An overview of integrated hydro-ecological studies in the MELMARINA Project: monitoring and modelling coastal lagoons—making management tools for aquatic resources in North Africa
- First Online:
- Cite this article as:
- Flower, R.J. & Thompson, J.R. Hydrobiologia (2009) 622: 3. doi:10.1007/s10750-008-9674-8
- 281 Downloads
As landscape disturbance and climate conspire to accelerate global environmental change towards unprecedented levels in the twenty-first century, the populated coastal regions of many countries are facing major threats to sustainability. Coastal water resources are particularly vulnerable in dry regions. In view of the expected severity of future environmental change in the Southern Mediterranean Region, the European Commission supported an integrated multidisciplinary project, MELMARINA, on monitoring and modelling coastal lagoons in Morocco, Tunisia and Egypt. This is a region where water management for people and for agriculture has been intense particularly during the twentieth century, yet long-term environmental monitoring and management of wetland ecosystems are under developed. Not only are biodiversity aspects at risk in coastal lagoons and wetlands but the goods and services that affect human welfare are also generally in decline. Co-ordinated hydro-ecological monitoring at key wetland lagoons was begun in 2003 with a view to establishing environmental baselines and calibrating site-specified hydro-ecological models. This article introduces the project and its results that range from lagoon typification and hydro-ecology to the application of hydro-ecological models. Detailed results and evaluations are presented in a linked series of themed scientific articles within this special issue. The present condition of the lagoons investigated essentially results from various hydrological modifications combined with eutrophication problems, yet all still remain valuable aquatic ecosystems. Adequate monitoring data are an essential part of reliable predictive modelling and, despite several data gaps, nutrient load reduction scenarios were undertaken to help target restoration aims. Implementation of aspects of the EU Water Framework Directive for achieving good ecological status of transitional waters is advocated. Nevertheless, as the twenty-first century advances the effects of global climate change are expected to amplify current stresses making intervention restoration and adaptation management even more imperative. Long-term sustainability depends upon detecting and measuring environmental change (long-term water quality and ecological quality) and incorporating the results into appropriate hydro-ecological models to facilitate the development of appropriate management initiatives.
KeywordsCoastal lagoons North Africa Mediterranean Hydro-ecological monitoring Modelling Environmental change
Reconciling environmental sustainability with human population growth and economic development is a challenging global problem. As the twenty-first century begins, growing demands for water and other resources, pollution, land-use intensification, together with climate change are all increasingly threatening the sustainability of natural resources in many countries (Tilman et al., 2001; Ragab & Prudhomme, 2002). This is particularly so for freshwater resources in most developing countries where pressing human needs often overwhelm other environmental considerations (Allan, 2000). In such countries, wetlands are amongst the most vulnerable natural ecosystems (Finlayson & Moser, 1992; Dugan, 1993; Zalidis et al., 2002), and problems of freshwater availability and land development are epitomised in North Africa. Across the Southern Mediterranean Region (SMR), water resources are under extreme and growing pressure (Hollis, 1992; Biswas, 1993; Sultan et al., 1999; Allan, 2000) with important implications for wetland resources (e.g. Finlayson et al., 1992, Smart, 2002). Nowhere is the loss rate of natural wetlands and associated water bodies more alarming than at the coastal margins.
Recent work points to hydrological disturbance, water pollution and land reclamation as the primary causes of coastal wetland resource loss around the Mediterranean (Flower, 2001; Özhan, 2005; Perez-Ruzafa et al., 2005). Many sites listed in wetland inventories for North Africa (Kerambrun, 1986; Hughes et al., 1997; Green et al., 2002) no longer exist or are severely degraded. Some remaining wetland lagoons nonetheless support high-value ecosystems that are not only important resources for local human populations (e.g. Benessaiah & Belhaj, 1999) but also contribute substantially to regional biodiversity (Pearse, 1996; Hughes et al., 1997; Chergiu et al., 1999; Khedr & Lovett-Doust, 2000; Flower, 2001; Green et al., 2002). The values of these sites are often recognised by international bodies (for example, the Ramsar Convention and MedWet, Farinha et al., 1996) and by national legislation (e.g. Tunisia’s Forest Act 1988). However, effective management of coastal wetland resources for sustainability requires reliable integrated information about current site status, past and projected changes in human usage and a variety of other environmental change processes. Such information is generally not currently available for North African wetland lakes and lagoons. The need to acquire this information is heightened given projected climate change. Precipitation is predicted to decline around the Mediterranean during the twenty-first century (e.g. Arnell, 2004). For example, it has been suggested that a substantial area will experience declines greater than 40 mm year−1 (or 17%) in successive 50-year periods and, as a result, the region is likely to experience acute water stress (Shindell, 2007). In addition, it has been suggested that the Mediterranean will experience some of the largest sea level rise-induced losses of coastal wetlands (Nicholls & Hoozemans, 1996; Nicholls et al., 1999; Nicholls, 2004).
