoleum of Syphax	Archeological Site	77	Department of Culture of Ain Temouchent Province www.m-culture.gov.dz/, accessed on 27 March 2022 National Office for the Management and Exploitation of Protected Cultural Assets of Ain Temouchent Province www.ogebc.dz, accessed on 27 Macrch 2022	8 May 2016
	Rachgoun Island	Historical Site	/	Department of Culture of Ain Temouchent Province www.m-culture.gov.dz/, accessed on 27 Macrch 2022	19 October 2009
International designation	Rachgoun Island	-RAMSAR Site “Wetland of International Importance” -Future MPA		Forest Department of Ain Temouchent Province http://www.dgf.org.dz/, accessed on 27 Macrch 2022	-5 June 2011 -Study in process launched in 2021 by www.rac-spa.org, accessed on 18 February 2022

. The information is synthesized to give rankings categorically (Table 2).

4.

Population density and touristic density

Population density and touristic density in coastal areas are widely studied topics due to interaction between human activities and coastal ecosystems [26,96,99,100]. In the SES approach, it is required to involve these parameters in coastal planning. Coastal zones are characterized by rapid urbanization and high population density (higher than the national and countryside cities’ averages) [18]. Gornitz et al. and Ariffin et al. [22,26] acknowledged that population density is an important variable to include in vulnerability studies of coastal areas. Population and tourism could exert a negative pressure on the coastal environment [68]. Urban sprawl is mainly due to tourism development. Rachgoun and Beni Saf are the most important tourist destinations in Ain Temouchent Province. The total number of tourists visiting the beaches of the Ain Temouchent Province during the summer period of 2017 reached 17,834,575 people, which is 42 times more than the resident population of Ain Temouchent Province (BPMD, 2017; Civil protection Department, 2017). For that, this study considers the influences of population and tourism as relevant criteria to estimate CVI. The importance of these parameters is correlated with the interdependence and the mutual interaction between coastal ecosystems and human activities. Since population can affect and be affected by coastal areas, it is concluded that population can be interpreted as an “economic” and as a direct “erosion-inducing” variable in the CVI study [96]. It can be considered as an “Economic variable” due to the potential impacts of natural hazards on the population [18,96,101,102,103], and an “erosion-inducing variable” exerting a negative pressure on the coastal system considering human-related alterations and overexploitation of natural resources [18,23,96,104], which leads to a significant increase in environmental degradation.

The number of tourists considered is from before the COVID pandemic, as tourist activity was exceptionally low during the pandemic. A 1 km buffer zone around the sandy beaches (the main tourist destination) was considered for this criterion (Figure 11).

5.

People’s Awareness

People’s Awareness refers to the response of the main question: are the coastal people aware of the vulnerability of their coastal socioecological system? People’s awareness about the coastal ecosystem is a widely studied subject [105,106,107,108,109]. The public awareness indicator is measured by means of a public opinion survey conducted through face-to-face interviews among a representative sample of the adult population (18 years and older). Primary data were collected through interviews, questionnaires, and surveys conducted between August 2018 and April 2022 with various stakeholders (local inhabitants, farmers, fishers, and tourists). Some questions pertaining to awareness were added through Google Forms during the country’s travel bans imposed in response to the COVID-19 pandemic.

Based on the literature [40,98,106,107,108,110,111,112,113], awareness was estimated based on three indicators: (i) familiarity and knowledge concerning coastal issues, which aims to detect whether stockholders understand the consequences and restrictions of their actions towards this specific ecosystem; (ii) environmental perception, which refers to their degree of concern for environmental process and cultural heritage; and (iii) preparedness, which aims to determine if they are prepared for disasters related to natural process. Adapted from relevant studies [105,107,111,114,115,116], the questions were developed by the authors. The estimation awareness of people is 56%, and rates as moderate vulnerability, as shown in (Table 2).

2.2.2. MCDM Method and CVI Estimation

The development of integrated CVI for the CSES requires MCDM, most commonly applied to decision-making issues [117,118]. MCDM is a very widely used approach to solve complex problems. It refers to an integrated approach that considers a certain number of criteria in decision making, i.e., the process that combines competing criteria with quantitative and qualitative outcomes to select the best option [119,120,121,122]. According to [123] MCDM techniques have been classified in a number of ways; multi-objective decision making (MODM) and multi-attribute decision making (MADM) are the most common classifications [122,123,124,125,126,127]. The distinction between MADM and MODM depends on the evaluation of criteria as an attribute and as an objective (an attribute is a parameter that describes any aspect of a given situation, and an objective is the suitable direction for a certain attribute [124]). MADM deals with problems with predetermined alternatives, whereas MODM aims to identify the best choice from an infinite set of alternatives [123,124,126,128].

