A Marine Biotic Index to Establish the Ecological Quality of Soft-Bottom Benthos Within European Estuarine and Coastal Environments A. BORJA*, J. FRANCO and V. P EREZ Department of Oceanography and Marine Environment, Technological Institute for Fisheries and Food (AZTI), Av. Satr ustegui 8, 20008 San Sebasti an, Spain In this paper, a marine Biotic Index (BI) for soft-bottom benthos of European estuarine and coastal environments is proposed. This is derived from the proportions of indi- vidual abundance in five ecological groups, which are re- lated to the degree of sensitivity/tolerance to an environmental stress gradient. The main dierence with previously published indices is the use of a simple formula that produces a continuous Biotic Coecient (BC) – which makes it more suitable for statistical analysis, in opposition with previous discreet biotic indices – not af- fected by subjectivity. Relationships between this coe- cient and a complementary BI with several environmental variables are discussed. Finally, a validation of the pro- posed index is made with data from systems aected by recent human disturbances, showing that dierent an- thropogenic changes in the environment can be detected through the use of this BI. Ó 2000 Elsevier Science Ltd. All rights reserved. Keywords: biotic index; ecological quality; diversity; benthos; soft-bottom; European coastal environments. Introduction Marine environmental quality control is undertaken usually by means of monitoring dierent parameters in water, sediment and sentinel organisms (i.e. Mussel Watch), as in the USA (OÕConnor, 1992), France (RNO, 1998) or Great Britain (Franklin and Jones, 1994). This control is centred on physico-chemical and ecotoxico- logical variables and, less usually, on biological vari- ables. Dauer (1993) stated that biological criteria are considered important components of water quality be- cause: (i) they are direct measures of the condition of the biota, (ii) they may uncover problems undetected or underestimated by other methods; and (iii) such criteria provide measurements of the progress of restoration eorts. New European rules (see Directive Proposal 1999/C 343/01, Ocial Journal of the European Communities 30/11/1999) emphasize the importance of biological in- dicators, in order to establish the ecological quality of European coasts and estuaries. Benthic invertebrates are used frequently as bio-indicators of marine monitoring, because various studies have demonstrated that macro- benthos responds relatively rapidly to anthropic and natural stress (Pearson and Rosenberg, 1978; Dauer, 1993). River ecology has an established long tradition in applying macrobenthos as bio-indicators; likewise some biotic indices have been proposed (Woodiwiss, 1964; Cairns et al., 1968; Chandler, 1970; ISO-BMWP, 1979, etc.). On the other hand, some attempts to provide useful ‘tools’ to measure ecological quality in the marine environment have been developed in Europe and North America (Hily, 1984; Majeed, 1987; Dauer, 1993; Grall and Gl emarec, 1997; Weisberg et al., 1997). All the aforementioned studies utilize soft-bottom communities to construct the indices, because macro- benthic animals are relatively sedentary (and cannot avoid deteriorating water/sediment quality conditions), have relatively long life-spans (thus, indicate and inte- grate water/sediment quality conditions, with time), consist of dierent species that exhibit dierent toler- ances to stress and have an important role in cycling nutrients and materials between the underlying sedi- ments and the overlying water column (Hily, 1984; Dauer, 1993). In this contribution, a marine Biotic Index (BI) is designed to establish the ecological quality of European coasts. This explores the response of soft-bottom com- munities to natural and man-induced changes in water quality, integrating long-term environmental conditions. Marine Pollution Bulletin Vol. 40, No. 12, pp. 1100–1114, 2000 Ó 2000 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0025-326X/00 $ - see front matter PII: S0025-326X(00)00061-8 *Corresponding author. E-mail address: [email protected] (A. Borja). 1100
15
Embed
a Marine Biotic Index To Establish The Ecological - ECASA · A Marine Biotic Index to Establish the Ecological Quality of Soft-Bottom Benthos Within European Estuarine and Coastal
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A Marine Biotic Index to Establish theEcological Quality of Soft-BottomBenthos Within European Estuarine andCoastal EnvironmentsA. BORJA*, J. FRANCO and V. P�EREZDepartment of Oceanography and Marine Environment, Technological Institute for Fisheries and Food (AZTI),Av. Satr�ustegui 8, 20008 San Sebasti�an, Spain
In this paper, a marine Biotic Index (BI) for soft-bottombenthos of European estuarine and coastal environments isproposed. This is derived from the proportions of indi-vidual abundance in ®ve ecological groups, which are re-lated to the degree of sensitivity/tolerance to anenvironmental stress gradient. The main di�erence withpreviously published indices is the use of a simple formulathat produces a continuous Biotic Coe�cient (BC) ±which makes it more suitable for statistical analysis, inopposition with previous discreet biotic indices ± not af-fected by subjectivity. Relationships between this coe�-cient and a complementary BI with several environmentalvariables are discussed. Finally, a validation of the pro-posed index is made with data from systems a�ected byrecent human disturbances, showing that di�erent an-thropogenic changes in the environment can be detectedthrough the use of this BI. Ó 2000 Elsevier Science Ltd.All rights reserved.