Most North African coastal lagoons are strongly exploited but few are used sustainably (Flower, 2001). The MELMARINA (monitoring and modelling coastal lagoons: making management tools for aquatic resources in North Africa) Project was conceived with the ambitious aim of establishing integrated hydrological and ecological monitoring at selected lagoons in three North African countries (Morocco, Tunisia and Egypt). A major aim was to improve understanding of the functioning of these ecosystems and to ascertain the impacts of environmental and management changes through hydro-ecological modelling. As such, the project comprised a multi-partner consortium formed with the objective of applying internationally integrated research to North African lagoonal systems in order to promote potentially sustainable management options for these aquatic resources. This article provides an overview of the project activities including site surveys, lagoon ecosystem monitoring and field instrumentation, palaeo-environmental studies and remote sensing. Together with the collection of secondary data, these activities have facilitated the establishment of environmental baselines. They have also provided data to develop hydro-ecological models which can make major contributions to lagoon management when parameterised and calibrated with sufficient site-specific data (e.g. Rasmussen et al., 2000; Gönenc & Wolfin, 2005). Such models can facilitate the simulation of future environmental change scenarios involving alternative management interventions, climate change and sea level rise.
The MELMARINA Project and its objectives
The implementation of integrated monitoring programmes at selected lagoons to establish space-time changes in hydrological and ecological characteristics and to introduce automated measurement and best practice procedures (with reference to the EU Water Framework Directive, Directive 2000/60/EC).
The establishment and monitoring of water resources and the extent of aquatic vegetation throughout each lagoon using the combination of remotely sensing imagery and field surveys.
The identification of key environmental variables controlling vegetation characteristics and water quality changes.
The development of dynamic hydro-ecological models using state-of-the-art model codes and the calibration of these models using data from the monitoring programme.
The simulation of site-specific environmental change scenarios using these models.
The development of integrated time series and GIS databases for the MELMARINA lagoons and making these data and the results of model simulations available to end-users, especially to national management agencies.
The MELMARINA consortium
Environmental Change Research Centre/Wetland Research Unit, UCL Department of Geography
DHI-Water and Environment
Institut Scientifique, Universite Mohamed V, Rabat
Institut National des Science et Technologies de la Mer, Salambo
Department of Geography, Faculte des Lettres et des Science Humaines de Sousse
National Authority for Remote Sensing and Space Science, Cairo
Rationale and methodology
Following mutual agreements on research protocols (Flower et al., 2003), specific tasks were assigned to institutions within the MELMARINA consortium. Each North African institution undertook specialist tasks and was responsible for the day-to-day operation of the monitoring programme (including hydrology, water quality, plankton, fish and aquatic vegetation) and the collection of existing data and other documentary information. Specialist tasks included zooplankton and plant analyses for Merja Zerga and Ghar El Melh (ISRabat), fish diversity and growth rates (INSTM), land-use analysis (FLSHS) and remote sensing (NARSS). UCL was responsible for analytical quality control, sediment coring and analyses, hydrological analyses, database management and GIS. They also acted as advisors to institutions undertaken monitoring activities. DHI-WE’s main responsibility was the hydro-ecological modelling work packages.
The primary results of the MELMARINA Project are summarised in Flower & Thompson (2006). This special issue draws on these results and comprises a suite of themed articles each dealing with the main results of specific aspects of the MELMARINA Project.
Site bathymetries, site characteristics and land use issues
The three primary lagoons (Merja Zerga, Ghar El Melh and Lake Manzala) are strongly contrasted according to many of their physical and biological characteristics. Ayache et al. (2009) present an overview of the three sites. All are shallow water bodies connected to the sea by at least one well-defined channel and, except in these channels, water depths are less than 2 m. The lagoons vary in size from less than 20 km2 (Merja Zerga) to ca. 700 km2 (Lake Manzala), and they are also different in terms of hydrological characteristics with Merja Zerga being flushed daily by the relatively large Atlantic tides whereas Ghar El Melh and Lake Manzala, being confluent with the Mediterranean, are much less affected by tidal cycles (see Thompson et al., 2009). All three lagoons occur at the termini of inland drainage basins of varying size. Merja Zerga receives freshwater from two inflows, one of which (the Nador Canal) is artificial being created as a result of major drainage works. Inflows from both sources are highly seasonal. Freshwater inflows to Ghar El Melh are also seasonal but are smaller unless the Mejerda River, which was historically diverted away from the lagoon, overspills its banks. Many of the small streams still draining to the site have been dammed reducing their already modest flows (Ayache et al., 2009). Consequently, both Merja Zerga and Ghar El Melh are dominated by marine water, especially in summer. On the other hand, Lake Manzala is predominantly freshwater, since it receives large year-round inflows of Nile River water via agricultural and domestic drainage. All three lagoons are delineated from the sea by sand bars, dunes and spits but the landscape on the inland margins of each has become strongly agricultural, especially during the latter part of the twentieth century, with intensive crop production for cereals, vegetables and fruit. The impacts of these activities are most vividly demonstrated around the southern margins of Lake Manzala where extensive areas of lake bed have been reclaimed for agriculture.