In the last few decades, MCDM has been considered as one the most effective approach to deal with sustainability and natural resource management issue, which involves ecological, social, and economic dimensions [123,124,129,130]. MCDM methods are frequently employed to address a range of geospatial decision-making issues, related mainly to the four main areas: location analysis; environmental modeling; suitability analysis; and finally, risk and vulnerability assessment [125,131]. MCDM helps decision makers to establish an appropriate policy framework aimed at effective management in line of sustainable development of the region. As coastal areas present multiple conflicting criteria linked to interactions between natural and socioeconomic variables, coastal planning and management is a complex issue in the context of decision making. Hence, several studies have applied MCDM as an integrated framework and analytical tool to assess coastal vulnerability [80,130,132,133,134,135]. For coastal vulnerability assessment, several multi-criteria decision-making (MCDM) techniques can be successfully used. Taherdoost et al. and Tanim et al. [127,130] describe a comprehensive list of MCDM that can be successfully used for vulnerability assessment in coastal zones, such as the analytical hierarchy process (AHP), principal component analysis (PCA), simple additive weighting (SAW), and the weighted product model (WPM), among others. Although not all the tools are attempted to be described in this study, the specific techniques suitable for CVI estimation, which are WPM and SAW, are used [136]. PWM is based on the multiplication of scores of all criteria [137] (Equations (1) and (2)). SAW is the most popular of the MCDM methods due to its simplicity; it involves a simple summation of scores representing the objective considering all parameters [120,130] (Equation (3)).

CVI Estimation 

The parameters’ category led to the selection of two main sub-indices: CVI_N, combining geological and physical parameters; and CVI_S, including socioeconomic parameters. The global ICVI integrates both natural and socioeconomic sub-indices.

In the absence of relevant information on the relative impact of indicators [120], the simple average method (SAM) was chosen, providing an equal weightage to variables to illustrate the interdependency and to comprehend the complexity of coastal vulnerability [120,121]. Each sub-index was calculated by the square root of the product of the ranked variables divided by the total number of variables (Equations (1) and (2)). The global ICVI was calculated as the sum of the two sub-indices (Equation (3)).

$$CV{I}_N=\sqrt{\frac{L\times S\times E\times Hs\times Pc\times Dr\times Sc}{7}}$$

(1)

$$CV{I}_S=\sqrt{\frac{Pd\times Lc\times Rn\times Pa\times Dc\times Ch\times Td}{7}}$$

(2)

$$\mathrm{ICVI}=CV{I}_N+CV{I}_s$$

(3)

where, L is the lithology; S is the slope; E is the elevation; Hs is the significant wave height; Pc is the proximity to coast; Dr is the distance from river; Sc is the shoreline change; Pd is the population density; Lc is the land cover; Rn is the road network; Pa is the people’s awareness; Dc is the designated conservation areas; Ch is the cultural heritage and Td is the touristic density.

2.2.3. Fieldwork

Planning decisions that incorporate local knowledge succeed in achieving the proposal goals [138,139]. Hence, fieldwork conducted between August 2018 and April 2022 played a major role in this study. The results of questionnaires and interviews were analyzed and synthesized to better understand the local dynamics and interactions by combining local and scientific knowledge, as well as to carry out this work using a participatory approach, which is now proven to be essential for sustainable development.

Table 4. Description of the sample of stakeholders.

Category	Number of Interviewees
Fishermen	118
Farmers	109
Local population	74
Tourists	57

resumes the fieldwork performed in this research.

3. Results and Discussion

Fourteen variables, namely lithology, slope, elevation, proximity to coast, distance from river, significant wave height, shoreline change, population density, number of tourists, LULC, road network density, people’s awareness, designated conservation areas, and cultural heritage, were considered for all the 27 km of the coastline of the study area. The results of the analysis were mapped in a GIS environment using 10 m of spatial resolution, projected on WGS84 UTM Zone 30N. Mapping CVI allowed for the identification of vulnerable areas. Three maps were obtained: a CVI_N map (Figure 12), a CVI_S map (Figure 13), and a CVI map (Figure 14).

