The ecological quality status of the Elbe estuary. A comparative approach on different benthic biotic indices applied to a highly modified estuary

Ecological Indicators 19 (2012) 118–129

Contents lists available at ScienceDirect

Ecological Indicators journal homepage: www.elsevier.com/locate/ecolind

The ecological quality status of the Elbe estuary. A comparative approach on different benthic biotic indices applied to a highly modified estuary Markus A. Wetzel a,b,∗ , Peter C. von der Ohe c , Werner Manz b , Jochen H.E. Koop a,b , Dierk-Steffen Wahrendorf a a

Department of Animal Ecology, German Federal Institute of Hydrology – BfG, Am Mainzer Tor 1, 56068 Koblenz, Germany Institute for Integrated Natural Sciences, University Koblenz – Landau, Universitätsstrasse 1, 56070 Koblenz, Germany c Department of Effect-Directed Analysis, UFZ Helmholtz Centre for Environmental Research, Permoserstraße 15, 04318 Leipzig, Germany b

a r t i c l e

i n f o

Keywords: Sediment contamination Water framework directive Benthic indicators W-statistic BOPA BO2A AMBI M-AMBI

a b s t r a c t Indices to assess the ecological status of water bodies according to the European Water Framework Directive (WFD) frequently produce widely differing results when applied to estuarine systems. Although several ecological indices have been employed to coastal environments and in estuaries in particular, there is still a lack of knowledge about their suitability for assessing the ecological status of heavily modified water bodies. Thus, we evaluated the performances of indices and fauna parameters (AMBI, MAMBI, BOPA, BO2A, W-value, Shannon diversity, species richness, abundance) that have been discussed in the WFD context using data on invertebrates dwelling in two typical morphological units: the navigation channel and the river bank habitats of Elbe estuary (Germany). In addition, we tested their ability to identify several environmental factors (grain size distribution and chemical sediment contamination). All indices were able to detect major changes in macrofauna composition along the estuarine salinity gradient and were able to differentiate between navigation channel and shallow bank habitats. A strong significant correlation was found with most indices with the exceptions of the W-value and the BOPA with mean grain size. Almost all indices signaled poor ecological quality in the coarser fairway sediments against the finer sublitoral bank sediments. However, AMBI and BOPA showed the opposite: both indicators classified the invertebrate assemblages from the navigation channel better compare than the shallower habitats. The correlation of ecological indices and parameters with sediment contaminants and the toxicity of the sediment calculated as toxic units showed a diverse picture: all indices, except species richness and the BOPA, had a certain significant correlation with several individual sediment pollutants, however, only one index, the W-value, was correlated significant with the majority of chemical pollutants (Pb, Cd, Cu, Ni, Hg, Zn, ␤-HCH, pp -DDD, and TBT) and the toxic units. Our results show clearly that ecological quality classification of heavily modified estuaries depends strongly on both the index and the habitat. Thus, we conclude that no index should be used on its own to estimate the ecological quality of estuaries. Further investigations and the improvement or development of such indices should place emphasis on their independence from the grain size spectrum of the sediments and on their good correlation with its pollution status. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction The establishment of the European Water Framework Directive (Directive 2000/60/EC of the European Parliament and of the Council) for the Community action in the field of water policy requires the definition of aims to achieve a “good” ecological qual-

∗ Corresponding author at: Department of Animal Ecology, German Federal Institute of Hydrology – BfG, Am Mainzer Tor 1, 56068 Koblenz, Germany. Tel.: +49 0 261 1306 5842; fax: +49 0 261 1306 5152. E-mail addresses: [email protected], [email protected] (M.A. Wetzel). 1470-160X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.ecolind.2011.08.007

ity (EcoQ) status of all water bodies including the estuaries by 2015. The first step towards this goal consists in assessing the current status of these water bodies. Benthic macrofauna plays a vital part in the assessment of the EcoQ, because they are an important component in the aquatic ecosystems and they may serve as sensitive indicators (Rosenberg et al., 2004). Therefore, the Water Framework Directive (WFD) calls for the development of tools for defining the EcoQ status of bodies of water. Several attempts have been made in the past to develop an index based system to estimate the EcoQ status which allows the translation of biological information, such as presence and abundance of macrofauna species, into five different EcoQ classes (high, good, moderate, poor, and bad): the Benthic Index (BI, Grall and Glémarec, 1997), the Azti

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

Marine Biotic Index (AMBI Borja et al., 2000), the M-AMBI (Borja and Franco, 2004; Muxika et al., 2007), and the biotic index BENTIX (Simboura and Zenetos, 2002). These indices use lists of species where species are assigned to ecological groups according to their sensitivity to stress. The index calculation is usually performed with a relatively simple formula. However, further simplification was suggested by Dauvin and Rullet (2007): their index requires only information about the overall abundance of opportunistic polychaetes and amphipods. Besides these recently developed indices, which were especially constructed to meet the requirements of the WFD, several community-descriptive parameters and indices exist that have been used in conjunction with the demands of the WFD. For instance species richness (e.g. Marín-Guirao et al., 2005; Simboura and Reizopoulou, 2007; Simboura and Zenetos, 2002; Dauvin et al., 2007; Borja et al., 2000), total abundance (e.g. Simboura and Reizopoulou, 2007; Dauvin et al., 2007; Borja et al., 2000), Shannon–Wiener diversity (e.g. Teixeira et al., 2007; Marín-Guirao et al., 2005; Simboura and Reizopoulou, 2007; Simboura and Zenetos, 2002; Dauvin et al., 2007; Borja et al., 2000; Labrune et al., 2006), Margalef diversity (e.g. Teixeira et al., 2007; Salas et al., 2004) and the W-value (e.g. Teixeira et al., 2007; Marín-Guirao et al., 2005; Salas et al., 2004), an index that evolved from abundancebiomass comparison (ABC) distribution curves (Warwick, 1986; Clarke, 1990). WFD related studies using the AMBI on its own, or in combination with other indices have been carried out in a number of cases along the European coasts from the Baltic Sea (Muxika et al., 2005), the English channel (Dauvin et al., 2007), the Atlantic coast (Borja et al., 2000, 2003; Salas et al., 2004; Muxika et al., 2005) to the Mediterranean (Simboura and Reizopoulou, 2007; Marín-Guirao et al., 2005; Muxika et al., 2005) and some studies even report their use in estuarine systems (Borja et al., 2000, 2003; Salas et al., 2004, Muxika et al., 2005; Dauvin et al., 2007; Teixeira et al., 2007). In general, applicability of the AMBI in estuarine systems proved to be successful. The index was used to detect different anthropogenically induced changes (Borja et al., 2000), approximated the distribution of organic matter and sediment grain size (Borja et al., 2003) and detected pollution point-sources (Muxika et al., 2005). Estuaries are the most productive and vulnerable marine coastal environments. Here, nutrient-rich freshwater mixes with highly oxygenated waters from the seas, making them one of the biologically most productive and vulnerable regions of the marine environment (Correll, 1978). In addition, estuaries themselves are specific habitats, characterized mainly by strong gradients (salinity, temperature) and by changes and fluctuations of these gradients due to the tidal regime making them unique habitats for a variety of brackish-water species. Moreover, they are also the anthropogenically most altered aquatic systems and susceptible to numerous and strong amounts of pressures. Estuaries have long been influenced by human activities like, dyke constructions, dredging, and pollution, so that today most estuarine ecosystems in industrialized regions are far from being a natural environment and can be considered as strongly disturbed ecosystems. One of the most prominent morphological features of modified estuaries is the presence of a deep navigational channel (fairway) that was prepared in place of shallow-water areas. Large rivers and their estuaries are important routes for navigation and the increasing sizes of ships in the world’s merchant fleets demands regular widening and deepening of the fairways. In the Elbe estuary navigational channel adjustments (fairway deepening) has been carried out gradually in the 20th century, starting with deepening from a primary depth of 3–4 m to 9 m in 1910. Then in 1930, the navigational channel was dredged down to 10 m, in 1962 to 11 m, from 1964 to 1969 to 12 m, and between 1974