During the first phase of MELMARINA, lagoon surveys were carried out and although activities focussed on the three primary lagoons, secondary sites were also visited to investigate surficial sediments through grab sampling and sediment coring (Fig. 3). In addition to bathymetric measurements, spatial assessment of sediment quality was undertaken at each primary lagoon and Flower et al. (2009) describe the variations found. Bioclastic material (mainly mollusc shell fragments) was most common in central areas of the lagoons where oxygenated marine water predominated and where silts were common. Sands predominated in the deeper marine channels where benthic biomass was low. Freshwater mollusc remains were only common in surficial sediment from the most western parts of Lake Manzala. Sediment cores from Merja Zerga indicated that below about 30 cm depth (either in the mainly clays and silts of the southern part of the lagoon or in the more sandy sediment in the north) marine shells were abundant and in the northern part they made up the bulk of the sediment. This was similar to the secondary site Lagoon Nador where shell debris constituted the sediment bulk. In cores from Ghar El Melh and Lake Manzala, shells (usually Cerastoderma) were occasionally present. In contrast, grab surveys of surface sediments contained few living molluscs indicating major and recent declines in marine shellfish abundance. A shift to more freshwater conditions (Birks et al., 2001; Flower, 2001) is probably a factor in Lake Manzala, but elsewhere eutrophication and oxygen depletion are suspected as the cause of shellfish decline.
Sediment cores were used for radio-isotope dating so that sediment accumulation rates could be calculated (Flower et al., 2009). Despite dating problems caused by low rain washout of the atmospheric radionuclide 210Pb (see also Appleby et al., 2001), 137Cs fallout usually provided a good marker for the 1963 sediment horizon in the lagoon sediments. Surface sediment accumulation rates (SARs) were similar in the primary lagoons at around 0.7–0.8 cm year−1. Past changes in SAR were also indicated but dating was restricted to about the past 50 years. Ghar El Melh and the Nile Delta lakes including Manzala typically showed a reduced SAR in the most recent period, whilst Merja Zerga showed little change and several secondary sites showed increased SARs. Reduced SAR probably indicated the diminished effects of floods and sediment inwash, either as a result of regulation of the Nile (for Lake Manzala) or impounding inflowing streams thereby compounding the earlier impacts of diversion of the Mejerda River (for Ghar El Melh). Irrespective of past changes in SAR, the current sediment accumulation rates indicate that considerable quantities of sediment are retained in the primary lagoons. Sediments of approximately between 2 and 5 kg (dry mass) m−2 year−1 are stored in each lagoon. The lowest dry weight sediment accumulation occurred in Lake Manzala and this reflects a relatively higher proportion of organic sedimentary material from within lake productivity. Consequently, and despite the strong similarity of contemporary sediment accumulation rates in the primary lagoons, sediment dynamics appear to be quite different between and within sites and are related to both within lagoon and within drainage basin processes as well as to the magnitude of sea water incursions.
