Vertical Accretion and Relative Sea Level Rise in the Ebro Delta Wetlands (Catalonia, Spain)
- First Online:
- Cite this article as:
- Ibáñez, C., Sharpe, P.J., Day, J.W. et al. Wetlands (2010) 30: 979. doi:10.1007/s13157-010-0092-0
- 240 Views
The Ebro Delta in Catalonia, Spain is an ecologically and commercially important wetland system under threat from sea level rise and marsh subsidence. Our principal hypothesis was that a brackish marsh that receives inorganic sediments and fresh water amendments from the Ebro River would exhibit significantly higher rates of soil accretion, resulting in a greater resistance to subsidence and sea level rise compared to isolated salt marsh habitats with no river subsidy. Marsh sites representative of the wetland ecosystems found in the Ebro Delta were selected based on plant community type, porewater salinity, and landscape position. The results supported the research hypothesis, suggesting that a brackish marsh that receives river subsidies exhibited a significantly higher (F3,4 = 31.6, P < 0.01) rate of vertical accretion compared to more hydrologically-isolated salt marsh systems. Accretion data showed that only the riverine-influenced brackish marsh site met the minimum predicted rate of relative sea level rise (RSLR range of 5–8 mm yr−1) for the Ebro Delta. This research provides the first quantitative record of marsh subsidence and accretion dynamics in the Ebro Delta using Surface Elevation Tables (SET), marker horizons, and 210Pb techniques, and also illustrates the importance of sediment and fresh water subsidies in deltaic environments.
KeywordsMarsh accretionSubsidenceSurface elevation table
Deltas such as the Ebro lie at the interface between terrestrial and coastal zones and support a myriad of ecosystem functions and economic values. The low elevation and abundance of water and nutrients often create ideal habitat for wetland ecosystems; these wetlands exhibit high primary productivity and assist in important functions such as nutrient removal and sediment retention. In addition, deltaic systems also serve as critical habitat for many species of wildlife and fisheries (Prat and Ibáñez 1995). The most recent provisional count (2008) of bird species in the Ebro Delta found that over 292,600 birds (91 individual species) overwinter in the delta.
Currently the Ebro Delta is undergoing coastal retreat at the river mouth because wave erosion and elevation loss in the delta are no longer offset by new sediments coming from the Ebro River. Sediment has been obstructed by the construction of an extensive systems of dams (n = 170) upriver of the delta that retain as much as 99% of the river sediment (Ibáñez et al. 1996a). The sediment deficit in the delta created by the dams, coupled with land subsidence (vertical sinking) from compaction, organic soil decomposition, accelerated sea level rise, and the already low elevation of the delta plain, puts the Delta and its wetlands at major risk for submergence, salt intrusions, and coastal erosion (Ibáñez and Prat 2003).
Marsh elevation relative to sea level is a function of numerous processes, such as eustatic sea level rise, sediment inputs, soil compaction, organic matter decomposition, subsidence, and vertical soil accretion, occurring at several time scales (Day et al. 1995, 1998; Mudd et al. 2009). Eustatic sea level rise (ESLR) has increased at a rate of 1–2 mm yr−1 over the last century and has further increased over the past 10 to 15 years to between 3.0 and 3.5 mm yr−1 (FitzGerald et al. 2008). Sea level rise is expected to accelerate over the next 100 years (IPCC 2007), and could reach one meter by 2100 (Rahmsdorf 2007; Pfeffer et al. 2008). In addition, subsidence has caused relative sea level rise (RSLR) to be much greater than the eustatic rate, especially in wetlands associated with deltas. For example, in the Mississippi Delta, relative sea level rise (RSLR) is about 1 cm yr−1, primarily due to tectonic subsidence and compaction of Holocene substrata (Dokka 2006; Törnqvist et al. 2008). In the Po Delta, ground water withdrawal from the 1940s to the late 1960s led to subsidence as high as 30 cm yr−1 (Carbognin and Tosi 2002).
Estimates of RSLR rates utilizing historic and current survey data from backshore flats of different ages from the La Banya spit in the Ebro Delta (Fig. 1) indicate mean RSLR rates ranging from 2.08 mm yr−1 over 132 years (1965–1833) to 6.26 mm yr−1 over 31 years (1965–1934) (Ibáñez et al. 1996b). To prevent excessive water-logging of wetlands, vertical accretion needs to keep pace with the local combined effects of eustacy and subsidence. Vertical marsh accretion depends on both mineral sediment inputs from river- or wind-derived flood waters and local organic matter production from plants (La Peyre et al. 2009).