Keywords: biotic index; ecological quality; diversity;benthos; soft-bottom; European coastal environments.
Introduction
Marine environmental quality control is undertakenusually by means of monitoring di�erent parameters inwater, sediment and sentinel organisms (i.e. MusselWatch), as in the USA (OÕConnor, 1992), France (RNO,1998) or Great Britain (Franklin and Jones, 1994). Thiscontrol is centred on physico-chemical and ecotoxico-logical variables and, less usually, on biological vari-ables. Dauer (1993) stated that biological criteria areconsidered important components of water quality be-cause: (i) they are direct measures of the condition of the
biota, (ii) they may uncover problems undetected orunderestimated by other methods; and (iii) such criteriaprovide measurements of the progress of restoratione�orts.
New European rules (see Directive Proposal 1999/C343/01, O�cial Journal of the European Communities30/11/1999) emphasize the importance of biological in-dicators, in order to establish the ecological quality ofEuropean coasts and estuaries. Benthic invertebrates areused frequently as bio-indicators of marine monitoring,because various studies have demonstrated that macro-benthos responds relatively rapidly to anthropic andnatural stress (Pearson and Rosenberg, 1978; Dauer,1993).
River ecology has an established long tradition inapplying macrobenthos as bio-indicators; likewise somebiotic indices have been proposed (Woodiwiss, 1964;Cairns et al., 1968; Chandler, 1970; ISO-BMWP, 1979,etc.). On the other hand, some attempts to provideuseful `tools' to measure ecological quality in the marineenvironment have been developed in Europe and NorthAmerica (Hily, 1984; Majeed, 1987; Dauer, 1993; Gralland Gl�emarec, 1997; Weisberg et al., 1997).
All the aforementioned studies utilize soft-bottomcommunities to construct the indices, because macro-benthic animals are relatively sedentary (and cannotavoid deteriorating water/sediment quality conditions),have relatively long life-spans (thus, indicate and inte-grate water/sediment quality conditions, with time),consist of di�erent species that exhibit di�erent toler-ances to stress and have an important role in cyclingnutrients and materials between the underlying sedi-ments and the overlying water column (Hily, 1984;Dauer, 1993).
In this contribution, a marine Biotic Index (BI) isdesigned to establish the ecological quality of Europeancoasts. This explores the response of soft-bottom com-munities to natural and man-induced changes in waterquality, integrating long-term environmental conditions.
Marine Pollution Bulletin Vol. 40, No. 12, pp. 1100±1114, 2000
Ó 2000 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
0025-326X/00 $ - see front matterPII: S0025-326X(00)00061-8
SamplingThe Department of Land Action, Housing and
Environment of the Basque Government has establisheda network of monitoring stations along the Basquecoast-line (North of Spain). This provides water, sedi-ment and biological quality information from 30 sam-pling stations (Fig. 1). The benthic sampling has beencarried out every February, from 1995 to 1998 using theresearch vessel `Ortze'.
At each of these stations, three replicates of benthoswere collected with a Van Veen grab (1215 cm2). Thesamples were ®ltered immediately, using a sieve of meshsize of 1 mm and ®xed in a solution of 4% formalin(Holme and McIntyre, 1971).
Sediment dataAt each station, a sediment sample was obtained to
determine redox potential, organic matter content andcontaminant levels (heavy metals and organic com-pounds). The redox potential was measured, on board,by means of an Orion 977800 platinum electrode whichwas connected to a Crison 501 pH-meter-milivoltimeter.
A 200 g sediment sample was dried at 80°C for 24 h,then it was washed with freshwater on a mesh of 63 lm.The dried residue was sieved on a column of eight sieves(size 31 lm to 4 mm). The percentages of gravel, sandand mud were calculated as: >2 mm fraction, 63 lm ± 2mm and <63 lm, respectively (Holme and McIntyre,1971).
The organic matter content was calculated by the losson ignition method: drying at 105°C, 24 h; then com-busting at 520°C, 6 h (Kristensen and Anderson, 1993).
Metal concentrations (As, Cd, Cu, Cr, Hg, Ni, Pb andZn) were analysed on the <63 lm fraction. Extractionwas made ®rst with nitric acid, during 15 h, at ambienttemperature; and second, with nitric and hydrochloricacids (1:3 in volume), using a microwave oven (130 W,4 min; 0 W, 1.5 min; 250 W, 5 min; 0 W, 2 min; 400 W,4 min). Detection was made by atomic absorption, using¯ame, graphite furnace and cold vapour techniques. Theanalytical procedure was checked with reference mate-rial (BCR marine sediment-harbour PACS-1); di�er-ences with this material were lower than 10%.
For PCB (eight congeners), DDT and HCH deter-mination a portion of the original sample was desiccatedwith anhydrous sodium sulphate and extraction wasmade with iso-octane, after conditioning and clean-up ofthe extract the analysis was made with an HP-5890 gaschromatograph. On the other hand, for PAH (10 com-pounds) determination, the extraction was made withether, and the analysis was made by means of HPLC.