119

and 1978 to 13.5 m. In 1999, a navigational depth of 14.4 m was reached. One consequence of this development was shrinking of the highly productive shallow water areas (water depth < 2 m); between 1896/1905 and 1981/1982 the Elbe estuary lost 26% of its shallow waters (Schirmer, 1994). The major requirement for an ecological index is its universal usability. In the case of the Water Framework Directive, benthic marine indices this means that they should be applicable in all marine environments ranging from undisturbed off-shore benthic communities to heavily modified water bodies like estuaries. Although, some indices have already been tested in estuarine environments (e.g. Borja et al., 2003; Salas et al., 2004; Muniz et al., 2005; Muxika et al., 2005; Puente and Diaz, 2008; Ranasinghe et al., 2009), no differentiation was made so far between different morphological structures within estuaries (navigation channel vs. shallow-water areas). The aim of this paper is to explore: (1) the suitability of the different indices in a highly modified estuary with the typical morphological structure elements of a deep navigation channel (fairway) and shallow-water areas, and (2) their ability in identifying different environmental factors like sediment grain-size distribution and its chemical contamination. 2. Material and methods 2.1. Study site The Elbe estuary is located at the southern coast of the North Sea and discharges the River Elbe (catchments size 148,268 km2 ) into the Wadden Sea. The estuary is characterized by diurnal mesotidal conditions (the mean tidal range at the Cuxhaven tidal gage is about 3 m) and by water temperatures ranging from approximately 0 ◦ C in the winter to 26 ◦ C in summer. Salinity can fluctuate between 0.3 and 2.6 PSU in the inner part of the estuary (station Grauerort, river-km 660.6) and from 1.2 to 22 PSU at the mouth (station Cuxhaven, river-km 725.2) depending on season, river runoff, and tidal cycle. The estuarine water body is usually completely mixed due to tidal currents and stream flow; however, when the tide is turning stratification may occur for short intervals (Carstens et al., 2004). The transitional water body of the Elbe estuary extends from river-km 630 to km 727.7 (Office of the River Elbe Water Quality Board, Arge-Elbe, www.arge-elbe.de) and covers an area of some 500 km2 . Anthropogenic modifications of the Elbe estuary have been going on since several centuries. Since the beginning of the 20th century, the estuary has been successively adjusted to the increasing average size of the ships in the merchant fleets to ensure the safe navigation to the port of Hamburg (Schuchardt et al., 1999); the last major fairway adjustment was carried out in 1999/2000 and the next action is already planned for. Today, a safe navigation depth of 14.4 m is maintained; annual maintenance dredging activities in the Elbe estuary move between 5 and 10 Mio. m3 of sediment per year. Most of the dredged sediment is relocated within the estuary (Rolinski and Eichweber, 2000). Pollution release into the river and the estuary declined considerably following the re-unification of Germany in 1990, as a result of the construction of sewage treatment plants, closure of polluting factories, and changes made in industrial production processes. The annual pollution discharge of the River Elbe has dropped from 1986 to 2007 by 29% to almost 100% in some cases, depending on the pollutant (Arge-Elbe, 2007), although historical pollutants are still present in sediments. In the estuary pollutant concentrations usually follows a declining gradient, with higher concentrations occurring close to the port of Hamburg and lower concentrations towards the mouth of the estuary. Higher concentrations are usually found also in zones of low flow velocities, where suspended particles and sediment-bound pollutants (e.g. trace metals)

120

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

2.3. Environmental factors

Fig. 1. Sample locations in the Elbe Estuary; the location of this estuary in relation to the North Sea is indicated in the small map. At each of the 48 sample stations marked by a black circle 4 van Veen grabs (0.1 m2 ) were collected for macrofauna analysis and one for granulometric and chemical analysis (total 192 grabs). The sample stations were positioned between river km 630–730 on the banks and in the fairway covering the entire transitional water body of the estuary. Triangles indicate kilometer mileage every 10 km; the position of the triangles indicates also the center of the navigational channel. The dotted line is the low water line.

accumulate. Mud flats usually have higher concentrations of pollutants than sandy areas.

2.2. Fauna sampling and preparation For the present study we collected samples at 48 stations to determine macrofauna abundance and composition of them in the Elbe estuary, Germany (Fig. 1). Sampling was carried out in October 2006. Sample stations spanned along the entire estuarine transitional waters zone from km 634 near the port of Hamburg to km 729 at the estuary mouth at Cuxhaven. In general, our collection scheme included samples from both banks and from the middle of the fairway, though in some cases only one bank was accessible. Sampling was carried out with the buoy tender “MS Vogelsand” of the Waterways and Shipping Office Cuxhaven and the working vessel “MS Störort” of the Waterways and Shipping Office Hamburg. Softbottom samples were taken by Van-Veen grab with 0.1 m2 surface area. At each station, 4 parallel grabs were taken. Macrofauna was sieved (0.5 mm) and preserved in ethanol (70% final concentration). Fauna was sorted under a dissection microscope and identified to the lowest taxon possible. Biomass (ash-free dry weight, AFDW) of species was analyzed using the ethanol-preserved specimens; Wetzel et al. (2005) have shown that no significant differences exist in biomass estimation between the different macrofauna fixation procedures. In brief: the dry weight (DW) of the samples was measured by placing them for 12 h in an oven at 55 ◦ C. The samples were cooled down to room temperature in a desiccator prior to weighing (3–4 h) to the nearest 0.1 mg. Then, the ash content of the samples was determined by placing them into a muffle furnace at 485 ◦ C for 12 h. Again, samples were cooled down to room temperature in a desiccator (6 h) before weighing. The AFDW was calculated by subtracting the ash-weight from the dry weight.

At each of the 48 stations, a Van-Veen grab with 0.1 m2 surface area was used to collect samples to estimate the environmental factors sediment grain size, organic matter content (total organic carbon, TOC), redox potential, pore water salinity, and nutrients (total N and P). In addition, we analyzed levels of trace metals and organic compounds; the selected compounds are persistent, have high ecotoxicological potential, and accumulate in sediments and biota. All pollutants were estimated for the 20 ␮m grain-size fraction. Sediment characterization was performed by sieving through a nested sieve set (mesh sizes: 4000 ␮m, 2000 ␮m, 1000 ␮m, 500 ␮m, 250 ␮m, 125 ␮m, 63 ␮m, and <63 ␮m). Medium grain size (Md) and sorting coefficient (QDI) were calculated. Total organic matter (TOM) was analyzed by gravimetric determination of sample weight loss. Pore-water salinity was measured using the WTW Multiline F/Set with a TeraCon 325 probe. Conductivities were converted to practical salinity unit values (PSU) according to the UNESCO recommendations (Fofonoff and Millard, 1983; Rathlev, 1980). Total nitrogen (Ntotal ) and total phosphorous (Ptotal ) in the sediment samples were measured according to DIN ISO 11261 for nitrogen and according to DIN EN ISO 11885-E22 for phosphorus. The trace metals AS, Cd, Cu, Pb, Ni, Hg, and Zn were extracted by dissolving in aqua regia (DIN ISO 11466) and the detection was performed by inductively coupled-plasma mass-spectrometry ICP-MS (Agilent 7500c). Mercury (Hg) concentrations were measured via atomic absorption spectroscopy (AAS) in a flow-injection system for atomic spectroscopy FIAS 400 (PerkinElmer Life And Analytical Sciences Inc.) and by atomic-fluorescence spectroscopy (AFS) with the Merkur (Analytik Jena AG). Polycyclic aromatic hydrocarbons (PAH) were analyzed using gas chromatography and a coupled mass-selective sensor (GC-MSD) with the HP 5890 GC (HewlettPackard GmbH). For statistical analysis we used a sum parameter based on 16 EPA PAHs (naphthalene, acenaphthylene, acenaphthene, 9H-fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benzo[a]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[a]pyrene, indeno[1,2,3-c,d]pyrene, dibenzo[a,h]anthracene, benzo[g,h,i]perylene). Concentrations of the low-volatile halogenated hydrocarbon hexaclorbenzol (HCB) were estimated by gas chromatography with a mass-selective detector using negative chemical ionisation (GC-MSD/NDI) with the Agilent 6890N GC (Agilent Technologies GmbH). Polychlorinated biphenyls (7 congeners: PCB 28, PCB 52, PCB 101, PCB 118, PCB 138, PCB 153, PCB 180) were analyzed by gas chromatography. Hexachlorocyclohexane (␤-HCH) was determined using gas chromatography coupled with a mass-selective sensor (GC-MSD) with the Agilent 6890N (Agilent Technologies GmbH) and Organochlorine pesticides (OCPs: DDT and metabolites, octachloro-styrene, octachloro-naphthalene) were determined using the same method. The organotin compound TBT was determined according to E ISO 19744:2003-08 using gas chromatography. For the purpose of quantifying toxic stress related to organic and metal pollutants, toxic units (TUs) are recommended, based on the approach by Sprague (1970). A separation was made because of different community effects that would be expected from organic and metal compounds, at least in invertebrate species (von der Ohe and Liess, 2004). To derive respective TUs, the measured compound concentrations were scaled to the inherent toxicity of each compound towards the standard test organism Daphnia magna: TUmax = log