Several studies have attempted to predict the fate of selected coastal wetlands subject to accelerated ESLR rates by comparing current and predicted rates of RSLR to measured rates of sediment accretion, and then calculating an accretion deficit, surplus, or balance (Bricker-Urso et al. 1989; Day et al. 1999; Pont et al. 2002; Morris et al. 2002; Reyes et al. 2004; Kirwan and Murray 2007). Several methods have been used to measure sedimentation and vertical accretion in wetlands. Near-surface horizon marker methods often do not span the time scale of the shallow subsidence processes affecting long-term accretion, such as decomposition and primary consolidation, resulting in an overestimate of this parameter (Cahoon et al. 1995). Therefore, coupling short term accretion measures (i.e., kaolinite marker horizons) with longer term sediment dating, such as 210Pb that integrates decomposition and compaction processes occurring within the first meter of sediment, can help to resolve this problem of time scale (DeLaune et al. 2003; Neubauer 2008). As an alternative to measuring accretion directly, the Surface Elevation Table (SET) method has been used to integrate both accretion and shallow subsidence over several meters, and it can be used to measure changes in marsh elevation over time (Boumans and Day 1993; Cahoon et al. 2002). To determine if wetlands are growing vertically at a rate sufficient to offset water level rise, measurements of vertical accretion alone are insufficient. Measurements must also be made of the rate of vertical elevation change because of shallow subsidence occurring in the upper soil profile. Shallow subsidence is defined as the difference between vertical accretion and surface elevation change (Day et al. 1999).
One objective of this study was to perform the first long-term study of sediment accretion/subsidence dynamics within two wetland types representative of typical coastal marsh communities commonly found in the Ebro Delta. Emergent marshes of the Ebro Delta can be classified by their salinity regime into the following types: salt (18–32 ppt), brackish (5–18 ppt), oligohaline (0.5–5 ppt), and fresh marshes (0–0.5 ppt). These marshes and their adjacent land areas within the Ebro Delta have undergone extensive human alteration, beginning in 1860 with the construction of the first irrigation canal from the Ebro River into the Delta for rice production (Rovira and Ibáñez 2007). From 1860 until today, the Ebro Delta has gradually been converted from mixed salt/brackish/fresh marsh and swamp communities to primarily rice agriculture (Cardoch et al. 2002). With the exception of the Garxal lagoon at the mouth of the Ebro River and salt marsh communities located along the outermost edges of the delta, an extensive system of canals and pumping stations has effectively placed every marsh habitat under strict hydrologic control, thus isolating the majority of these systems from the Ebro River and Mediterranean Sea, and eliminating most of the fresh marsh habitat.
The critical question at the heart of this research was—do deltaic marshes that receive regular sediment and freshwater river subsidies possess a greater ability for vertical marsh growth and are they more resistant to accelerated sea level rise than nearby salt marshes with no direct river connection? The principal hypothesis was that sediment accretion in the marsh communities possessing a combination of organic material accretion and Ebro River sediment inputs (i.e., the brackish marsh at Garxal lagoon) would demonstrate the greatest rate of vertical marsh growth. Possessing information regarding which landforms are more resistant to sea level rise and land subsidence will allow resource managers in the Delta to track and better predict the potential evolution of these different habitats, and develop effective management solutions for preserving these ecosystems.
Study Site Descriptions
Four study sites were selected; one site representing riverine-associated brackish marsh habitat, and three representing typical salt marshes isolated from the Ebro River and occupying different landscape positions (i.e., low, intermediate, and high elevation). As a result of past and present anthropogenic influences, only the brackish marshes of Garxal lagoon at the Ebro River mouth receive any of the inorganic river sediments historically associated with the majority of wetland habitats in the Ebro Delta. Therefore, the Garxal marsh (Garxal) was selected as the representative study site of riverine-associated brackish marsh habitat in the Delta. The three salt marsh sites chosen included two sites at Buda Island (Buda Backshore and Buda Lagoon) and one site at the abandoned Migjorn River channel (Migjorn). The Buda Backshore, Buda Lagoon, and the Migjorn marsh sites each possessed different landscape positions (see description below) and thus potentially different accretion dynamics that necessitated examining the three salt marsh subtypes. The main vegetative and hydrogeomorphic features of the four sampling sites are described below (for more detail see Curcó et al. 2002).