Water quality dataThe mean bottom oxygen concentration was mea-
sured with a CTD Sea-Bird 25, or with a portable YSI-
55 oxymeter. Salinity was measured with the same CTD,or with a Kahlsico SR10 induction salinometer.
Biological dataThe identi®cation was undertaken in the laboratory
by means of a binocular microscope (4±40�). Aftercomputing the mean abundance of each taxon, at eachsampling station, the macrobenthic community struc-ture was described calculating the following descriptors(Washington, 1984): richness (number of identi®edtaxa); abundance (N: ind mÿ2); numerical diversity(Shannon Wiener H0n: bits indÿ1); biomass (Dry Weight,B: g mÿ2); and biomass diversity (Shannon Wiener H0b:bits gÿ1).
BI modelThe model here developed is based on that ®rst used
by Gl�emarec and Hily (1981) and then by Hily (1984),which utilizes soft-bottom benthos to construct a BI.
Soft-bottom macrobenthic communities respond toenvironmental stress (i.e. the introduction of organicmatter in the system) by means of di�erent adaptivestrategies. Gray (1979) summarizes these strategies intothree ecological groups: r (r-selected: species with shortlife-span, fast growth, early sexual maturation and lar-vae throughout the year); k (k-selected: species withrelatively long life, slow growth and high biomass); andT (stress tolerant: species not a�ected by alterations).
Salen-Picard (1983) has proposed four progressivesteps relating to stressed environments: (i) initial state(in an unpolluted situation, there is a rich biocenosis inindividuals and species, with exclusive species and highdiversity); (ii) slight unbalance (regression of exclusivespecies, proliferation of tolerant species, the appearanceof pioneering species, decrease of diversity); (iii) pro-nounced unbalance (population dominated by pollutionindicators, very low diversity); and (iv) azoic substrata.
Following these four steps, Hily (1984) and Gl�emarec(1986) have stated that the soft-bottom macrofaunacould be ordered in ®ve groups, according to their sen-sitivity to an increasing stress gradient (i.e. increasingorganic matter enrichment). Their concept is similar tothat developed for the Infaunal Index for SouthernCalifornia, described by Mearns and Word (1982) andFerraro et al. (1991). These groups have been summa-rized by Grall and Gl�emarec (1997), as outlined below.
Group I. Species very sensitive to organic enrichmentand present under unpolluted conditions (initial state).They include the specialist carnivores and some deposit-feeding tubicolous polychaetes.
Group II. Species indi�erent to enrichment, alwayspresent in low densities with non-signi®cant variationswith time (from initial state, to slight unbalance). Theseinclude suspension feeders, less selective carnivores andscavengers.
Group III. Species tolerant to excess organic matterenrichment. These species may occur under normalconditions, but their populations are stimulated by
organic richment (slight unbalance situations). They aresurface deposit-feeding species, as tubicolous spionids.
Group IV. Second-order opportunistic species (slightto pronounced unbalanced situations). Mainly smallsized polychaetes: subsurface deposit-feeders, such ascirratulids.
Group V. First-order opportunistic species (pro-nounced unbalanced situations). These are deposit-feeders, which proliferate in reduced sediments.
The distribution of these ecological groups, accordingto their sensitivity to pollution stress, provides a BI witheight levels, from 0 to 7 (Hily, 1984; Hily et al., 1986;Majeed, 1987).
In the aforementioned monitoring network of sam-pling stations, together with other studies developed byAZTI along the Basque coastline within the last ®veyears (Borja et al., 1995, 1999a,b), more than 900 taxahave been identi®ed. These species are representative ofthe most important soft-bottom communities present atEuropean estuarine and coastal systems. The taxa havebeen classi®ed (list in Appendix A) according to theabove ecological groups, following Majeed (1987),Dauer (1993), Weisberg et al. (1997), Grall andGl�emarec (1997) and Roberts et al. (1998). Only about
12% of the taxa have not been possible to be assigned toan ecological group.
Based upon HilyÕs model (Hily, 1984; Hily et al., 1986;Majeed, 1987), Fig. 2 shows the theoretical distributionof relative abundance of each ecological group, along apollution gradient.
A possible limitation in the utilisation of the model ofHily is that each BI has a discreet value and its calcu-lation is not systematized. In order to improve the index,a single formula is proposed here. This is based upon thepercentages of abundance of each ecological group,within each sample, to obtain a continuous index (theBiotic Coe�cient (BC)), where
The above-mentioned ecological groups (GI, GII,GIII, GIV and GV) are summarized in Table 1. Speciesnot assigned to a group were not taken into account.These species represent only a mean abundance of 1.4%,for the total number of samples.
In this way, use of the BC can derive a series ofcontinuous values, from 0 to 6, being 7 when the sedi-ment is azoic. Nonetheless, the BC can be compared tothe Grall and Gl�emarec (1997) BI, as adapted in thispaper (Table 1). The result obtained is a `pollutionclassi®cation' of a site which is a function of the BC.Consequently, this represents the benthic community`health', represented by the entire numbers of the BI.