3.1. Coastal Vulnerability Index Mapping

3.1.1. Natural Coastal Vulnerability Index

Figure 12 presents the result of CVI_N mapping based on seven geological and physical parameters representing the natural component. The CVI_N values varied from 0.37 to 81.83 and it was classified into five categories using natural breaks classification (Jenks). The extremely high and high vulnerability classes are mainly marked in the narrow strip next to the port and in the downstream of the mouth of Tafna Valley at a distance of 3 km. This value is linked to the low coast and sandy beaches (Rachgoun and Wells) where both geological and physical parameters are assigned high and very high vulnerability classes. The lowest elevation, gentle slope, and proximity to coast associated with the lithology type of sandy beaches and dunes, allowing for a shoreline dynamic, are the main parameters explaining this value. The vulnerability values decrease clearly, going upstream of the Tafna Valley and moving away from both the banks (right and left). Along the Tafna Valley, the vulnerability ranks at a moderate level due to the combination of three key parameters: slope, elevation, and distance from the river. The same rank is assigned around Rachgoun Island (moderate vulnerability) as result of low elevation and shoreline proximity combined with the potential impact of waves. These values decrease clearly moving away from the banks of Tafna Valley. This finding is favorable due to relevant parameters such as high elevation, solid lithology, and distance to the coastline and from the river, which increase going upstream of the Tafna Valley and moving away from the two banks. In addition to these two main parameters explaining the low vulnerability value, the lack of impact of shoreline dynamics and waves, are both believed to be very low due to the distance between this area and the coastline, as they only affect the strip closest to the coast.

3.1.2. Socioeconomic Coastal Vulnerability Index

CVI_S was mapped based on seven main parameters: LULC, population density, number of tourists, people’s awareness, road network density, designated conservation areas, and cultural heritage. CVI_S registered values from 1 to 17.66 classified into five classes according to natural breaks classification (Figure 13). The extremely high vulnerability and high vulnerability classes were found mainly around low coasts (sandy beaches) where socioeconomic stakes have a considerable weight. This value is associated with impacts related to socioeconomic activities combining urban development, a dense road network, and tourism pressure. The designated conservation area status and cultural heritage are the main parameters contributing to these findings. Rachgoun Island, which has internationally designated status as a future MPA and RAMSAR site, is located in the moderate vulnerability area. Rachgoun’s ZEST is found in the extremely high vulnerability class, whereas archeological sites (Siga and Beni Rhenane) are in the moderate vulnerability area. The vulnerability class decreases with decreasing human activities, where it is found that areas of moderate vulnerability along the coast are mainly associated with socioeconomic parameters assigned to low and moderate vulnerability classes, such as touristic activity, road network density, and population density. Outside the narrow coastal strip, the moderate vulnerability class is mainly correlated to the scattered settlements and road network. Low vulnerability areas are generally located in zones with little socioeconomic interest. Very low vulnerability areas are found in untouched zones characterized by their steep, difficult to access topography. Hence, the high concentration of socioeconomic activities on the narrow strip of the coast. The polled population showed an average awareness, which balanced the results of CVI_S; more awareness will lead to low socioeconomic vulnerability and vice versa.

3.1.3. Global Integrated Coastal Vulnerability Index

The ICVI is the composition of CVI_N and the CVI_S. The analysis performed revealed that the two main urban centers located on the coast of the Ain Temouchent Province (Beni Saf and Rachgoun) are found in areas of high to extreme vulnerability (Figure 14). These low coasts are mainly exposed to various natural and anthropogenic pressures where both the natural and socioeconomic sub-index are ranked in the highest vulnerability classes. The two agglomerations host most of the coastal population, with an average population density of 814 hab/km² (Department of Construction and Urban Planning, 2020). Socioeconomic development in this area is mainly linked to the seaside and tourist activities. The number of tourists marked was about 10,111,220 between June and September of 2017 (Department of Civil Protection, Ain Temouchent Province) representing 56.69% of the total number of tourists on all the beaches of Ain Temouchent Province, hence the almost permanent threat of risks linked to anthropogenic pressure. On the other hand, these areas, located in a narrow coastal strip, are susceptible to multiple risks related to flooding, submersion, pollution, erosion, and their repercussions, which include loss of life and properties [140]. It turns out that some habitations were abandoned due to the constant risk of erosion (Figure 15), indicating extremely high vulnerability. Our study reveals regression of the shoreline (estimated from 0.01 m to 2.55 m/year) on average by 37 years in the exploited low coast. Settlements in the Rachgoun agglomeration are mainly built on sedimentary lithology (sand of beach and ancient dune). Photos taken during the field trip (2019 and 2022) show some affected coasts by the shoreline dynamic (Figure 15). Among the 74 surveyed residents in Rachgoun, Madrid, and Wells agglomerations, 90% of them confirmed suffering from humidity-related problems affecting health (respiratory diseases) and properties (construction cracks and rusting of metal equipment). The field survey reveals that 53% of the local population noted beach regression compared with the past and 39% noted landslides along the coast, while 47% of residents on Tafna’s two banks complained of pollution-related damage, including insect bites and stings (which cause skin allergies), and a foul smell caused by human waste along the Tafna Valley, which accumulates in its mouth. Human health has been harmed because of this pollution [82]. According to our survey with the Department of Epidemiology and Preventive Medicine, beach pollution was reported several times due to pollutants (lagoon wastewater and household waste) spilled by Oued Tafna at Rachgoun beach, by unauthorized discharges from hotels near Wells beach, and by water that escapes from sewage networks. This was based on site visit sewage observed at the “Sidi Boucif” beach, reported as prohibited for swimming due to pollution.