Salt Marshes (Buda Backshore, Buda Lagoon, and Migjorn)
The Buda Backshore marsh site was located in a marine-influenced backshore area of Buda Island (Fig. 1). Plant communities at Buda Backshore were dominated by Arthrocnemum glaucum (Delile) Ung.-Sterb. and displayed a low degree of cover (10–20%) and mean vegetation height. The Buda Backshore marsh was also situated at an intermediate relative elevation compared to the three salt marsh sites examined during this study. The Buda Lagoon marsh was dominated by a community of Sarcocornia fruticosa (L.) A.J. Scott that covered nearly 100% of the land surface area. This marsh area was situated at the lowest relative elevation of the three salt marsh sites. The Migjorn marsh was located just behind the beach/dune system, close to the mouth of the abandoned Migjorn River channel (Fig. 1) and was situated at the highest relative elevation of the three salt marsh sites examined. The plant community was dominated by Sarcocornia fruticosa. The two dominant species in the salt marsh sites, Sarcocornia fruticosa and Arthrocnemum glaucum often co-occur in the Ebro Delta, therefore, we believe the salt marsh plant communities were relatively equivalent.
Brackish Marsh (Garxal)
The Garxal brackish marsh site bordered a lagoon area directly influenced by river discharge and therefore receives a periodic influx of fresh water, nutrients, and sediments (Fig. 1). This marsh was formed over the last five decades as a result of the most recent change of the river mouth. The shallow lagoon is partially opened to the river and is separated from the sea by a continuous sand barrier. Along the south edge of the lagoon, there is a belt of brackish marshes dominated by Phragmites australis (Cav) Steudelbut with an occurrence of Scirpus maritimus L. The sampling plots were located in the Phragmites/Scirpus community.
Soil Granulometry, 210Pb Vertical Accretion, and Chemical Analysis
To provide a general idea of soil granulometry and soil chemistry differences that may impact accretion processes, soil cores were taken from one random location inside the Buda Lagoon (salt marsh) and Garxal (brackish marsh). The Buda Lagoon salt marsh site was chosen as the representative sample site for the soil analysis as it possessed soil characteristics indicative of the Buda Backshore and Migjorn marshes, and thus was representative of a “typical” salt marsh in the Ebro Delta. Soil cores were collected to a depth of 20–56 cm (depending on soil marsh thickness) with a cylindrical PVC corer of 11.5 cm internal diameter. To improve the efficiency of core extraction, the top of the corer was sealed with a screw-top before extracting the sample from the sediment. To attain a sufficient weight of soil for analysis, especially in the more organic layers, composite samples were made from several replicates. Cores were sliced in 2–5 cm layers (depending on soil depth) of known volume, weighted to determine wet weight, and dried to a constant weight at 60°C. Soil bulk density and water content were calculated from these data. Samples were washed by hand through a 2 mm sieve and homogenized mechanically for 8 h. Soil texture was determined by Robinson’s method (Page et al. 1982), except for sandy soils, where the method described in Dupuis (1969) was used. The following particle size classes were measured: sand (diameter between 2 and 0.05 mm), silt (0.05 mm < d < 0.002 mm), and clay (d < 0.002 mm). In samples with more than 35% of organic matter content it was not possible to do granulometric analysis. Total carbon and nitrogen content (expressed as % carbon and nitrogen) was determined following Page et al. (1982) using a Carlo Erba NA 1500 analyzer. Organic matter content was measured by loss on ignition at 500°C for 12 h.
Long-term accretion estimates were also obtained from these cores using measurements of 210Pb along the soil profile (Edgington et al. 1991) from both the Garxal and Buda Lagoon marshes. Long-term accretion rates (210Pb dating) were measured based on the vertical distribution of 210Pb in soil cores following the methodology of Radakovitch et al. (1999). 210Pb activity profiles were estimated by gamma spectrophotometry analyses from subsamples of the same material used in the soil analysis from Garxal and Buda Lagoon marshes. The samples were homogenized and placed in 65 cm3 Petri dishes or 560 cm3 Marinelli receptacles, depending on the availability of material. Samples were counted in the CRII-RAD Laboratory (France) using a gamma hyperpure germanium “N” type detector (EGG/ORTEC, Type GMX) coupled to a multichannel analyzer (type NUCLEUS) and calibrated by a pitchblende gamma multiray source. Samples were counted for 20 to 90 h.