Results
The mean and standard error values of grain size andphysical characteristic associated with each of the sam-pling stations (17 estuarine and 13 littoral) are listed inTable 2. The water depth range is very large at each ofthe stations (under Mean High Water Neap to 24 m inthe estuaries and 30±35 m associated with the littoralsamples). Mean salinity, at bottom water, ranges from16.2 to 35.3 in estuaries, but is restricted within thecoastal areas (35.3±35.5).
The range in the percentage of oxygen saturation isvery high within the estuaries (43±119%), but ranges inthe littoral stations from 92% to 97%. The organic
Fig. 2 Theoretical model, modi®ed from Hily (1984), Hily et al. (1986)and Majeed (1987), which provides the ordination of soft-bottom macrofauna species into ®ve ecological groups (GroupI: species very sensitive; Group II: species indi�erent; GroupIII: species tolerant; Group IV: second-order opportunisticspecies; Group V: ®rst-order opportunistic species), accordingto their sensitivity to an increasing pollution gradient. Therelative proportion of abundance of each group in a sampleprovides a discreet BI with eight levels (0±7) and an equivalentcontinuous BC (values between 0 and 6).
TABLE 1
Summary of the BC and BI (modi®ed from Grall and Gl�emarec, 1997).
Site pollution classi®cation Biotic Coe�cient Biotic index Dominating ecological group Benthic community health
Unpolluted 0:0 < BC � 0:2 0 I NormalUnpolluted 0:2 < BC � 1:2 1 ImpoverishedSlightly polluted 1:2 < BC � 3:3 2 III UnbalancedMeanly polluted 3:3 < BC � 4:3 3 Transitional to pollutionMeanly polluted 4:5 < BC � 5:0 4 IV±V PollutedHeavily polluted 5:0 < BC � 5:5 5 Transitional to heavy pollutionHeavily polluted 5:5 < BC � 6:0 6 V Heavy pollutedExtremely polluted Azoic 7 Azoic Azoic
1103
Volume 40/Number 12/December 2000
matter content in the sediments is higher in the estuaries(2.3±28.2%) than in littoral zone (3.4±6.8%). This cor-responds to a higher range of the mud content within thesediments (0.3±80.1% and 0.1±17.3%, respectively). Theredox potential ranges from )185 to 410 mV within theestuaries, and from )84 to 405 mV within the littoralsamples.
From 30 stations, some 114 samples of benthos havebeen obtained over a 4 year period. These samplescorrespond to di�erent environments (estuarine, littoral,intertidal, subtidal) and physico-chemical characteristics(reduced and oxidized sediments, hypoxia and oversat-uration in the bottom waters, poor organic matterproportion and enrichment, etc.).
After the application of the BC, considering its cor-respondence with the BI (Table 1), the results were: 2samples with a BI� 0; 23 samples of BI� 1; 48 samplesof BI� 2; 15 samples of BI� 3; 7 samples of BI� 4; 6samples of BI� 5; 6 samples of BI� 6; and 7 samples ofBI� 7.
Fig. 3 shows the results obtained by comparing dif-ferent biological parameters, on samples having thesame biotic indices. The BI� 7 is equivalent to an azoicsite, so all the biological parameters are equal to 0 inthese particular samples.
The mean abundance increases from 36.7 ind mÿ2
(BI� 0) to 2 559 ind mÿ2 (BI� 6), with the exception ofBI� 5, with a value of 456 ind mÿ2 (Fig. 3(a)). Withinthe lowest of the Biotic Indices (0, 1 and 2), the standarderror of the mean is very small; it is progressively largerin the highest.
Statistical analyses were made considering the BCbecause, as this coe�cient can derive continuous values,it is more suitable for this purpose than the BI. Takinginto account all the samples analysed, the non-para-metric Spearman rank correlation between the abun-dance and the BC is not statistically signi®cant(p > 0:05).
On the other hand, biomass (Fig. 3(b)) increases from0.1 g mÿ2 (BI� 0) to 14.3 g mÿ2 (BI� 4). However forBiotic Indices 5 and 6, dominated by small opportunisticspecies, the biomass is lower than 4 g mÿ2. There is no astatistically signi®cant correlation between biomass andBC (Spearman rank correlation).
Fig. 3(c) shows the mean richness of the samples.Except in the case of BI� 0, with a mean richness of 2,in the other biotic indices the richness decreases pro-gressively from 26 to 27 species (BI� 1 and 2) to 0species (BI� 7). Richness and BC are highly correlated(p < 0:001, Spearman rank correlation).
TABLE 2
Physico-chemical characterisation of sampling stations, showing mean and standard error (SE) values of some sedimentological and waterparameters.a
Stationnumber
Stationtype
Depth(m)
Salinity Dissolvedoxygen (ml lÿ1)
% Oxygensaturation
% Sand % Mud % Organicmatter
Redoxpotential (mV)
Mean Mean� SE Mean� SE Mean� SE Mean� SE Mean� SE Mean� SE Mean� SE
a E: estuarine site; L: littoral site; I: intertidal site.