 C  i LC50i

max

where i is the compound, Ci is the measured environmental concentration of compound i, LC50i is the respective acute lethal

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

concentration in a standard toxicity test (48 h). The endpoint of the acute LC50 was chosen, because it is representative for the observed community effects (Liess and von der Ohe, 2005; Schäfer et al., 2007), and comparable effect data were available of many compounds (von der Ohe and Liess, 2004). The laboratory-derived data were extracted mainly from a database maintained by von der Ohe et al. (2005), the pesticide manual (Tomlin, 2003), the ECOTOX database (USEPA, 2008), the RIVM e-toxbase database (De Zwart, 2002) and the screening information data sets (IPCS, 2008). Concentrations below the LOQ were excluded from the computation of TUs in order to avoid overestimation of risks by including compounds that were likely to be absent (Clarke, 1998). Using only the maximum TU (TUmax ) selected from individual compound values accounts for the minimum expected effect based on the most potent toxicant in the mixture. The lower end of the toxicity range was set to 1/10,000 of the acute LC50 , corresponding to a log TU of −4, which was assumed to be a rather protective concentration level. Sublethal concentrations above a log TU of −3 are expected to have effects on the biota. 2.4. Index calculation For the present study we chose indices that are widely discussed in the scientific literature and may find application for purposes of the European Water Framework Directive. The following indices were incorporated in this analysis: AMBI, M-AMBI, W-value, BOPA, Shannon diversity, number of species and abundance (Table 1 gives an overview about the indices and their EcoQ limits). All indices were calculated on individual samples. The common sample size dependent indices like species richness (S; number of identified taxa) and Shannon–Wiener diversity (H ) were calculated for the 0.1 m2 sampling surface for each individual grab. Species abundance (density, N) of individuals was calculated per 0.1 m2 (N 0.1 m−2 ). The W-statistic (Clarke, 1990) is based on the abundancebiomass comparison (ABC) distribution curves (Warwick, 1986). A method which has been successfully used in detecting the influence of oil pollution (Gray, 1979; Warwick and Clarke, 1994), industrial pollution (Pearson and Rosenberg, 1976), and of the dumping of municipal sludge on macrobenthic communities. The W-statistic transforms the information of the ABC curves into a single value and allows the testing of hypotheses when replicate samples are present. The W-statistic has been developed before the WFD; hence it was developed without a classification scheme in mind. Nevertheless, according to Teixeira et al. (2007) W-values higher than 0.2 are assumed to indicate a “good/high” ecological level, values between 0.1 and 0.0 a “moderate” status, and values below −0.1 a “bad” status. The AMBI (Borja et al., 2000; Borja and Franco, 2004) and its successor, the M-AMBI, is based on ecological models (Glémarec and Hily, 1981). Both indices were calculated following the instructions given by Borja and Muxika (2005) and using the AMBI© software (Borja et al., 2003; http://www.azti.es). The AMBI is constructed so that worsening of EcoQ results in an increase in AMBI, while the M-AMBI follows the conception of most other indicators were increasing values signalize enhancement of the EcoQ. The BOPA (benthic opportunistic polychaetes amphipods index; Dauvin and Rullet, 2007) has evolved from empirical investigations of polychaete/amphipod ratios following an oil spill gradient in soft-bottom communities (Gomez Geistera and Dauvin, 2000). So far, the BOPA index and its predecessor the opportunistic polychaete/amphipod index has been successfully applied to investigate oil spills (Nikitik and Robinson, 2003) and to characterize the ecological status of soft-bottom communities in the English Channel (Dauvin and Rullet, 2007). The BOPA results range between zero (no opportunistic polychaetes present) and 0.30103, with low

121

values indicating a “high”, i.e. unpolluted ecological status while high values stand for a bad, or extremely polluted status; the WFD classification status follows the AMBI scheme. The BOPA has been adapted to transitional waters by including oligochaetes into the BOPA formula resulting in the BO2A (Dauvin and Ruellet, 2009; de-la Ossa-Carretero and Dauvin, 2010). 2.5. Data analysis and statistics Changes in species assemblages along the estuarine salinity gradient were identified by cluster analysis. The similarity matrix was calculated using the Bray Curtis similarity index, with doublesquare-root-transformed abundance data (Kruskal and Wish, 1978; Clarke and Ainsworth, 1993; Clarke and Warwick, 1994). Numbers of clusters were estimated using Rousseeuw’s (1987) technique of silhouette width. The silhouette value measures the degree of confidence in a particular clustering assignment. Analysis of similarities (ANOSIM) pairwise randomization test was applied to assess if significant differences in species compositions between sampling sites and sampling dates were present. Characteristic species within each community were identified by examining their constancy (frequency). Constancy is the relative consistency of occurrence of a species throughout a community (number of samples where the species is found, divided by the number of samples × 100). The constancy of species and the mean number of species were calculated for each of the different communities identified in the cluster analysis. Species were ranked according to their constancy (values ranging from 1.0 = a species that was present in all samples to 0.0 = a species that was not present) from the highest to the lowest. We define characteristic species as those highest ranking species that equal the mean number of species within the community. Differences in indices between benthic communities were analyzed using one-way ANOVA followed by a post hoc test (multiple t-tests). Prior to index calculation samples containing less than 20 individuals were omitted prior to the calculation (e.g. dela Ossa-Carretero and Dauvin, 2010); 18 samples from our total of 192 benthos samples fulfilled this criteria. In order to account for the Type I error associated with multiple t-tests p-values were Bonferroni-adjusted. Associations between the indices (N, S, H , Wstatistic, AMBI, BOPA, BO2A) and different environmental factors (sediment grain size: Md, QDI; organic matter and nutrients: TOM, N, P; trace metals: AS, Cd, Cu, Pb, Ni, Hg, and Zn; organic pollutants: HCB,  PCB7 , ␤-HCH, pp -DDE, pp -DDD,  PAHEPA , TBT); were tested for significance using Pearson’s product moment correlation coefficient (e.g. Quinn and Keough, 2002). Statistical calculations and figure compilations were performed using the freely available R computer language, Version 2.11. (Ihaka and Gentleman, 1996; R Development Core Team, 2008). 3. Results Community analysis of the benthic fauna showed a distinct overall difference in community composition along the estuarine gradient at km 680/690 (Fig. 2) that is expressed in the formation of two main clusters. Analysis of similarity (ANOSIM) results indicated that four significantly different communities were present (global R = 0.52, p < 0.001): the inner-fairway community from km 630 to 680 differed significantly from the outer-fairway community from km 690 to 730 (R = 0.30, p < 0.001), the inner-bank community (km 630–680) differed significantly from the outer-bank community (km 690–730, R = 0.71, p < 0.001), the inner-bank community (km 630–680) differed significantly from the inner-fairway community (km 630–680, R = 0.41, p < 0.001), and the outer-bank community (km 690–730) differed significantly from the outer-fairway community (km 690–730, R = 0.29, p < 0.001). These changes in

122

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

Table 1 Classes associated with the different ecological quality (EcoQ) status proposed within the WFD. Literature sources are given at the bottom of the table. EcoQ

S

H

High Good Moderate Poor Bad

>60 45–60 30–45 15–30 0–15

>4 3–4 2–3 1–2 <1

AMBI 0.0–1.2 1.2–3.3 3.3–4.3 4.3–5.5 5.5–7.0

M-AMBI

W-valuea

W-valueb

BOPA/BO2A

>0.85 0.55– 0.85 0.39–0.55 0.20–0.39 <0.20

0.1–1.0 0.1–1.0 −0.1 to 0.1 <−0.1 <−0.1

0.50–1.00 0.15–0.49 −0.14 to 0.14 −0.49 to −0.15 −0.50 to −1.00

<0.02452 0.13002–0.02452 0.19884–0.13002 0.25512–0.19884 >0.28

S: Borja et al. (2003); H : Labrune et al. (2006); AMBI: Borja et al. (2003); M-AMBI: Bald et al. (2005), Borja et al. (2007); BOPA/BO2A: Dauvin and Rullet (2007), de-la Ossa-Carretero and Dauvin (2010). a Teixeira et al. (2007). b Marín-Guirao et al. (2005).