The CVI registers a high value at a distance of 3 km from the coast, and along a 500 m stretch of the Tafna River, resulting from a combination of natural and socioeconomic variables. Natural criteria (low altitude, gentle slope, and proximity to the coast and to the Tafna Valley with a fragile recent lithological formation of low wadi terraces) interacting with socioeconomical activities, including touristic activity, the type of urban occupation scattered along the narrow strip of the coastline (600 m from the Tafna mouth on the left bank and 300 m on the right bank), and agricultural activity, generated high global vulnerability value. Going a little away from the coast and remaining on a narrow strip of the Tafna Valley, the vulnerability is classified as moderate due to moderate vulnerability factors. Proximity to the Tafna Valley, the low altitude, and the sedimentary lithological type are the main natural factors, while anthropogenic activity is associated with agriculture (as the dominant activity), according to which small settlements are distributed. Based on field surveys, 61% of the farmers are concerned about the impact of marine intrusion on irrigation water and soil quality. For Rachgoun Island, the main natural factors are elevation and wave impact, whereas socioeconomic drivers are the designated conservation area status and the high impact of touristic activity. All the polled fishers (100%) confirmed the decrease in fishing stocks, and 73% of them link this decrease to the impact of pollution caused by human activities along the coast.

Coastal and marine areas can produce a set of services that are useful to human societies, essentially linked to the exploitation of natural resources (fish) and the regulation of the environment (water quality, pollination), or even constituting cultural services (recreation, natural heritage) (ocean-climate.org, (accessed on 1 November 2022). Going inland reduces human pressure in the absence of activities that depend on the maritime façade. In this study, the majority of the region’s economic activities, such as tourism, fishing, and urbanization, are concentrated in a narrow strip along the lower coasts (Rachgoun, Madrid, and Wells), which explains the evolution of the global ICVI. Socioeconomic factors multiply natural vulnerability; hence, the further we move away from these low coasts, the lower the vulnerability values.

3.2. Vulnerability Relative Impact

By the use of the two sub-indices, it is possible to evaluate the degree of human activity’s impact on natural aspects and then to implement adequate spatial planning strategies for the region. Based on the two sub-indices (CVI_N and CVI_S), a vulnerability relative impact diagram was developed (Figure 16). It consists of determining integration requirement degree according to characterization and the vulnerability class of each area. The purpose is to demonstrate that complexity in coastal areas can be well managed with good information about all component of the coastal system (natural and socioeconomic sides), which leads to efficient blue planning of the coastal zone based on integration process. Four areas have been identified: A, B, C, and D. A (low vulnerability area) includes low natural and socioeconomic vulnerability. It corresponds to a coherent ecosystem with good functioning, allowing for the sustainable use of coastal resources. In this area, socioeconomic activities are few and/or well integrated with the natural system. This range represents the suitable situation where the integration process is well applied or there is no integration requirement in case of no threatening human activity; B (moderate vulnerability) results from low natural vulnerability and high socioeconomic vulnerability, where human activities are developed affecting the natural side of the area. This category needs more concern about the natural side, and requires improvement of the integration process in order to maintain the resilience of this system. Integration needs to be improved in this area, requiring adaptive management based on a participatory approach; C (high vulnerability) combines the high natural vulnerability with the few anthropogenic activities. It refers to zones where human activities are developed slowly, mainly because of difficult natural conditions. It is a potential area for development of sustainable socioeconomic activities with a high integration requirement. D (very high vulnerability) results in natural vulnerable areas with dense human activity, amplifying the vulnerability degree. This area presents a risk for environmental and socioeconomic stakes. Urgent integration action is needed in order to restore/preserve the coastal ecosystem and prevent any future deterioration while allowing for the sustainable use of coastal resources.