MELMARINA hydro-ecological monitoring was undertaken in two ways: by employing in situ instrumentation to record hydrological and some water chemistry changes and by undertaking regular field sampling for water quality and ecology over 15 months during 2003–2004. Additional data for this period, most notably meteorological station and tide gauge records, were obtained from appropriate secondary sources (Fig. 3). Thompson et al. (2009) describe the field instrumentation and monitoring activities with an emphasis on the hydrological work. Several digital water level recorders were installed in each lagoon using supporting frameworks designed according to local site conditions. In addition, one conductivity, temperature and pressure (CTD) logger was employed at each site with its location being rotated between monitoring stations. As backup, water level readings from stage boards were also made on monthly monitoring visits and using local observers. Water samples (for water chemistry and for plankton) from several strategically selected monitoring stations within each lagoon were acquired on each visit. Unsurprisingly, large daily changes in water level were recorded at Merja Zerga due to the influence of Atlantic tides and these were much less at the two Mediterranean sites. Water level changes at all sites generally diminished with distance from the sea connections. Freshwater inflows were highly seasonal at Merja Zerga and Ghar El Melh being dependant on rainfall over the inland catchments. Therefore, except in the winter months, freshwater inflows to these sites were characterised by very low or no flow. At Lake Manzala, however, regulation of the Nile provided by upstream impoundments, most notably the High Aswan Dam, and the development of perennial irrigation in the Nile Delta have greatly diminished the seasonality of freshwater inflows (e.g. Stanley & Warne, 1993; Randazzo et al., 1998). The balance of freshwater inflows and tidal cycles largely regulates the salinities of each lagoon so that summer water levels are maintained by seawater at Merja Zerga and Ghar El Melh. At the latter site, because tidal flushing is relatively small, hypersalinity develops in these months. Lake Manzala displays a persistent and strong salinity gradient where the southern and western parts are essentially freshwater with higher salinities in the north where sea water enters the lake.
Phytoplankton communities in coastal lagoons are of major importance for food web structures and ecosystem health. However, water quality changes, especially through eutrophication, are major problems for coastal lagoons influenced by human activity. These changes can alter the species composition of plankton communities but precise species changes are difficult to predict. Plankton monitoring of the three MELMARINA primary lagoons is described by Ramdani et al. (2009). Phytoplankton numbers are shown to fluctuate markedly with decreased numbers during November–January but with the spring increase occurring around March. The highest cell number was found during the summer period when relatively larger individuals were dominant. The summer increase at Merja Zerga was attributed to the increase in cell numbers of Pseudonitzschia spp. and Pleurosigma elongatum which reached 24,400 cells l−1. Diatoms and dinoflagellates comprised the bulk of the crops in Merja Zerga and Ghar El Melh and a total of 211 phytoplankton taxa were identified. Only occasionally were harmful toxin producing dinoflagellates recorded during the 2003–2004 monitoring period. As in most lakes and lagoons, changes in phytoplankton species were related to nutrient levels rather that to salinity changes.
Seasonal vegetation surveys were undertaken at each of the primary lagoons (Flower & Thompson, 2006). Results from the transects undertaken at the margins of Merja Zerga and Ghar El Melh revealed the dominance of ruderal plants indicating the disturbed nature of the landscape. A variety of marginal plants was present with little seasonal succession. Within lagoon vegetation at Lake Manzala comprised a complex mosaic of plants dominated by emergent species (Phragmites and Typha). Submerged aquatic plants (most notably Ceratophyllum and Potamogeton) were also common in the less saline and less polluted parts of the lake in particular those in the west of the site. Vegetation data were compared with environmental data obtained from the hydro-ecological monitoring programme in order to identify the key controlling variables affecting the distribution of vegetation. Ordination analysis showed that water salinity exerted the strongest influence upon vegetation distribution with water depth and nutrients having more localised influence.
Kraïem et al. (2009) describe the sampling of fish communities within the three MELMARINA primary lagoons. Lake Manzala supports a markedly different fishery from that in Ghar El Melh and Merja Zerga. Its ichthyofauna consisted essentially of freshwater species (especially tilapia) some of which are characteristic of the Nile system. Furthermore, Manzala supports by far the most important fishery with an annual catch of around 8,000 tonnes. In comparison, fish production in Merja Zerga and especially Ghar El Melh is small and is declining. The exploited fishery within Merja Zerga mainly comprises two species of mullet (Mugil cephalus and Liza ramada) and eels (Anguillaanguilla) together with shellfish including Cerastoderma and Ruditapes. The Ghar El Melh fishery is very restricted by the characteristics of the site: summer hypersalinity develops in the central and western parts and water quality problems due to nutrient enrichment occur particularly in the area of Ghar El Melh town (the largest population centre on the lagoon shore, Ayache et al., 2009). Such problems are not so manifest at Merja Zerga because of the flushing effect of the large Atlantic tides. Fish growth rates and condition show small differences for mullet in the three lagoons reflecting the ecological state during the last 10 years. As with catch decline, these are associated with disturbed sea connections, increasing eutrophication and intensification of human activities generally. Remarkably, however, overall fisheries production in Lake Manzala has remained relatively stable in the last few decades despite major changes in fish species and widespread pollution and intense fishing activity.
