The coastal ecosystems of the northwestern Arabian Gulf mostly consist of low lying tidal flats backed by hyper saline marsh (sabkha) and desert. As populations increased, much of this coast has been subject to infill and seaward extension for housing, recreation and industrial development. This has led to degradation of the highly productive intertidal and shallow subtidal marine ecosystems upon which fisheries depend. Further degradation can be expected with future sea level rise, as high evaporation rates will induce hyper saline conditions.
Rather than building out into the sea using dredging and infill, an alternative approach is to construct cities on land, safely above future high water levels, by excavating saline desert of low productivity to create marine waterways. Sabah Al-Ahmad Sea City in Kuwait was first connected to the sea in 2004 and, with 3 phases completed, now has 84 km of shoreline in the desert. The waterway system was modeled before excavation and relies solely upon tides and wind for seawater circulation. To date seawater quality has been excellent, meeting all criteria demanded by the Environment Public Authority.
Within the waterways, a full range of man-made marine habitats have been created. These include intertidal beaches, tidal flats, mangrove and salt marshes on islands, together with subtidal sand and rock benthic habitats. These have been monitored with daily physical and annual biological surveys. Importantly, this data provides new information on the natural colonization rates and development of Gulf soft substrate intertidal and subtidal marine communities. Within 5 years of opening to the sea, all artificially created marine habitats have a species richness and abundance close to, or exceeding, that of similar open sea natural habitats in Kuwait. Over 1000 species of macrobiota now exist within the desert waterways including 100 species of fish and shellfish. Present work describes the development of these marine communities and provides a baseline for recruitment rates of Gulf marine fauna into non-polluted habitats.
The shallow sea forming the Gulf is situated in a subtropical arid zone, bordered by oil and gas rich states undergoing rapid population expansion. The low sloping Arabian coast line is mostly sedimentary and is subject to intense development, despite the potential for flooding due to climate change driven sea level rise (SLR) (NASA; flood map, 2010). Natural climatic stresses include high temperatures and salinity, but western Gulf coastal ecosystems are also degraded by a wide range of anthropomorphic impacts (Sale et al., 2011). Apart from oil spills and other discharges, the most serious impacts are caused by dredging and infill for residential, recreation and industrial developments (see Sheppard et al., 2010 for detailed review).
The marine ecology of Gulf coastal ecosystems is now relatively well known (Sheppard et al., 1992; Jones et al., 2002) and has been recently summarized by Sheppard et al. (2010) and Sale et al. (2011). All these authors recognize the high ecological value and productivity of intertidal and shallow subtidal habitats including tidal mud and sand flats, salt marshes, mangroves and subtidal sea grass beds.
In contrast to most recent Gulf coastal developments, created by infilling ecologically valuable inshore marine habitats, Sabah Al-Ahmad Sea City, for 100,000 people, was designed to be built on low lying coastal desert in the south of Kuwait. This in effect brought the sea to the desert rather than the reverse (Figure 1). As the best estimate for SLR, this century has a rise between 190 and 590 mm (Bindoff and Willebrand, 2007); much of this low lying coastal sabkha is destined to be flooded by the sea during the next 100 years. With elevated air temperatures, any flooded areas of desert will become hyper saline due to evaporation which is similar to the Khors Al Ama and Khor Al Mufateh (Clayton, 1983) upon which the city is centred. These khors already have a low ecological value due to high salinity (Jones et al., 2007); hence flooding due to SLR is likely to have a further negative impact on this already stressed coastal ecosystem.
The city extends the original khors by excavating sabkha desert material to produce a series of interconnected channels 200 m wide and 3–4 m deep on average (Figure 1). Excavated material is impacted and forms the basis for housing, services and infrastructure which is safely situated above future high water levels with an elevation of +4 to +5 m KLD (Kuwait Land Datum). Groynes and marinas are lower to provide access to water, but are at +3.2m above KLD. Site design takes full advantage of both wind and tide to provide energy free natural flushing (Ealey et al., 2002) and tidal gates have been introduced in phase A3, which is some 5–6 km from the open sea. The original khors reached over 5 km inland with salinities in excess of 100 psu. A1, the first of 5 phases, was opened to the sea in 2004 and with the opening of phase A3 in 2009, some 84 km of shore line have now been completed (Figure 1).