community structure along the estuarine gradient correspond very well with the known location of the estuarine turbidity maximum zone and the zone with the highest change in salinity at km 680–690 (Rolinski and Eichweber, 2000). Further analysis differentiated between these four community groups (inner fairway, inner bank, outer fairway, and outer bank, for species abundances of these groups see Table 2) and comparisons between bank and fairway were made only between inner and outer communities. Changes in relative dominance of higher macrofauna taxa and mean abundance (number of individuals m−2 ) along the estuary is given in Fig. 3, while Fig. 4 gives the results of the comparisons between sublitoral bank sediments and fairways divided between the inner and the outer estuarine parts are shown in Fig. 3. From km 630 to km 730 (Fig. 3) mean abundance was fairly high from km 630 to 660, with values ranging from approximately 213 individuals 0.1 m−2 at km 640 to 958 individuals 0.1 m−2 at km 630. In contrast, mean abundance from km 670 to 730 ranged between 41 individuals 0.1 m−2 and 263 individuals 0.1 m−2 . Relative abundance of taxonomic groups changed most

drastically at km 680, whereby from km 670 upstreams mainly oligochaets dominated the picture, and from km 680 to 730 polychaetes and crustaceans are the dominating taxonomic groups. Comparisons between sublitoral bank and fairway communities (Fig. 4) show that the upper estuarine part (km 630–680) was dominated by oligochaetes (94% of the all taxa) on the sublitoral banks. Characteristic for these sediments were mainly species of the family Tubificidae (Limnodrilus hoffmeisteri, Limnodrilus claparedianus, Potamothrix moldaviensis, and two other unidentified species of the family Tubificidae, constancy: 67%, 38%, 33%, 92%, and 53%), however, the spionid polychaete Marenzelleria sp. was also found with high constancy (33%) in these sediments. In contrast, the sediment of the fairway in the upper estuarine part was dominated by crustaceans (66%; species with the highest constancy were Neomysis integer and Bathyporeia pilosa with 60% and 57%, respectively) and polychaetes (28%; constant species: Marenzelleria sp. with 46%). Oligochaetes accounted only for 8% of the macrofauna in which one, not exactly identifiable, tubificid species was found in 50% of all samples. Mean abundance was lower in upper-fairway sediments

Table 2 Mean abundance of macrobenthos species (ind. 0.1 m−2 ), mean total abundance (ind. 0.1 m−2 ), total richness, and mean total biomass (mg afdw 0.1 m−2 ) collected in the four different groups (inner estuary bank, inner estuary navigational channel, outer estuary bank, outer estuary navigational channel) in the Elbe estuary in October 2006. Inner estuary (<685 km) Bank

Outer estuary (>685 km) Navigation channel

Bank

Navigation channel

Amphipoda Bathyporeia pilosa Corophium lacustre Corophium volutator Gammarus salinus

8 – – –

74 29 – –

1 – 11 13

36 – – –

Mysidacea Mesopodopsis slabberi

1

–

5

6

Polychaeta Boccardiella ligerica Marenzelleria sp. Marenzelleria neglecta Marenzelleria viridis Neanthes succinea

8 5 1 – –

31 8 4 – –

1 55 5 – 16

– 104 1 13 –

16 109 9 84 2 7 – 12 287

43 49 1 2 4 – – 23 143

– – – – – – 11 – 1

– – – – – – – – – –

Clitellata Limnodrilus claparedianus Limnodrilus hoffmeisteri Limnodrilus udekemianus Paranais litoralis Potamothrix hammoniensis Potamothrix moldaviensis Tubificoides heterochaetus Tubificidae sp. 1 Tubificidae sp. 2 Bivalvia Pisidium moitessierianum Other taxa Total number of species Total abundance (ind. 0.1 m−2 ) Total biomass (mg afdw 0.1 m−2 )

3

–

–

12

4

14

1

43 563 121

36 414 124

47 134 1061

18 160 163

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

123

Table 3 Results of the one-way ANOVAs (F-values with (df) and p values) and pairwise post hoc comparisons (Bonferroni adjusted p-values) for mean total abundance (N), mean species richness (S), mean Shannon–Wiener diversity (H ), mean W-value, mean AMBI, mean BOPA, and mean BO2A. N

S



H

W-value

AMBI

M-AMBI

BOPA

BO2A

F(3, 170) = 6.691, p < 0.001 Fairway 630–680 Bank 630–680 Bank 680–730 630–680 Bank

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 0.720 p = 0.014 p = 1.000 p = 1.000

F(3, 170) = 13.056, p < 0.001 Fairway 630–680 Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 1.000 p = 0.047 p < 0.001 p < 0.246

F(3, 170) = 23.188, p < 0.001 630–680 Fairway Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 1.000 p < 0.001 p < 0.001 p = 0.002

F(3, 170) = 4.535, p < 0.001 Fairway 630–680 Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 1.000 p = 0.001 p = 0.037 p = 1.000

F(3, 170) = 56.511, p < 0.001 630–680 Fairway Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p < 0.001 p < 0.001 p = 0.003 p < 0.001

F(3, 170) = 18.592, p < 0.001 Fairway 630–680 Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 1.000 p < 0.001 p < 0.001 p = 0.016

F(3, 170) = 55.573, p < 0.001 630–680 Fairway Bank 630–680 Bank 680–730 Bank 630–680

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p < 0.001 p < 0.001 p = 1.000 p = 0.230

F(3, 170) = 14.1, p < 0.001 Fairway Bank Bank Bank

vs. vs. vs. vs.

Fairway Bank Fairway Fairway

680–730 680–730 680–730 630–680

p = 0.168 p = 1.000 p = 0.289 p < 0.001

630–680 630–680 680–730 630–680

*

* *** ***

*** *** **

** *

*** *** ** ***

*** *** *

*** ***

***

Blank = not significant. * p < 0.05. ** p < 0.01. *** p < 0.001.

●

730

●

710

●

1500 ●

690

●

●

●

●

670

375

40 20

●

750

● ●

650

630

abundance (N ind. 0.1 m−2)

60

80

Gastropoda Malacostraca Oligochaeta Polychaeta

0

Fig. 2. Cluster diagram of averaged macrofauna data (samples were averaged in 10 km groups following the kilometer mileage), along the estuarine gradient of the Elbe Estuary. A distinct difference in community composition was present between km 640–680 and km 690–730. ANOSIM results indicated also the existence of differences between fairway and sublitoral bank communities in both parts.

rel. abundance (%)

km

0

640

630

650

660

670

680

690

700

710

720

730

(417 ind. 0.1 m−2 ) compared with the sublitoral bank sediments (564 ind. 0.1 m−2 ). In the outer estuarine section (km 690–730) the sublitoral bank sediments were dominated by polychaetes (65%), crustaceans (23%), and oligochaetes (10%). Species found in high constancy were the polychaetes Marenzelleria sp. (77%), Neanthes succinea (69), Nephtys hombergii (33%), Pygospio elegans (30%), the

Fig. 3. Development of relative abundance of higher taxa and mean total abundance (N ind. m−2 ) along the estuarine gradient of the Elbe Estuary from km 630 to 730. Only taxa with relative abundance above 5% are shown.

fairway

700 525 175

350

80 60 40

0 bank

fairway

abundance (N ind. 0.1 m−2)

bank

●

20

350 0

175

●

●

0

525

80 60 40 20

●

0

rel. abundance (%)

700

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

abundance (N ind. 0.1 m−2) rel. abundance (%)

124

Fig. 4. Comparisons of relative abundance of higher taxa and mean total abundance (N ind. m−2 ) between sublitoral bank and fairway. Left graph shows the comparison for the outer estuarine section (km 690–730) and the right graph shows the comparison for the inner estuarine part (km 630–680). Only taxa with relative abundance above 5% are shown.