1104
Marine Pollution Bulletin
Numerical diversity (Fig. 3(d)) shows a similar patternto that of richness. There is a progressive decrease in themean values, from 3.5 bits indÿ1 (BI� 1) to 0 bits indÿ1
(BI� 7), with the exception of BI� 0 which is associatedwith a low value (0.6 bits indÿ1). Biomass diversity hasvalues of about 1.6 bits gÿ1, between BI� 1 to 4; then, itdecreases to 0 (BI� 7). Both variables are correlatedwith BC (p < 0:001 and p < 0:05 for numerical andbiomass diversities, respectively), using Spearman rankcorrelation.
The relationships between some of the sedimento-logical and water quality parameters and biotic indices
are shown in Fig. 4. BI� 0 is associated with the highestmean redox potential (Fig. 4(a): 360 mV). This param-eter becomes progressively lower, with BI� 7 having amean potential which is very reduced (±125 mV). Sam-ples with low biotic indices (0 and 1) are associated withless than 2% of mud (Fig. 4(b)) and the values increaseto 63% (BI� 7). Some anomalies were detected inBI� 5 and 6, which present 10±20% of mud. The or-ganic matter content has a similar pattern of distribu-tion to that of granulometry (Fig. 4(c)). Data on themean bottom dissolved oxygen content are presented inFig. 4(d). The highest value corresponds to BI� 0
Fig. 4 Mean and standard error values of di�erent sedimentologicaland water quality parameters obtained on samples having thesame biotic indices. (a) redox potential; (b) percentage of mud;(c) organic matter content; and (d) bottom dissolved oxygencontent.
Fig. 3 Mean and standard error values of di�erent biologicalparameters obtained on samples having the same biotic indices.(a) abundance; (b) biomass; (c) richness; and (d) diversity(derived from number of individuals N and biomass B).
1105
Volume 40/Number 12/December 2000
(6.5 ml lÿ1), decreasing to 2.6 ml lÿ1 at BI� 7. TheSpearman rank correlations between these variables andBC are highly signi®cant (p < 0:001).
The mean concentrations relating to some of theheavy metals in the sediments associated to each BI areshown in Fig. 5. Arsenic and mercury contents do notreveal any clear pattern of distribution with the BI.Other metals present increasing concentrations fromBI� 0 to BI� 7, with the exception of some speci®cpeaks (BI� 3, for chromium and nickel; and BI� 6, forlead and copper) and troughs (BI� 4 and 5, for cad-mium, nickel and zinc). Except arsenic and mercury allthe metals are positively correlated with BC (p < 0:01,Spearman rank correlations).
On the other hand, the organic compounds (Fig. 6) donot show a similar pattern to that of the metals. OnlyPCB increase in their concentrations from BI� 0 toBI� 7; however the di�erences are very small. PAH is attheir smallest concentrations in BI� 5 and 6. The onlysigni®cant correlation is found between BC and PCB(p < 0:05, Spearman rank correlation).
Comparing the percentage of samples of each BI thatgoes beyond the ER-L (or E�ects Range-Low, repre-senting concentrations below which adverse e�ects tofauna are expected to occur rarely (Long et al., 1995)),the data presented in Fig. 7 shows that BI� 0 does notinclude samples that surpass these limits for metals andorganic compounds. Normally, the other biotic indices
Fig. 5 Mean and standard error values of eight heavy metal contentsin sediments obtained on samples having the same bioticindices.
Fig. 6 Mean and standard error values of four organic compoundcontents in sediments obtained on samples having the samebiotic indices.
1106
Marine Pollution Bulletin
increase progressively in the percentage of samples sur-passing these limits (see data presented for arsenic,mercury, nickel, lead, copper, chromium, PCB andDDT).
Discussion
Many of the biotic indices developed in the literature(Clements et al., 1992; Mouthon, 1993; Stark, 1993;Grall and Gl�emarec, 1997; Roberts et al., 1998, etc.)have been based on the paradigm of Pearson and Ro-senberg (1978), as stated by Weisberg et al. (1997) indeveloping their own index. The paradigm states thatbenthic communities respond to improvements in hab-itat quality in three progressive steps: the abundanceincreases; species diversity increases; and dominantspecies change from pollution-tolerant to pollution-sensitive species.
This generally accepted paradigm has been adaptedfrom Grall and Gl�emarec (1997) in this contribution, inorder to obtain an European BI. This should be able todistinguish easily estuaries and coastal reference sitesfrom polluted sites, with di�erent levels of anthropo-genic or natural degradation.
The index derived provides a semi-quantitative mea-surement of the degree of impact on soft-bottom mac-rofauna, which is re¯ected by changes in the qualitativeand quantitative community composition.