The environmental challenge has been an attempt to create artificial man-made intertidal marine habitats, including sand beaches, mangrove and salt marsh, together with rock revetments. Subtidally artificial fine, medium sand and rock benthic habitats have been constructed. When the city is completed these new marine habitats will extend for some 30 km2 in the desert and compensate for most of the 20 km2 of open sea coastal habitat lost in Kuwait by unrelated reclamation projects during 1975–1994 (Bishop, 1999). As the waterways are semi enclosed (Figure 1), they can be protected from oil spills and unregulated fishing, providing nursery grounds for fish, shellfish and other wildlife.
Since the initiation of the project daily water quality measurements have been taken, together with annual ecological surveys. Results have been reported annually to the Environment Public Authority (EPA) and preliminary results published (Al-Jamali et al., 2005; Jones et al., 2007; Sheppard et al., 2010). Present work demonstrates that it is possible to recreate functional intertidal and subtidal marine ecosystems based on artificial substrata, and more importantly provides the first estimates of the time scale required for colonization by marine communities. This data provides a baseline for the recovery dynamics and rates of non-polluted soft substrate communities in the Gulf, and compliments the work by Burt et al. (2011) for hard substrate communities.
Materials and Methods
Physiochemical data collection
Water quality measurements have been made fortnightly from stations within phase A1 by Kuwait Institute of Scientific Research (KISR) since 2003. These ceased in 2008 when this phase passed to the developers, but continued in phase A2 which was flooded in 2006. Measurements included temperature and salinity, total suspended solids, chemical and biological oxygen demand, nitrite, nitrate ammonia and silica levels. Quarterly, the waterways are analysed for total organic carbon, trace metals and faecal streptococci; all results are reported directly to EPA. In addition, La'Ala, Al-Kuwait Real Estate Co. KSC conducts daily water monitoring (salinity, temperature, dissolved oxygen and pH) from all phases for quality control.
Biological data collection
Initially, environmental monitoring of phase A1 waterways was conducted by Coastal Science and Engineering in 2004. Beginning in 2005, annual marine ecological surveys have been conducted for all marine habitats using methodologies recommended by PERSGA (2004). These include taking 25×25×15 cm deep core samples in triplicate at mid and low tide levels on all shore habitats, and sieving to retain macrofauna greater than 0.5 mm. Subtidally triplicate VanVeen grab samples are taken, together with triplicate Ocklemann sledge samples for epibiota. These methods are similar to those used in the United Nations Claim Commission (UNCC) surveys in both Kuwait and Saudi Arabia (UNCC, 2005) hence data bases are comparable.
Surveys collect quantitative data on the biota colonizing sand beaches, mudflats, mangroves salt marshes on islands, rocky areas, together with soft and hard subtidal benthos in all phases. (See Figure 1 for sampling sites.) In addition, the water column is sampled for phytoplankton (monthly), ichthyoplankton (monthly) and finfish and shellfish with traps and nets (annually). A monthly bird census and quarterly underwater video and photography are taken using SCUBA. Further details are given in Jones et al. (2007). In view of the large number of samples taken for analysis, phase A1 and A2 are surveyed in alternate years, although a few permanent transects are monitored annually. Over 150 samples are analysed every year, and results compared using PRIMERv 5 (2001).
Water quality measurements
Despite the removal of islands in Phase A2 in 2008, all water quality measurements have remained similar to those recorded for the open sea and within the limits set by EPA over the period 2004–2010. This confirms the predictions by the original flushing model (Ealey et al., 2002) and demonstrates that even in a high evaporative climate, seawater of bathing quality can be achieved in a semi enclosed desert environment.