crustaceans Mesopodopsis slabberi (63%) and Corophium volutator (30%), as well as the oligochaet Tubificoides heterochaetus (41%). In addition, the mud snail Hydrobia ulvae was consistently found in samples from the outer sublitoral bank sediments (30% constancy). The outer estuarine fairway sediments were populated mainly by polychaetes (73%; Marenzelleria sp. was present in 96% of all samples) and crustaceans (26%). From the crusteaceans mainly B. pilosa, M. slabberi, and N. integer were found in high constancy (85%, 85%, and 25%, respectively). Mean abundance did not show any significant difference between outer-fairway sediments (161 ind. 0.1 m−2 ) and outer-bank sediments (135 ind. 0.1 m−2 ; Table 3). All indices underwent basically the same overall development along the estuarine gradient (Fig. 5), though AMBI, BOPA, and BO2A values are more or less mirror-inverted due to the calculation processes which associated high values with a “bad” EcoQ and low values with a “good” ecological status. The mean number of species was highest (S = 10) in the outer part of the estuary at km 730; further upstream in the estuary the number of species had a declining tendency with the lowest values of S = 5 at km 700, 680, and 670 and reaching at km 630 a value of S = 7. According to the suggestion of Teixeira et al. (2007) for the conversion of species richness indices to WFD EcoQ status groups, a value of S < 25 would be considered to characterize a “bad” ecological status (see also Table 1). The Shannon–Wiener diversity displayed a similar pattern: the highest value of H = 1.9 was observed at km 730, while in all other stations along the estuary it dropped to values below H = 1.2. Translated into the WFD EcoQ status: the outmost portion of the estuary (km 730) would be considered “poor”, at km 720 this status would be even “bad” with some small recovery at km 710 into the “poor” status, while the rest, except km 630 (“poor”) would be considered “bad”. A similar trend, though with some stronger fluctuations, is observed in the mean W-value with higher values in the outer part of the estuary (W = 0.15 at km 730) and lower ones in the upper portion, km 630: W = 0.03. WFD wise, most parts of the estuarine gradient would be classified as “moderate” (cf. Table 1) with one exception: at km 730 the mean W = 0.15 would lead to “good/classification. The AMBI results followed this overall pattern but although in a much more detailed classification: from km 630 to 640 the EcoQ status would be considered “poor” (AMBI values ranged from 5.2 to 5.9), from km 650 to 670 the status is described as “moderate” (AMBI values from 4.1 to 4.2), and from km 680 to 730 the status is “good” and sometimes even “high” (AMBI range 1.4–2.9). The MAMBI identified the outer most section of the estuary (km 730) as good (M-AMBI = 0.7), while the rest of the estuary is “moderate” to “poor” (M-AMBI = 0.3–0.5). The BOPA was highest in the outer part of the estuary (km 690–730, range = 0.15–0.2) indicating a “poor” to “moderate” EcoQ status. Up the estuary from km 680, the BOPA values declined drastically with values that would normally correspond to a “good” or “high” EcoQ status. However, this result is due

Fig. 5. Development of indices along the estuarine gradient of the Elbe Estuary from km 630 to 730. From top to bottom the graphs show: (1) species richness (S) and Shannon–Wiener Diversity (H ), (2) W-value, (3) relative abundance of AMBI groups (only groups with relative abundance above 5% are shown) and the mean AMBI value, and (4) BOPA BO2A values. Error bars represent one standard error. Class ranges associated with the different ecological quality (EcoQ) status ratings (dashed lines) are given for S, H , W-value, AMBI, BOPA, and BO2A. For the origins of the classification ranges see Table 1. Number of samples (n) used for this figure are given below the top graph.

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

125

to the natural decline of polychaete species following the estuarine gradient from the estuary mouth inwards. Because the index is based on the proportion of amphipods and polychaetes, the natural loss of one group along the estuarine gradient leads to incorrect results. The BO2A overcomes this natural restriction of the BOPA in the estuarine transitional waters by including oligochaetes. The BO2A showed an almost similar development like the BOPA along the outer portion of the estuarine gradient from km 690 to 730 with values ranging between 0.19 and 0.24. From km 680 inwards, the BO2A climbed up to 0.3 at km 630 indicating a bad EcoQ status. The BO2A showed the same tendency like the AMBI, though the classification according to the WFD scheme is sometimes different. Comparing the ecological indices between sublitoral bank and fairway in the outer and in the inner estuary gave a more differentiated picture (Fig. 6): almost all indices (S, H , W-value, AMBI, M-AMBI, BOPA, and BO2A) had lower values in the fairway sediments compared with the sublitoral bank sediments. These differences were significant (p < 0.001, see Table 3 for ANOVA post hoc comparison results) for species richness, Shannon–Wiener diversity, W-value, AMBI, and M-AMBI while this comparison was not always significant for the BOPA and the BO2A. The mean BOPA and the BO2A were significantly smaller (p < 0.001) only in the fairway community of the upper estuary. The highest mean AMBI value (AMBI = 5) was found in the community of the sublitoral shallowwater bank communities in the upper part of the estuary, while the lowest value was calculated for the fairway of the outer estuary (mean AMBI = 1). The comparison between the indices and several environmental factors (sediment grain size: Md, QDI; organic matter and nutrients: TOM, Ntotal , Ptotal ; trace metals: AS, Cd, Cu, Pb, Ni, Hg, and Zn; organic pollutants: HCB,  PCB7 , ␤-HCH, pp -DDE, pp -DDD,  PAHEPA , TBT) using Pearsons product momentum correlation coefficients is summarized in Table 4. All indices, except the Wvalue and the BOPA, showed a significant negative correlation with the mean grain size, while no significant correlation was found for grain size sorting coefficient. Pore-water salinity was significant for Shannon–Wiener diversity (H ), AMBI, and BOPA while a significant correlation with the redox potential was noted for the AMBI and the BO2A. Among the nutrients total phosphorous (Ptotal ) showed significant correlations with some indices: N, W-value, and BOPA. All indices, except species richness (S) and the BOPA, showed some significant correlations with sediment pollutants (Table 4). However, only one index, the W-value, was correlated with the majority of chemical pollutants (Pb, Cd, Cu, Ni, Hg, Zn, ␤-HCH, pp DDD, and TBT). Other indices (N, H , AMBI, and BOPA) had some significant correlation to a few pollutants, while some indices (S, M-AMBI, BO2A) showed no significant correlation to any pollutants. However, the picture is rather blurred: the total abundance (N) was significantly correlated with Cu and TBT concentrations, the Shannon–Wiener diversity to ␤-HCH, the AMBI to HCB and TBT, and the BOPA to Cd, HCB, and TBT.

4. Discussion In this study we applied different EcoQ indices that are discussed for application according to the EU Water Framework Directive to the Elbe estuary, a highly modified estuary which is characterized by a deep navigation channel and areas where sediments are heavily polluted. Our investigation showed that with most ecological indices (AMBI, BO2A, H , N, S) significant differences could be observed between the fairway and sublitoral shallow-water communities, while no such differences were found with the Wvalue. W-value, S, and H all followed a similar pattern with inferior

Fig. 6. Comparisons of indices between sublitoral bank and fairway. Left graphs show the comparisons for the outer estuarine section (km 690–730) and the right graphs show the comparisons for the inner estuarine part (km 630–680). Error bars represent one standard error. Class ranges associated with the different ecological quality (EcoQ) status ratings (dashed lines) are given for S, H , W-value, AMBI, BOPA, and BO2A. For the origins of the classification ranges see Table 1. Number of samples (n) used for this figure are given below the top graph.

126

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

Table 4 Results of Persons product moment correlation coefficient between the indices (S, N, H , W-statistic, AMBI, BOPA, BO2A) and different environmental factors (mean grain size, sediment sorting, % TOM, pore water salinity, and redox potential), nutrients (total P, total N), heavy metals (AS, Cd, Cu, Pb, Ni, Hg, and Zn), and organic pollutants (HCB,  PCB7 , ␤-HCH, pp -DDE, pp -DDD, PAH6 ,  PAHEPA , 1.2-benzpyren, TBT20 ␮m ). Concentration ranges are given. Range

S

N

H’

W-value

AMBI

M-AMBI

BOPA

BO2A

54–256 1.2–2.36 0–24 0.5–25

−0.40** 0.06 0.13 0.18

−0.35* −0.28 −0.06 −0.23

−0.25 0.19 0.10 0.38**

−0.14 0.34 0.22 0.30

−0.54*** −0.02 0.28 −0.44**

−0.42** 0.03 0.13 0.31*

−0.07 −0.01 −0.03 0.61***

−0.62*** −0.03 0.28 −0.08

mg kg−1 mg kg−1

592–5733 2308–52,632

0.09 0.06

0.47** −0.06

−0.13 0.04

−0.35* 0.19

0.24 0.12

0.11 0.12

−0.35* −0.22

0.05 0.02

mg kg−1 mg kg−1 mg kg−1 mg kg−1 mg kg−1 mg kg−1 mg kg−1 ␮g kg−1 ␮g kg−1 ␮g kg−1 ␮g kg−1 ␮g kg−1 mg kg−1 mg kg−1 mg kg−1