Vertical Accretion (Kaolinite Marker Horizons) and Wetland Surface Elevation Change
Average accretion rate and elevation change using marker horizons and surface elevation tables (SETs) from the replicate plots (A and B) from each of the brackish and salt marsh types examined within the Ebro Delta. Values were calculated using SAS version 9.1
Elevation Change (mm yr−1 ± SE)a
Vertical Accretion (mm yr−1 ± SE)a
Period of Record (years)b
Shallow Subsidence (mm yr−1 ± SE)a
Dominant Plant Species
Brackish Marsh (With River Connectivity)
6.61 ± 2.36
5.03 ± 0.33
Plot A −9.5yr Plot B −3yr
1.57 ± 2.41
Salt Marsh (No River Connectivity—Intermediate elevation)
4.89 ± 2.36
1.32 ± 0.33
Plot A −9.5yr Plot B −1 yr
3.57 ± 2.41
Salt Marsh (No River Connectivity—Low elevation)
4.02 ± 2.36
1.74 ± 0.33
Plot A −1 yr Plot B −1 yr
2.30 ± 2.41
Salt Marsh (No River Connectivity—High elevation)
1.36 ± 2.36
0.89 ± 0.33
Plot A −3 yr Plot B −3 yr
0.47 ± 2.41
Marsh surface elevation data collected from the SETs were averaged for each of four fixed positions and then averaged across all four positions to obtain one average value of elevation change at each SET location (n = 2 SETs per marsh site). The resulting mean was subtracted from the initial reading taken after SET installation to obtain the elevation change. Elevation change values at each point in time were then regressed using Sigma Plot version 10.0 (Systat Software, Inc., San Jose, CA). The resulting slope of the regression equation provided a relative rate of elevation change at each SET location. The resulting rates for each plot (n = 2) at each marsh site (i.e., Buda Backshore, Buda Lagoon, Migjorn, and Garxal) were averaged to capture any within site variability in subsidence and accretion and analyzed using ANOVA analysis of fixed effects (Type III test) in SAS version 9.1 (SAS Institute, Cary, N.C.). This analysis was utilized to determine if any significant differences among average accretion rate, average rate of elevation change, and average rate of shallow subsidence was evident for any of the sites using α = 0.05 (SAS Institute, Inc., Cary, NC). The marker horizon data from each of the plots within the four marsh types were also averaged and used in conjunction with the SET data to determine shallow subsidence.
This equation assumes a constant sedimentation rate and a constant input of exogenous 210Pb over time, therefore, the decrease of 210Pb can be primarily attributed to the radioactive decay of 210Pb over the time series.
The soil of the brackish marsh had an organic upper layer (0–12 cm), with organic matter contents higher than 35%, that prevented the measurement of particle sizes. Layers deeper than 12 cm showed an increase in sand content with depth, with maximum values between 70% and 80% while the silt and clay fractions decreased (Fig. 5). Bulk density increased gradually from 0.09 g cm−3 at the soil surface to 0.56 g cm−3 at about 20 cm and then rapidly to 1.82 g cm−3 at 30 cm. Organic matter, total nitrogen and total organic carbon contents were highest at the soil surface (73.4, 1.97, and 35.1%, respectively) and decreased to low values (3.48, 0.06, and 5.59%, respectively) at the bottom of the core, which were similar to the most sandy layers at the salt marsh site.
Our analysis shows a clear difference in marsh accretion rates between the brackish and salt marshes examined in this study. Although only receiving a fraction (1%) of the historic sediment subsidy from the Ebro River (Rovira and Ibáñez 2007), the brackish marsh exhibited the highest rate of marsh accretion. This can be attributed to the connectivity between the marsh, the Ebro River, and the Mediterranean Sea. The position of the brackish marsh at Garxal allows the deposition of inorganic sediments during high river flow and marine storm events. Nutrients in river water also contribute to high rates of organic soil formation. Similar results have been found in other deltaic wetlands such as the Mississippi (Baumann et al. 1984; DeLaune et al. 2003), Rhône Deltas (Hensel et al. 1999), and Venice lagoon (Day et al. 1999). The inorganic sediments not only provide a nutrient subsidy but also help buffer the marsh against subsidence and sea level rise. Thus, mineral materials may be more important in driving vertical accretion in freshwater marshes than has been reported for salt marshes (Neubauer 2008).