As the BI has been established on the basis of analysisof samples obtained from a monitoring network, with aprevalence of polluted sites, there are only two unpol-luted samples (BI� 0) which correspond to a `normal'community (sensu Grall and Gl�emarec, 1997). The di-versity results do not correspond to those expected fromthe aforementioned paradigm, because the richness is
very low. Conversely, samples with BI� 1 (also unpol-luted in the present proposal, corresponding to animpoverished community) or higher, BI� 2±6 (corre-sponding to slightly to heavily polluted sites) havewell-de®ned values of biological parameters; this is asmight be expected from the results of Pearson andRosenberg (1978).
Some biotic indices, or Coe�cients of Pollution (i.e.Bogdanos and Satsmadjis, 1985) do not appear to besuitable for application in some cases. This is due to thelack of sensitivity of these indices to intermediate pol-lution levels (MAFF, 1993), corresponding with slightlypolluted areas. Hily (1984) and Grall and Gl�emarec(1997) have described similar di�culties.
The above limitation appears to be due to a generalunder-estimation of the faunal abundance in compari-son with unpolluted areas. This is because faunalabundance will increase under slight to moderate pol-lution, but numbers of species can either stay constantor show only a slight increase. In the present proposal,this problem appears to be eliminated because the ap-proach has a high sensitivity at these levels, with well-de®ned values in the biological parameters.
Organic enrichment and muddy bottoms, associatedwith subsequent low redox potential and hypoxia, arerelated with opportunistic species (Majeed, 1987) in`heavily polluted' levels, according to the BI (BI� 5±7).Diaz and Rosenberg (1995) have suggested that benthicinfaunal mortality could be initiated when the oxygenconcentration falls below 2 ml lÿ1. Ritter and Montagna(1999) have recently proposed that 3 mg lÿ1 (� 2.14 mllÿ1) de®nes the breakpoint between normoxic and hy-poxic benthic communities. The mean oxygen concen-tration obtained for BI� 7 indicates that life could bevery limited in those sites. However, within BI� 6, there
Fig. 7 Percentage of samples of each biotic index that goes beyond theER-L (or E�ects Range-Low, representing concentrationsbelow which adverse e�ects are expected to occur rarely), forseven heavy metals and three organic compounds.
1107
Volume 40/Number 12/December 2000
are some situations of very low oxygen concentrationwhich explain the presence of species which are resistantto severe or moderate hypoxia. These species are clas-si®ed within ecological Groups IV and V.
Samples with BI� 6 and 7 are associated with sitesthat experience periodic hypoxia, consisting of repeatedbrief periods (days or weeks, in the case of BI� 6) orseasonal hypoxia (months, in the case of BI� 7), thatgenerate mass mortality or complete elimination of themacrofauna. Some of the samples with BI� 7 arelocated within the Bilbao estuary, for which S�aiz-Salinas(1997) and Gonz�alez-Oreja and S�aiz-Salinas (1998) havedemonstrated that the oxygen limitation represents thekey factor in the estuarine defaunation of samplingstations within the estuary.
Physico-chemical results related to the BI (see Fig. 4)have some unexpected results at the level of BI� 5 and6. The trend of increasing percentages of mud and or-ganic matter, together with decreasing redox potential,break-down at these particular levels. BI� 5 and 6correspond to high percentages of ecological Group V(with a mean of 77.5% of species in BI� 5, together with92.7% in BI� 6). These species are mainly deposit-feeders. As such, they could modify the proportion oforganic matter in the sediments on which they feed and,subsequently, modify the grain size composition of thesediments. The optimal grain size may be di�erent forthe settling larvae, juveniles and adults of a variety ofdeposit-feeders (Snelgrove and Butman, 1994), changingtheir physico-chemical properties. For example, Hall(1994) has stated that faecal pellets of benthic inverte-brates modify the grain size of the sur®cial sediments.
In spite of the fact that hypoxia seems to control thepresence of the groupings with BI� 7 and that organicmatter content is very important in ascribing samples tothe BI, ecotoxicological e�ects appear to play only a
secondary role in the analyses; however it may have hadan e�ect in the longer time, as cited by S�aiz-Salinas(1997) for the Bilbao Estuary.
In order to validate the derived BI, for more generalapplication, four stations from the 30 stations sampledhave been selected for more detailed analysis. The evo-lution of the percentage of ecological grouping, the BIand the BC, derived for between 1995 and 1998, at thesestations is shown in Fig. 8.
Within the Urdaibai estuary (Figs. 1 and 8(a)) theresults obtained from a single station, in the inner partof the estuary, shows a dominance of ecological GroupIII. This is characteristic of estuarine communities lo-cated at sites with organic matter inputs. The derived BIshows that, in 1995 and 1996, the site is slightly polluted(BI� 2) due mainly to the aforementioned organicmatter enrichment. In 1997 and 1998, the BI increasesprogressively (BI� 3 and 4); as such classifying thisstation as `meanly polluted'. Such a trend is caused bythe increasing dominance of ecological Group V, whichindicates the presence of opportunistic species. On theother hand, the BC increases gradually with time, whichindicates a rising contamination in this site during thelast years.