0–60 0–128 0–4 0–78 0–58 0–3 0–691 0–27 0–85 0–7 0–20 0–43 0–6.4 0–0.7 0–0.6

−0.20 0.10 0.04 0.09 −0.02 0.03 0.08 0.10 0.01 −0.15 −0.01 0.01 0.01 0.00 0.29

−0.18 −0.04 −0.37 * −0.26 −0.08 −0.20 −0.28 −0.35* −0.33 −0.10 −0.33 −0.10 −0.24 −0.30 −0.40*

−0.05 0.09 0.00 0.11 −0.09 0.03 0.11 0.15 −0.17 0.07 −0.04 0.01 −0.03 −0.12 0.31

Mean grain size Grain size sorting TOM Porewater salinity

␮m

Ptotal Ntotal As Pb Cd Cu Ni Hg Zn HCB  PCB7 ␤-HCH pp -DDE pp -DDD  PAHEPA Benzpyren TBT TUorganica TUmetals

% PSU

−0.11 0.04 0.03 0.04 −0.02 0.01 0.01 0.14 0.05 −0.08 0.01 0.04 0.04 0.07 0.27

0.10 0.21 0.28 0.34* 0.15 0.26 0.32 0.21 0.18 0.24 0.10 0.14 0.20 0.21 0.65***

−0.11 −0.19 −0.20 −0.26 −0.33 −0.21 −0.24 −0.17 −0.06 −0.48** −0.12 −0.20 −0.13 −0.13 −0.08

−0.20 −0.36* −0.32* −0.39* −0.38* −0.29* −0.37* −0.25 −0.21 −0.39* −0.29 −0.32* −0.28 −0.27 −0.40**

0.07 0.05 0.24 0.22 0.00 0.14 0.24 0.34* 0.11 0.03 0.20 0.04 0.13 0.11 0.49**

−0.06 −0.11

0.44** 0.28

−0.30 −0.22

−0.55*** −0.41**

0.29 0.17

0.12 0.08

−0.23 −0.19

0.28 0.13

Blank = not significant. * p < 0.05. ** p < 0.01. *** p < 0.001.

estimates for the fairway communities and better estimates for the shallow-water communities. However, opposite ratings were given by AMBI and BOPA, which both rated shallow-water communities as worse than fairway communities, a fact that contradicts common environmental perception that shallow-water communities are ecological valuable and deserve protection. These observations corresponded very well with the grain-size distribution pattern in the estuary: sediments of the fairway were significantly coarser than those in the shallow-water environments, what finds its expression in the results of the correlation analysis. The AMBI was highly significantly correlated with mean grain size, while the BOPA and the W-value did not show a significant correlation towards granulometric parameters in our study; obviously, the AMBIs EcoQ status is highly dependent on the substrate type. In addition, we were able to demonstrate in our study that only one index, the W-value, was significantly correlated (using Pearson’s product-moment correlation coefficient) to most of the sediment pollutants investigated (Pb, Cd, Cu, Ni, Hg, Zn, ␤-HCH, pp -DDD, and TBT). All other indices showed almost no (S, BOPA) or only a few significant relations to some sediment pollutants (AMBI, BOPA, N, H ). Between January and September, prior to our sampling campaign, 2 Mio. m3 were dredged in the navigation channel in the inner estuary (km <685), and a little more than 3.6 Mio. m3 were dredged in the outer section (km >685) (Waterways and Shipping Office Cuxhaven, personal communications). Despite these large amounts of dislocated sediment we found relatively high abundances in some species. Dredging in the Elbe estuary follows a rather patchy pattern; only there where the navigation channel does not have the target depth, sediment is removed. This procedure results in a pattern of relatively small patches depleted of fauna that are easily re-colonized by adult individuals which had been suspended into the water column by shear stress through floodflow in the river, the tidal currents, or the direct impact of large ships propellers. In addition, the bottom structure in the Elbe

estuary, especially in the navigation channel most likely enhances mixing of sediments and benthic invertebrates: in the navigation channel large areas of dunes several meters high are present that are known to travel with a speed of nearly 1 m per day as a function of the rivers floodflow towards the river mouth (Gehres and Winterscheid, 2011). This dune movement creates a highly mobile sediment layer that is able to transport benthic invertebrates and therefore makes the easy and fast re-colonization of defaunated patches possible. Although we do not yet know how fast previously dredged patches are re-colonized, the highly dynamic sediment transport in combination with natural (river flow and tidal regime) and vessel-induced shear stress (from ships propellers) most likely promotes fast re-colonization of such defaunated areas, so that the direct traces of dredging should be visible for a few weeks only. The most striking findings of our study is that only one index, the W-value, showed a good correlation with most sediment pollutants and to the toxicity of the sediments estimated using toxic units. Moreover, this index was the only one, which did not show any significant correlations to sediment parameters like mean grain size and sediment sorting. A feature that is also displayed by the lack of significant differences between the samples from the coarser navigational channel and the finer bank sediments. In our study, the W-value was the index that was based on the approach of kstrategists (large animals with high biomass and long reproductive cycle) vs. r-strategists (small animals with low biomass and short reproductive cycle), while all other indices (AMBI, M-AMBI, and BOPA) were based on the Pearson and Rosenberg (1978) model, which describes the successional changes of benthic macroinvertebrates as a function of organic pollution. Therefore, our results seem to favor the r vs. k strategist approach for WFD indicators. However, can we be sure that this approach is really the model of choice in terms of a reliable and robust indicator to estimate the EcoQ potential of estuaries? Previous studies (e.g. Dauer et al., 1993;

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

Chainho et al., 2007) have shown that the W-values response can be explained by the high dominance of a few species in the estuary that results in a species abundance and biomass distribution similar to that generated by opportunistic species in disturbed sites. In our case low (negative) W-values were most prominent in the inner part of the Elbe estuary where exactly this situation occurs: the inner part is dominated by oligochaet species which fit the description of r-strategists with their small body size and their high reproducibility. Therefore, we cannot rule out that our findings may have been influenced by this. So, our results concerning the correlation between the W-value and the sediment pollution and sediment toxicity are maybe simply due to the coincidence that sediment pollution occurred just in the estuarine area were oligochaetes are naturally found in high abundances. So, how can we find out, whether our results are just a pseudo correlation between the EcoQ index and environmental factors like anthropogenically induced pollution gradients in the estuarine environment? One possible solution is the incorporation of historical knowledge. About the Elbe estuary we know that many biomass rich species like molluscs (Unio tumidus, Anodonta cygnea) and gastropods (Theodoxus fluviatilis, Lymnaea stagnalis) occurred in sometimes high abundances in the today oligochaet-dominated part of the estuary at the end of the 19th century (Petermeier et al., 1996); today these species are missing. With this knowledge our findings appear to be not just pure coincidence but rather the result of several decades of anthropogenic activities and sediment contamination within the estuary which eliminated many sensitive species. A second approach would be the comparison of polluted and unpolluted estuaries. Some recent studies (e.g. Bald et al., 2005; Blanchet et al., 2008) have already gone this way. However, in most cases, these studies lack the inclusion of the W-value or they have not performed correlation analysis with sediment pollutants. The W-value method (Clarke, 1990) seemed to work adequately in the estuarine environment of the Elbe by indicating most of the sediment pollutants with significant correlations. However, scanning the scientific literature this does not seem to be the case always: Marín-Guirao et al. (2005) found no such correlations with sediment metal concentrations in their studies of the coastal waters off Portmán and the Mar Menor coastal lagoon (Spain) and Teixeira et al. (2007) observed similar misclassifications in the Mondego estuary (Portugal). In cases when pollution consists only of metal contaminants some authors (Warwick et al., 1987; Anderlini and Wear, 1992; Marín-Guirao et al., 2005) have reported that W-values outcomes did not identify a pollution gradient and recruitment events can lead to classification of unstressed communities as highly stressed (Beukema, 1988; Dauer et al., 1993). However, other authors (e.g. Austen et al., 1989; Ritz et al., 1989) found this method to be a sensitive indicator of the effects of sediment contamination on macrobenthic communities. Another point of importance that needs to be addressed here is the fact that the choice of the indicator itself strongly influences the outcome of the EcoQ status characterization. In our study, the EcoQ status if the Elbe estuary ranged from “good” to “bad”, depending on the indicator applied. This alone is very much dissatisfying; however, when this problem of diverging classification status between different indicators will not be solved in the near future, this will most likely lead to strong disputes between stakeholders and environmentalist. At present, we have a situation, that with the choice of the index a better, or more optimistic, assessment can be obtained. Because of the economic importance of many estuaries, such a change from the pure scientific discussion to a more public dispute is most likely to happen in the near future. To clear this situation, not only reliable indicators have to be identified, but also adaptations between WFD classifications between indicators have to be made in order to get an even picture.