In addition to the sediment subsidy, the lower salinity conditions created by the periodic flux of fresh water into the marshes promote the accumulation of organic material as salinity and sulfate concentrations are kept low, thereby reducing the degree of sulfate reduction and subsequent rapid organic carbon decomposition. The low organic matter content observed in the typical salt marsh soil profile compared to the brackish marsh soils in Fig. 4 supports this assertion. The moderate salinity of these marshes also promotes the growth of plant species like Phragmites australis that aid marsh accretion through dense root zone development and generally low rates of decomposition. The riverine inputs of sediment from the Ebro River to the Delta have been drastically reduced during the last few decades due to dam construction (Ibáñez et al. 1997); this is a general problem of most Mediterranean and world deltas (Day et al. 1995; Ericson et al. 2006; Day et al. 2007; Blum and Roberts 2009; Syvitski et al. 2009). In addition, the construction of infrastructure such as artificial levees, dikes, canals, and roads have all cut off sediment inputs to most wetlands in the Ebro and other deltas. Impoundments such as these are exacerbating the impacts of relative sea level rise on deltaic systems (Bryant and Chabreck 1998; Pont et al. 2002). Decreased accretion of fluvial sediment resulting from upstream sediment capture in artificial impoundments and consumptive losses of runoff from irrigation are the primary determinants of RSLR in nearly 70% of the deltas (Ericson et al. 2006; Syvitski et al. 2009).
The salt marshes of this study possessed no hydrologic connectivity with the Ebro River, and even though they displayed positive mean vertical accretion rates, these rates are insufficient to keep pace with the IPCC projected rate of sea level rise or the predicted RSLR rates in the Ebro Delta. The hydrologically-isolated nature of the salt marshes in this study meant that they received little inorganic sediment and fresh water inputs (two factors we believe to be critical in maintaining marsh elevation). This inference is further supported by the soil chemistry/granulometry results that found lower organic matter content, higher bulk density, lower nitrogen, and lower carbon content in the salt marsh surface soils compared to the riverine-associated brackish marsh. Hydrologically-isolated systems like the salt marsh habitats of this study are thus subjected to high salinity and high organic matter decomposition rates that reduce the amount of marsh soil accretion, making them highly vulnerable to sea level rise impacts such as salt intrusion and water logging, leading to low marsh primary production and eventually wetland deterioration (Day et al. 1995).
Using the 2007 IPCC projection (IPCC 2007) of a mean ESLR of 3.1 mm yr−1 and a mean subsidence rate ranging from 2 mm yr−1 in the central parts of the Delta (Ibáñez et al. 1997) to 6 mm yr−1 in the most active depositional areas near the sea (this study), the estimated RSLR rate for the Ebro Delta wetlands likely ranges from 5 to 8 mm yr−1, as a general estimate. Previous studies assumed a mean value of at least 3 mm yr−1 for the whole deltaic plain (Ibañez et al. 1997), and the inorganic sediment deficit was estimated to be approximately 1,300,000 m3 yr−1 (Ibañez et al. 1996a). However, the results of this study suggest a higher sediment deficit that will be exacerbated as the rate of sea level rise accelerates during the present century. Under this scenario and considering the rates of RSLR in comparison to the measured rates of marsh accretion and elevation change, all of the wetland habitats within the Ebro Delta will be adversely affected. These changes are likely to come in the form of gradual submergence and, in the case of the fresh and brackish marshes, conversion to higher salinity marshes, resulting in the removal of the dominant plant Phragmites australis. The salt marshes will likely convert to open water and beach habitat due to subsidence, coastal erosion, and hyper saline conditions. Recent publications suggest that sea level rise will likely be a meter or more by 2100, indicating that the problems of the Ebro Delta will be even more severe (Rahmsdorf 2007; FitzGerald et al. 2008; Pfeffer et al. 2008).
Proposed solutions to mitigate the effects of RSLR on coastal wetlands and promote the recovery of the marsh structure and functioning range from classical engineering approaches based on protection structures (dikes) to new ecological engineering approaches based on restoring the sediment fluxes to the coast (Rovira and Ibáñez 2007) and reintroducing river input to the Delta (i.e., Day et al. 2007). However, under a scenario of climate change and increasing energy scarcity, the approaches based on heavy infrastructure interfering with natural fluxes of water, sediment, and nutrients (rather than in using them in a controlled way) will be less feasible. A more effective approach would be to promote ecological engineering schemes based on the recovery and management of riverine sediment inputs, like those being planned and implemented in some parts of the Mississippi Delta (Day et al. 2007, 2009).