The increase in the BI could be the result of dredgingactivities undertaken along this particular estuary,within the last few years. At the same time, there arechanges in the sediment composition, the abundance ofsuspended matter, etc. These provide the basis for anincrease in the opportunistic species at this particularlocation.
In February 1995, the station in the Oria estuary(Figs. 1 and 8(b)), was located some 500 m landward ofthe mouth. Group I was dominant and the BI (2) pro-vides a classi®cation of the site as `slightly polluted'. In1995 and 1996, some channelling works were under-
Fig. 8 Evolution of the percentage of ecological groups (I±V), the BIand the BC, derived for between 1995 and 1998, for the stationsshowed in Fig. 1: Station 10; Station 21; Station 13; and Station24.
1108
Marine Pollution Bulletin
taken in this estuary, extending the mouth of the estuarysome 500 m o�shore. This development has led to anincrease in the distance from the mouth of the estuary tothe sampling station, with a subsequent change in thephysico-chemical conditions (Borja et al., 1999a). Thischange resulted in an increase in the mud and organicmatter content, together with a decrease in dissolvedoxygen. This change in the physical setting provides anexplanation for the increase in the dominance of eco-logical Groups III and V (more characteristic at theinner part of the estuary), modifying the BI to `meanlypolluted' (3). The BC increased during this period, from1.4 to 3.5.
The Lea is a small estuary within the Basque Country(Figs. 1 and 8(c)) which, in the past 4 years, has beensubjected to a sewerage plan, eliminating urban andindustrial e�uent discharges into the estuarine waters.The estuary was dominated by the opportunistic GroupV in 1995, with a BI of 3 (meanly polluted). Followingthe introduction of the sewerage scheme, the ecologicalGroup I, composed of species that are sensitive to pol-lution, increased in its dominance. This represented, in1997 and 1998, nearly 100% of the community.Throughout these two last years, the BI is 0. The BCdecreases from 4.2 in 1995 to 0.4 in 1996 and near 0 in1997 and 1998. So, in the last two years this station canbe classi®ed as an unpolluted site.
Finally, within the coastal area of Momp�as, near SanSebasti�an (Figs. 1 and 8(d)), there is an importantchange in the ecological group composition between1995 and 1996. At the beginning of this period, there is aco-dominance of Groups V, I and II. However, there is aclear dominance of Group V from 1996 to 1998. At thesame time, the BI changes from 2 (slightly polluted) to5±6 (heavily polluted). The BC, which was 3.0 in 1995,increased to values between 5.3 and 5.9 during the lastthree years. This particular coastal area has receivedlarge amounts of domestic and industrial waste from theSan Sebasti�an area since the 1970s. Further, in 1995 and1996, some sewerage works were constructed and animportant volume of urban and industrial polluted wa-ters, derived from nearby areas such as Pasajes orTximistarri, were diverted to Momp�as. The waste in-cludes contaminants (heavy metals and organic com-pounds) and a high amount of organic matteroriginating from the paper manufacturers (Franco et al.,1999).
Conclusions
The BC, proposed here as a BI to establish the eco-logical quality of the soft-bottom benthos within theEuropean coastal environments, takes into account thefaunal composition. As such, it ascribes each species toan ecological grouping, according to their sensitivity toan increasing stress gradient.
The di�erent composition, in terms of the abundanceof the various ecological groups in these samples pro-
vides a continuous BC (with values between 0 and 6).This is referenced to a BI, representing quality of bot-tom conditions in a discreet range from 0 (unpolluted)to 7 (extremely polluted). This composition is governedby the physico-chemical factors within the sedimentsand the overlying water column in terms of: organicmatter content; percentage of mud within the sediments;dissolved oxygen content within the bottom waters; andthe concentration of pollutants.
Biological parameters (such abundance, richness,biomass or diversity) provide a useful (and morebroadly applicable) description of each level of the BI. Itis considered (as described by Dauer, 1993) that bio-logical criteria may complement toxicity and chemicalassessment methods, to serve as independent evaluationsof the ecological quality of marine and estuarine eco-systems.
Validation of the model developed shows that di�er-ent anthropogenically changes in the environment canbe detected through the use of the BI, including altera-tions to the natural system such as dredging, engineeringworks, sewerage plans and the dumping of pollutedwaters. On the other hand, the BC provides a moreaccurate view of the evolution of the ecological status ofa particular location. Further, the fact that this partic-ular coe�cient can derive continuous values makes itmore suitable for application to statistical analysis thanthe BI (i.e. temporal trend analysis).
The BI proposed here is relatively simple and can,meaningfully, be applied when attempting to determinethe ecological status of European coastlines. Althoughthis index has been developed in the Bay of Biscay, themethodology can be applied for other European coastalareas, only conditioned by the assignation of the speciesto the ecological groups described here. In fact, many ofthe species compiled in the Appendix A are present inNorth Sea and Mediterranean. So, the index may beimproved with the contributions of newly assignedspecies from these seas and further examples of its moregeneral application and validation.