127

5. Conclusion Our results clearly show that the EcoQ classification of heavily modified estuaries depends strongly on (1) the considered biotic index and (2) the choice of the sample sites (habitat). We conclude that no single index should be used on its own to estimate the EcoQ of an estuary and that the index should be able to associate the ecological status of the benthic communities with the chemical contamination of their habitat. Sediment contamination is still on of the most pressing problems in many rivers and estuaries in industrialized countries. In our study the only index able to reflect contaminated sediments was the W-value. We therefore hope that this index will be used in upcoming studies on the ecological quality status of estuaries more often so that the basic parameters for its application can be identified. Further investigations and the improvement or development of indices should place emphasis on their independence from the grain size spectrum of the sediments and on their good correlation with the estuaries pollution status. Acknowledgements We are greatly indebted to help from our colleagues Angela Koppers, Annika Linke (German Federal Institute of Hydrology – Bfg), and Regina Bönsch (IfaÖ consulting company) who were involved in sampling and/or laboratory treatment of fauna samples. Many thanks go to the crews of the ships “MS Vogelsand” and “MS Störort” from the Waterways and Shipping Offices in Cuxhaven and Hamburg for their efficient help and their hospitality on the ships throughout the sampling campaign. We would also like to acknowledge Prof. Jean Claude Dauvin for his comments on a previous version of this manuscript that greatly improved our manuscript. The views and conclusions contained in this paper are those of the authors and should not be interpreted as representing the opinions or policies of the German Federal Institute of Hydrology (BfG) or the German Government. References Anderlini, V.C., Wear, R.G., 1992. The effect of sewage and natural seasonal disturbances on benthic macrofaunal communities in Fitzroy Bay, Wellington, New Zealand. Marine Pollution Bulletin 24, 21–26. Arge-Elbe, 2007. Arbeitsgemeinschaft für die Reinhaltung der Elbe. Die Wassergüte der Elbe im Jahre 1997. Accessible at http://www.fggelbe.de/tl fgg neu/veroeffentlichungen.html. Austen, M.C., Warwick, R.M., Rosado, M.C., 1989. Meiobenthic and macrobenthic community structure along a putative pollution gradient in southern. Portugal Marine Pollution Bulletin 20, 398–405. Bald, J., Borja, A., Muxika, I., Franco, J., Valencia, V., 2005. Assessing reference conditions and physico-chemical status according to the European Water Framework Directive: a case-study from the Basque Country (Northern Spain). Marine Pollution Bulletin 50, 1508–1522. Beukema, J., 1988. An evaluation of the ABC method abundance/biomass comparison as applied to macrozoobenthic communities living on tidal flats in the Dutch Wadden Sea. Marine Biology 99, 425–433. Blanchet, H., Lavesque, N., Ruellet, T., Dauvin, J., Sauriau, P., Desroy, N., Desclaux, C., Leconte, M., Bachelet, G., Janson, A.-L., Bessineton, C., Duhamel, S., Jourde, J., Mayot, S., Simon, S., de Montaudouin, X., 2008. Use of biotic indices in semienclosed coastal ecosystems and transitional waters habitats – implications for the implementation of the European Water Framework Directive. Ecological Indicators 8 (4), 360–372. Borja, A., Franco, J., Muxika, I., 2004. The biotic indices and the water framework directive: the required consensus in the new benthic monitoring tools. Marine Pollution Bulletin 48 (3–4), 405–408. Borja, A., Franco, J., Pérez, V., 2000. A marine biotic index to establish the ecological quality of soft-bottom benthos within european estuarine and coastal environments. Marine Pollution Bulletin 40 (12), 1100–1114. Borja, A., Josefson, A.B., Miles, A., Muxika, I., Olsgard, F., Phillips, G., Rodríguez, J.G., Rygg, B., 2007. An approach to the intercalibration of benthic ecological status assessment in the North Atlantic ecoregion, according to the European Water Framework Directive. Marine Pollution Bulletin 55 (1–6), 42–52. Borja, A., Muxika, I., 2005. Guidelines for the use of AMBI (AZTI’s Marine Biotic Index) in the assessment of the benthic ecological quality. Marine Pollution Bulletin 50, 787–789.

128

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129

Borja, A., Muxika, I., Franco, J., 2003. The application of a marine biotic index to different impact sources affecting soft-bottom benthic communities along European coasts. Marine Pollution Bulletin 46, 835–845. Carstens, M., Claussen, U., Bergemann, M., Gaumert, T., 2004. Transitional waters in Germany: the Elbe estuary as an example. Aquatic Conservation: Marine and Freshwater Ecosystems 14, 81–92. Chainho, P., Costa, J., Chaves, M., Dauer, D., Costa, M., 2007. Influence of seasonal variability in benthic invertebrate community structure on the use of biotic indices to assess the ecological status of a Portuguese estuary. Marine Pollution Bulletin 54, 1586–1597. Clarke, J., 1998. Evaluation of censored data methods to allow statistical comparisons among very small samples with below detection limit observations. Environmental Science and Technology 32, 177–183. Clarke, K., 1990. Comparison of dominance curves. Journal of Experimental Marine Biology and Ecology 138 (1–2), 130–143. Clarke, K., Ainsworth, M., 1993. A method of linking multivariate community structure to environmental variables. Marine Ecology Progress Series 92, 205–219. Clarke, K., Warwick, R., 1994. Similarity-based testing for community pattern: the 2-way layout with no replication. Marine Biology 118, 167–176. Correll, D.L., 1978. Estuarine productivity. BioScience 28 (10), 646–650. Dauer, D., Luckenbach, M., Rodi, A., 1993. Abundance biomass comparison ABC method: effects of an estuarine gradient, anoxic/hypoxic events and contaminated sediments. Marine Biology 116, 507–518. Dauvin, J., Rullet, T., 2007. Polychaete/amphipod ratio revisited. Marine Pollution Bulletin 55, 215–224. Dauvin, J.-C., Ruellet, T., Desroy, N., Janson, A.-L., 2007. The ecological status of the Bay of Seine and the Seine estuary: use of biotic indices. Marine Pollution Bulletin 55, 241–257. Dauvin, J.-C., Ruellet, T., 2009. The estuarine quality paradox: is it possible to define an ecological quality status for specific modified and naturally stressed estuarine ecosystems? Marine Pollution Bulletin 59, 38–47. De Zwart, D., 2002. Observed regularities in species sensitivity distributions for aquatic species. In: Posthuma, L., Suter, G., Traas, T. (Eds.), Species-Sensitivity Distributions in Ecotoxicology. CRC-Press, Lewis, Boca Raton, FL, USA, pp. 133–154. de-la Ossa-Carretero, J.A., Dauvin, J.-C., 2010. A comparison of two biotic indices, AMBI and BOPA/BO2A, for assessing the ecological quality status (EcoQS) of benthic macro-invertebrates. Transitional Waters Bulletin 4, 12–24. EC, 2000. Directive 2000/60/EC of the European Parliament and of the Council of 23 October 2000 establishing a framework for community action in the field of water policy. Official Journal of the European Communities L 327 (L 327/1), 1–72. Fofonoff, N., Millard Jr., R.C., 1983. Algorithms for computation of fundamental properties of seawater. Endorsed by Unesco/SCOR/ICES/IAPSO Joint Panel on Oceanographic Tables and Standards and SCOR Working Group 51. Unesco Technical Papers in Marine Science 44. Gehres, N., Winterscheid, A., 2011. Untersuchung der Dynamik von Transportkörpern sowie deren Oberwasserabhängigkeit an ausgewählten Flussabschnitten in der Tideelbe. Technical Report BfG-1710. Bundesanstalt für Gewässerkunde, Koblenz. Glémarec, M., Hily, C., 1981. Perturbations apportées à la macrofaune benthique de la baie de Concarneau par les effluents urbains et portuaires Acta Oecologica. Oecologica Applicata 2, 139–150. Gomez Geistera, L., Dauvin, J., 2000. Amphipods are good bioindicators of the impact of oil spills on soft-bottom macrobenthic communities. Marine Pollution Bulletin 40, 1017–1027. Grall, J., Glémarec, M., 1997. Using biotic indices to estimate macrobenthic community perturbations in the Bay of Brest. Estuarine Coastal and Shelf Science 44 (Suppl. A), 43–53. Gray, J., 1979. Pollution-induced changes in populations. Philosophycal Transactions of the Royal Society of London Series B 286, 545–561. Ihaka, R., Gentleman, R., 1996. R: a language for data analysis and graphics. Journal of Computational and Graphical Statistics 5, 299–314, Screening information data set (SIDS) for high production volume chemicals, accessible at http://www.inchem.org/pages/sids.html. IPCS, 2008. Screening information data set (SIDS) for high production volume chemicals. Accessible at http://www.inchem.org/pages/sids.html. ISO11261, 1997. Soil quality – determination of total nitrogen – modified Kjeldahl method (ISO 11261:1995). ISO11466, 1997. Soil quality – extraction of trace elements soluble in aqua regia (ISO 11466:1995). ISO11885, 2007. Water quality – determination of selected elements by inductively (ISO 11885-E22). Kruskal, J.B., Wish, M., 1978. Multidimensional Scaling, vol. 07-011. Sage Publications, Newbury Park, CA. Labrune, C., Amouroux, J., Sarda, R., Dutrieux, E., Thorin, S., Rosenberg, R., Grémare, A., 2006. Characterization of the ecological quality of the coastal Gulf of Lions (NW Mediterranean). A comparative approach based on three biotic indices. Marine Pollution Bulletin 52, 34–47. Liess, M., von der Ohe, P., 2005. Analyzing effects of pesticides on invertebrate communities in streams. Environmental Toxicology and Chemistry 24, 954–965. Marín-Guirao, L., Cesar, A., Marín, A., Lloret, J., Vita, R., 2005. Establishing the ecological quality status of soft-bottom mining-impacted coastal water bodies in the scope of the Water Framework Directive. Marine Pollution Bulletin 50 (4), 374–387.