Artificial marine habitats
The current shoreline extends for 84 km throughout the first 3 phases, and most is artificially created sand beaches. These are composed of quarried marine sand, washed to remove fines and held in place by rock groynes. Table 1 shows the recruitment of biota larger than 0.5 mm into these beaches over time. In phase A1 mean species richness has risen from 3.5 ± 0.7 to 31.5 ± 3.1 in 2008, with an abundance of 3982 ± 292 m−2. Phase A2 shows a similar, but more rapid increase, in both diversity and abundance over time. Phase A3, sampled just 8 months after flooding already equals both the diversity and abundance found in open sea natural sand beaches (Table 1). The UNCC data is taken from a survey of 13 beaches along the coast of Kuwait using similar methodology (n = 26).
The ratios for dominant macro faunal taxa for Sea City beaches were initially 60.91% Polychaeta and a low diversity for Mollusca (0–12%). However, after 5 years ratios in phase A1 were 41% Polychaeta, 9.1% Mollusca, 31% Crustacea, and after 4 years, ratios in phase A2 were 44% Polychaeta, 18% Mollusca and 18% Crustacea.
Mangroves and salt marshes
Areas of islands in phase A1 have been planted with Avicennia marina grown in an onsite nursery from genetically selected seed (Maguire et al., 2000). After initial problems with low winter temperatures and correct positioning, mangroves were established with a mean height of over 50 cm flowering and fruiting in 2010. The development of macrobiota associated with these mangroves is summarized in Table 2 which showed that by 2010 species richness reached 20 ± 2.02 and abundance reached 1488 ± 512 m−2. Salt marsh, mainly Halocnemon strobilaceum, was already present in the project area and had been transplanted onto the phase A1 islands. Initially survival was poor, but once the correct tidal height was established by constructing micro channels, plants thrived and recruited naturally.
Table 3 shows that recruitment rose from a single pioneer species Ceratonereis sp. in 2005 to a community of 15.6 ± 2.0 species in 2010 with an abundance of 240 ± 48.5 m−2. The new salt marsh now includes most of the key species Nasima, Macrophthalmus, Cerithidea, found in natural marshes.
These currently comprise 8.5 km2 of excavated channels with fine sandy benthic sediments. In all phases part of the benthos was lined with rocks to promote settlement of attached epibiota. Table 4 shows recruitment of soft benthic biota since waterways were connected to the sea. Within 4 years, phase A1 exceeded both the diversity and abundance found in open sea benthos. Phase A2 followed a similar pattern with mean abundance exceeding that of the open sea within 2 years. However, although species diversity was following the phase A1 recruitment curve, removal of the islands in 2008 from phase A2 has temporarily reduced the rate of recruitment (Table 4). In all phases initial high abundances of opportunistic species declined as diversity increased. Settlement of macroalgae, Pinctada beds and coral species enhanced benthic biodiversity to over 395 macro species with 175 species occurring at a single mixed rock and sand benthos station in phase A1 in 2010.
In phase A1 the ratios between dominant taxa settling have not changed greatly over the 6 year period since the channels were dredged. Polychaeta formed 59%, Mollusca 23% and Crustacea 13% in 2004 and 52%, 24% and 15% in 2010. In 2006, A2 polychaetes dominated with 90%, molluscs were absent and crustaceans formed 10%. However, the survey was conducted only 6 months after this phase was opened to the sea, and only 10 benthic species were collected (Table 4). By 2009 the ratios were 43% polychaetes, 31% molluscs and 12% crustacean taxa. In phase A3 in 2010, 9 months after flooding, polychaetes again dominated with 66%, molluscs formed 14% and crustaceans 11%; only 33 species were present in total with a mean of 23 per station (Table 4).
By 2010 the total macrobiota present throughout the waterways exceeded 1000 species, including some new to science (Jones and Nithyanandan, in press).