In our case study, despite the vulnerable natural characteristics in the narrow coastal strip and the downstream of the mouth of Tafna Valley, development of the most significant socioeconomic activities is accelerating. For example, urbanization is still growing close to the river and coast, where the natural characteristics are the most vulnerable. In addition to touristic activity, the narrow strip hosts the most populated urban agglomerations (Beni Saf and Rachgoun). The Cumulative impacts amplify the vulnerability degree and threaten environmental and socioeconomic stakes. Therefore, urgent integration action is needed for sustainable development scenarios. Despite the adoption of the national strategy of ICZM in Algeria, it is still difficult to adopt on the ground, and decisions related to coastal development are still fragmented. The implementation of ICZM involves multiscale integration, among related sectors; in this context, the proposed tool could enhance the effectiveness of ICZM plans.

The outcome of the discussion offers prominent learned lessons and managerial insights on how to sustainably manage CSES.

Theoretically, the current study supports policy and coastal management from several points of view:

(i) Integrated, long-term coastal management should combine social and physical management processes rather than being just focused on the physical ones. (ii) It helps coastal managers and planners to identify vulnerable regions (at scale of 10 m), providing information about impacts of selected criteria on the coastal zone. In return, it enhances the capacity of decision makers to establish appropriate coastal management strategies. (iii) Findings can support the assessment of integrated coastal vulnerability, providing a comprehensive database and maps to enable planners to prioritize their stakes related to the coastal zone and assess combined effects under various idealized alternatives in order to choose the best plan. The evaluation of the impacts of the changes against different management scenarios may reduce the possibility of unsustainable management decisions. (iv) The developed GIS-MCDM database supports long term planning while allowing for the incorporation of future scenarios. It allows decision makers to analyze future scenarios by incorporating potential changes after the establishment of the new MPA of Rachgoun. (v) The adoption of CSES (bottom-up approach) helps to broaden local perspectives for multiscale management problems. In other words, it provides suitable ground for raising stakeholders’ awareness and participation. It improves participation and knowledge exchange about the different managerial activities, relevant interacting factors, and their influence on the intervention coastal zone. (vi) The development of the ICVI map, through selecting and ranking the relevant parameters that multiply the vulnerability of the region, enhances environmental monitoring and management, while establishing coastal management strategies. (vii) The ICVI incorporates conflicting criteria and enhances analysis of multiscale conflict. The integration of fieldwork and local stakeholder’s experiences (fishers, agricultures, and inhabitants) in the analysis of the results improves integration actions and defines a concrete vision for coastal planning considering local specificities. More specifically, the developed index offers quantification of the integrated coastal vulnerability, and crossed with field data from relevant stakeholders, improves current knowledge. Therefore, it helps in establishing an effective management vision.

4. Conclusions

The CSES is defined by its interacting natural and socioeconomic characteristics, hence the specificity of each coastal zone. Efficient management of coastal areas requires the adoption of integrated approaches considering the local characteristics of natural and human components as well as their interactions in order to implement integrated management strategies for sustainable development. This study develops an integrated coastal vulnerability index (ICVI) for the coastal area of Rachgoun, Northwest Algeria, using an MCDM based on an integrated approach. A multihazard vulnerability assessment involves the actual interactions on the coastal zone without depending on one process without the other. Coastal vulnerability is expressed by integrating two main components: natural and socioeconomic. Fourteen parameters were chosen for this study: seven to represent the natural component (lithology, elevation, slope, significant wave height, shoreline change, proximity to coast, and distance from river) and seven for the socioeconomic component (LULC, road network density, population density, touristic density, people’s awareness, and designated conservation areas and cultural heritage).

The results reveal that the narrow band of less than 3 km located mainly around the urban areas of Rachgoun and Beni Saf is the most vulnerable area. Its natural vulnerability is mainly related to the sedimentary lithological type, the low altitude, and the proximity to the coast and to Oued Tafna, while touristic activity and urbanization are the most influential socioeconomic factors. Therefore, the two sub-indices are important determinants for the assessment of coastal vulnerability based on a socioecological approach.

This study relies on the integration of natural and socioeconomic components to determine the area of interest, where the coastal area is not limited to the narrow strip of land–sea contact, but rather the coastal system defined by its natural and human components in mutual interaction describing the CSES. Two sub-indices representative of natural vulnerability zones and social vulnerability zones were developed, the integration of which gives the CVI. The mapping results were crossed with field data to better discuss the study results with the field realities.

This multidisciplinary study serves as a decision support tool for the implementation of sustainable spatial planning and management strategies of the Ain Temouchent region. It can constitute a reference study for an efficient classification of “Rachgoun” marine protected area, and for effective blue planning of Ain Temouchent Coast to avoid bad development scenarios experienced by big Algerian coastal cities, such as Oran and Algiers, whose repercussions spiral out of control.