Muniz, P., Venturini, N., Pires-Vanin, A., Tommasi, L., Borja, A., 2005. Testing the applicability of a marine biotic index (AMBI) to assessing the ecological quality of soft-bottom benthic communities, in the South America Atlantic region. Marine Pollution Bulletin 50, 624–637. Muxika, I., Ángel, B., Bald, J., 2007. Using historical data, expert judgment and multivariate analysis in assessing reference conditions and benthic ecological status, according to the European Water Framework Directive. Marine Pollution Bulletin 55, 16–29. Muxika, I., Ángel, B., Bonne, W., 2005. The suitability of the marine biotic index (AMBI) to new impact sources along European coasts. Ecological Indicators 5, 19–31. Nikitik, C.C.S., Robinson, A.W., 2003. Patterns in benthic populations in the Milford Haven waterway following the ‘Sea Empress’ oil spill with special reference to amphipods. Marine Pollution Bulletin 46, 1125–1141. Pearson, T., Rosenberg, R., 1978. Macrobenthic succession in relation to organic enrichment and pollution of the marine environment. Oceanography and Marine Biology Annual Review 16, 229–311. Pearson, T., Rosenberg, R., 1976. A comparative study of the effects on the marine environment of wastes from cellulose industries in Scotland and Sweden. Ambio 5, 77–79. Petermeier, A., Schöll, F., Tittizer, T., 1996. Die ökologische und biologische Entwicklung der deutschen Elbe. Ein Literaturbericht. Lauterbornia H 24, 1–95. Puente, A., Diaz, R.J., 2008. Is it possible to assess the ecological status of highly stressed natural estuarine environments using macroinvertebrates indices? Marine Pollution Bulletin 56 (11), 1880–1889. Quinn, G., Keough, M., 2002. Experimental Design and Data Analysis for Biologists. Cambridge University Press. 2008. R Development Core R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3900051-07-0. Ranasinghe, J.A., Weisberg, S.B., Smith, R.W., Montagne, D.E., Thompson, B., Oakden, J.M., Huff, D.D., Cadien, D.B., Velarde, R.G., Ritter, K.J., 2009. Calibration and evaluation of five indicators of benthic community condition in two California bay and estuary habitats. Marine Pollution Bulletin 59 (1–3), 5–13. Rathlev, J., 1980. Eine einfache Salzgehaltsformel für den praktischen Gebrauch, die der neuesten Unesco-Empfehlung entspricht. Deutsche Hydrographische Zeitschrift 3, 124–131. Ritz, D.A., Lewis, M.E., Shen, M., 1989. Response to organic enrichment of infaunal macrobenthic communities under salmonid seacages. Marine Biology 103, 211–214. Rolinski, S., Eichweber, G., 2000. Deformations of the tidal wave in the Elbe estuary and their effect on suspended particulate matter dynamics. Physics and Chemistry of the Earth (B) 25 (4), 355–358. Rosenberg, R., Blomqvist, M., Nilsson, H.C., Cederwall, H., Dimming, A., 2004. Marine quality assessment by use of benthic species-abundance distributions: a proposed new protocol within the European Union Water Framework Directive. Marine Pollution Bulletin 49 (9–10), 728–739. Rousseeuw, P., 1987. Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics 20, 53–65. Salas, F., Neto, J., Borja, A., Marques, J., 2004. Evaluation of the applicability of a marine biotic index to characterize the status of estuarine ecosystems: the case of Mondego estuary (Portugal). Ecological Indicators 4, 215–225. Schuchardt, B., Schirmer, M., Janssen, G., Nehring, S., Leuchs, H., 1999. In: Jong, F., de Bakker, J., van Berkel, C., Dahl, K., Dankers, N., Gatje, C., Marencic, H., Potel, P. (Eds.), Wadden sea quality status report 1999. Common Wadden Sea Secretariat, Trilateral Monitoring and Assessment Group, Quality Status Report Group, Wilhelmshaven, Germany, pp. 175–186. Schäfer, R., Caquet, T., Siimes, K., Mueller, R., Lagadic, L., Liess, M., 2007. Effects of pesticides on community structure and ecosystem functions in agricultural streams of three biogeographical regions in Europe. Science of the Total Environment 382, 272–285. Schirmer, M., 1994. In: Lozán, J., Rachor, E., Reise, K., von Westernhagen, H., Lenz, W. (Eds.), Ökologische Konsequenzen des Ausbaus der Ästuare von Elbe und Weser. In: Warnsignale aus dem Wattenmeer. Blackwell, Berlin, pp. 164–171. Simboura, N., Reizopoulou, S., 2007. A comparative approach of assessing ecological status in two coastal areas of Eastern Mediterranean. Ecological Indicators 7 (2), 455–468. Simboura, N., Zenetos, A., 2002. Benthic indicators to use in ecological quality classification of Mediterranean soft bottom marine ecosystems, including a new biotic index. Mediterranean Marine Science 3, 77–111. Sprague, J., 1970. Measurement of pollutant toxicity to fish, II. Utilizing and applying bioassay results. Water Research 4, 3–32. Teixeira, H., Salas, F., Pardal, M., Marques, J., 2007. Applicability of ecological evaluation tools in estuarine ecosystems: the case of the lower Mondego estuary (Portugal). Hydrobiologia 587, 101–112. Tomlin, C., 2003. The Pesticide Manual, A World Compendium. BCPC Publications, Hampshire, UK. 2008. ECOTOX 4.0 ecotoxicology database. Accessible at USEPA, http://cfpub.epa.gov/ecotox. von der Ohe, P., Kuhne, R., Ebert, R., Altenburger, R., Liess, M., Schuurmann, G., 2005. Structural alerts – a new classification model to discriminate excess toxicity from narcotic effect levels of organic compounds in the acute daphnid assay. Chemical Research in Toxicology 18, 536–555.

M.A. Wetzel et al. / Ecological Indicators 19 (2012) 118–129 von der Ohe, P., Liess, M., 2004. Relative sensitivity distribution of aquatic invertebrates to organic and metal compounds. Environmental Toxicology and Chemistry 23, 150–156. Warwick, R.M., 1986. A new method for detecting pollution effects on marine macrobenthic communities. Marine Biology 92, 557–562. Warwick, R., Clarke, K., 1994. Relearning the ABC – taxonomic changes and abundance. Marine Biology 118, 739–744.

129

Warwick, R.M., Pearson, T.H., Ruswahyuni, 1987. Detection of pollution effects on marine macrobenthos: further evaluation of the species abundance/biomass method. Marine Biology 95, 193–200. Wetzel, M.A., Leuchs, H., Koop, J.H.E., 2005. Preservation effects on wet weight, dry weight, and ash-free dry weight biomass estimates of four common estuarine macro-invertebrates: no difference between ethanol and formalin. Helgoland Marine Research 59 (3), 206–213.