The use of remotely sensed data for monitoring of wetlands has significant potential, especially in large sites where traditional field techniques are problematical and are unable to provide large spatial coverage (e.g. Hess et al., 2003; Cózar et al., 2005). When combined with ground surveys aimed at calibrating the spectral signatures of different wetland plant communities, remote sensing techniques provide an exceptional means of distinguishing spatial vegetation cover as well as determining the extent of open water and marginal disturbance (e.g. Schmidt & Skidmore, 2003; Castañeda & Herrero, 2005; Maheu-Giroux & de Blois, 2005; Fig. 3). Remote sensing approaches can also be employed to investigate water quality issues within water bodies. Ahmed et al. (2009) apply remote sensing techniques to the three MELMARINA primary lagoons and discuss the issues of geometric correction and image classification, primarily to distinguish vegetation patterns. Of the three sites, Lake Manzala has by far the most extensive cover of aquatic macrophytes and these plant communities displayed 13 different spectral patterns which were categorised into different groups, depending on species (most plant communities were dominated by emergent plants Phragmites and Typha or by the floating Eichhornia). A particularly valuable aspect of remote sensing imagery is that there is now approximately 35 years of data available enabling assessments of long-term changes in the nature of sites to be investigated. For Lake Manzala, examination of the older data is used by Ahmed et al. (2009) to track the reduction in the area of open water as the southern margin of the lake has been reclaimed. They also show significant fluctuations in the area of emergent vegetation. Similarly, Ayache et al. (2009) employ a combination of historical maps, aerial photographs and satellite imagery to track large-scale landscape evolution at Merja Zerga and Ghar El Melh. The expansion of the delta within the first of these sites which is associated with the Nador Canal is clearly identified, whilst a seaward movement of the sand bars separating Ghar El Melh from the Mediterranean is revealed.
Two further articles (Rasmussen et al., 2009a, b) describe the development of coupled hydrodynamic–ecological models for two of the primary lagoons (Ghar El Melh and Lake Manzala). Hydrodynamic models were developed using the MIKE 21 finite element modelling systems (Fig. 3). These models were parameterised using available historical data and the results of field survey and monitoring programmes. Parameterisation included the specification of lagoon bathymetry, upstream (freshwater inflows) and downstream (tidal curves) boundary conditions, wind speed and direction and evaporation. Hydrodynamic models were primarily calibrated through comparisons with salinity and observed water levels within the lagoons. Subsequently, the ECO Lab model (DHI-WE) was linked to these hydrodynamic models in order to simulate water quality. This necessitated further parameterisation including the specification of water quality at the upstream and downstream boundaries. Simulated water chemistry was subsequently compared to field data, whilst biomass distribution simulated by ECO Lab was also compared to vegetation surveys. Rasmussen et al. (2009a, b) describe how model results are able to highlight specific aspects of the functioning of the primary lagoons including the relative importance of freshwater inflows and exchanges to the sea and their influence upon salinity, circulation patterns and sources and distribution of nutrients within the lagoons. A range of scenarios and their simulation developed for each lagoon involving site-specific impacts are subsequently discussed and their results evaluated. It is, for example, shown that a 25% reduction in nutrient load to Ghar El Melh coupled with similar declines in suspended sediment will reduce macroalgae and increase transparency permitting the re-establishment of extensive Ruppia beds which have declined in recent years (Rasmussen et al., 2009b). Increasing filtration by bivalves, the populations of which have also declined over the past decades, is shown to enhance this recovery. Scenarios investigated using the model of Lake Manzala (Rasmussen et al., 2009a) focussed on the impacts of reducing nutrient loading with particular reference to water quality and the restoration of aquatic vegetation.
Coastal lagoons of the SMR are valuable economic and ecological resources, yet they are also highly impacted by human activities which include hydrological modification, pollution and habitat loss. Pressures on these environments are set to amplify during the twenty-first century as demands for water continue to grow whilst the impacts of climate change and sea level rise will increase. The long-term sustainability of coastal lagoons will depend upon improved understanding of ecosystem functioning and the ability to detect and measure environmental change. The development of management plans for coastal lagoons which maintain or improve environmental quality whilst also enhancing the well-being of local human populations can benefit from hydro-ecological models capable of representing the often complex hydrodynamic and ecological processes operating within them. This form of modelling will be important in the development of mitigation measures in the face of climate change and sea level rise. By addressing all these issues and identifying a series of key issues for environmental science and management of coastal lagoons in the SMR (see Thompson & Flower, 2009), the MELMARINA Project has provided the foundations for the sustainable management of these aquatic resources.
The MELMARINA Project was financed by the EU Framework V INCO-Med Programme (Grant ICA3-CT2002-10009). The authors of this article, who coordinated the project, thank the other members of the MELMARINA team for their contributions which were instrumental in the success of the project. We gratefully acknowledge the support provided by the UCL Department of Geography and in particular Maria Rodriguez.