Annual diving, seine netting and trapping surveys have revealed a total of 90 species of fish within the waterways of which 80 are commercial species. In addition 4 species of penaeid shrimp, Portunus pelagicus and Sepia species inhabit the waterway system, Ichthyoplankton surveys show species are breeding or entering the system as post larvae. The rise in trap catches in successive years, together with the presence of fish eggs and larvae, indicated that the waterways act as spawning, nursery and feeding grounds for the open Gulf.
As all sewage water is treated on site and does not enter the waterways, pollution is minimized. Both phase A1 and 2 waterways are now lined with residential housing, but occupancy appears to have had minimal impact as marine diversities continue to rise in both phases. Boat traffic has risen dramatically with jet skis, power and sailing boats. This has been beneficial, creating waves and preventing settlement of fine particulates on the sand beaches. A marine laboratory is maintained on site and conservation information is available for residents and visiting groups. Guided tours for students from British schools were conducted during 2009 to create awareness.
The key to the successful colonization by marine life in the Sea City is the high quality of seawater circulating throughout the waterways. Original computer models developed by Buro Happold were tested in physical models by Kuwait Institute of Marine Science before construction commenced (Ealey et al., 2002). These allow for future SLR with higher tides channeling excess water to back of site marshes which rapidly drain back to the sea using circulation based on wind and tidal energy. Phase A3, some 3–4 km from the open sea, has tidal gates enhancing flow to areas remote from the sea.
The construction of man-made intertidal habitats, beaches, tidal flats, mangrove, salt marsh and rock revetments provides a unique opportunity to measure colonization dynamics and time scales for marine biota settling into pristine habitats. Previous estimates of recovery time for Gulf marine soft substrate biota have been based upon oil impacted shores and extend over 5–14 years (Jones et al., 1996).
Table 1 indicates that colonization by marine biota into sand beaches may reach the species richness and abundance levels found on natural shores within 1 to 4 years. As phase A3 receives water from A2, it is likely that sand beach fauna already established in A2 provided the seed for phase A3. Initial composition of taxa settling on beaches is dominated by polychaetes and crustaceans with few molluscs. However, dominance ratios change over time with polychaetes reducing to 42%, molluscs rising to 14% and crustaceans 24% after 4 years. This compares with natural open sea beach ratios of 32% polychaetes, 29% molluscs and 24% crustaceans (UNCC, 2005 n = 320). In general, molluscs often dominate on Gulf beaches (Basson et al., 1977; Jones, 1986) which are more exposed to wave action.
Mangroves and salt marshes are more difficult to grow as tidal level, exposure and sediment characteristics are critical; however, once established they rapidly attract key biota. Table 2 shows that mangroves reached the species richness and abundance found in natural northwest Gulf mangroves (Jones et al., 2002) within 5 years. Salt marsh colonization by macrobiota was more rapid in reaching levels found on natural marshes within 3 years (Table 3).
Table 4 demonstrates that species richness and abundance levels, comparable to those seen in natural benthic habitats in Kuwait, were reached within 4 years of opening phase A1 waterways. Abundances in all three phases exceed natural benthic levels within a year due to large scale settlement of a few opportunistic species such as the polychaetes Ceratonereis, Lumbrinereis and crustaceans Amphithoe and Apseudes. These dominate the taxa initially, but after 4–5 years ratios are closer to those seen in open sea benthos (Richmond, 1996; Al-Yamani et al., 2009; Joydas et al., 2011). Jones et al. (1996) also found initial recovery of oiled benthos was dominated by polychaetes.
The Sea City design has allowed the development of coastal habitation, without fear of future flooding, in a high quality recreational marine environment. It is both economically and environmentally beneficial to utilize desert sabkha for construction rather than infilling and dredging biologically productive intertidal and subtidal habitats. Present work demonstrates that it is possible to create man-made intertidal and subtidal marine habitats which are colonized by marine biota found in natural habitats within a period of 3–4 years. Annual marine surveys reveal important information on the dynamics of community development for a range of coastal habitats, and at the same time, demonstrate that it is possible to create extensive conservation areas for marine wildlife and fisheries.
We dedicate this paper in memory of Khalid Yousef Al-Marzouk, the vision behind this project.