Finally, this index facilitates the understanding ofcomplex benthic data, summarizing a considerableamount of ecological information into a single repre-sentative value.
This study was supported by di�erent contracts undertaken betweenthe Department of Land Action, Housing and Environment of theBasque Government and AZTI. One of the authors (V. P�erez) wassupported by a grant from the Department of Agriculture and Fishingof the Basque Government. We thank the sta� of the Department ofOceanography of AZTI for their assistance during the ®eld samplingand laboratory analyses and the INSUB Group that was charged ofthe taxonomical analysis. We wish to thank also Professor MichaelCollins (School of Ocean and Earth Science, University of South-ampton, UK) and an anonymous referee for kindly advising us onsome details of this paper.
Bogdanos, C. and Satsmadjis, J. (1985) Quantitative e�ect of sedimentcoarseness and depth on the macrobenthos of an unpolluted andclosed Mediterranean Gulf. Revue Internationale d' Oceanographiemedicale 77/78, 74±85.
Borja, A., Valencia, V., Garc�õa, L. and Arresti, A. (1995) Lascomunidades bent�onicas intermareales y submareales en SanSebasti�an-Pasajes (Guip�uzcoa, N de Espa~na). Actas IV ColoquioInternacional de Oceanograf�õa del Golfo de Vizcaya (Santander)165±181.
Borja, A., Franco, J., Belzunce, M. J. and Valencia, V. (1999a) Red deVigilancia y Control de la calidad de las aguas litorales del Pa�õsVasco (oto~no 1997±verano 1998). Departamento de Ordenaci�on delTerritorio, Vivienda y Medio Ambiente, Gobierno Vasco, 333pp+appendices.
Borja, A., Belzunce, M. J., Franco J. and Castro, R. (1999b)Seguimiento ambiental de los estuarios del Nervi�on, Barbad�un yButr�on durante 1998. Consorcio de Aguas Bilbao-Bizkaia, 228pp+appendices.
Cairns, J., Douglas, W. A., Busey, F. and Chaney, M. D. (1968) Thesequential comparison index ± a simpli®ed method for non-biologists to estimate relative di�erences in biological diversity instream pollution studies. Journal of Water Pollution Control. Fed 40,1607±1613.
Chandler, J. R. (1970) A biological approach to water qualitymanagement. Water Pollution Control 69, 415±422.
Clements, W. H., Cherry, D. S. and Van Hassel, J. H. (1992)Assessment of the impact of heavy metals on benthic communitiesat the Clinch River (Virginia): evaluation of an index of communitysensitivity. Canadian Journal of Fisheries and Aquatic Sciences 49,1686±1894.
Dauer, D. M. (1993) Biological criteria, environmental health andestuarine macrobenthic community structure. Marine PollutionBulletin 26 (5), 249±257.
Diaz, R. J. and Rosenberg, R. (1995) Marine benthic hypoxia: a reviewof its ecological e�ects and the behavioural responses of benthicmacrofauna. Oceanography and Marine Biology Annual Review 33,245±303.
Ferraro, S. P., Swartz, R. C., Cole, F. A. and Schults, D. W. (1991)Temporal changes in the benthos along a pollution gradient:discriminating the e�ects of natural phenomena from sewage-industrial wastewater e�ects. Estuarine, Coastal and Shelf Science33, 383±407.
Franco, J., Borja, A., Belzunce, M. J. Valencia, V. (1999) Campa~na demedici�on de variables biol�ogicas y f�õsico-qu�õmicas en el estuario del r�õoOiartzun y �area costera pr�oxima a cala Murgita. Departamento deObras Hidr�aulicas y Urbanismo de la Diputaci�on Foral deGipuzkoa, 197 p.+appendices.
Franklin, A. and Jones, J. (1994) Monitoring and surveillance of non-radiactive contaminants in the aquatic environment and activitiesregulating the disposal of wastes at sea, 1992. Aquatic EnvironmentMonitoring Report, MAFF 40, 1±83.
Gl�emarec, M. (1986) Ecological impact of an oil-spill: utilisation ofbiological indicators. IAWPRC-NERC Conference, July 1985.IAWPRC Journal 18, 203±211.
Gl�emarec, M. and Hily, C. (1981) Perturbations apport�ees �a lamacrofaune benthique de la baie de Concarneau par les e�uentsurbains et portuaires. Acta Oecologica Oecologia Applicata 2, 139±150.
Gonz�alez-Oreja, J. A. and S�aiz-Salinas, J. I. (1998) Exploring therelationships between abiotic variables and benthic communitystructure in a polluted estuarine system. Water Research 32 (12),3799±3807.
Grall, J. and Gl�emarec, M. (1997) Using biotic indices to estimatemacrobenthic community perturbations in the Bay of Brest.Estuarine, Coastal and Shelf Science 44 (suppl. A), 43±53.
Gray, J. S. (1979) Pollution-induced changes in populations. Philo-sophycal Transactions of the Royal Society of London. Series B, 286,545±561.
Hall, S. J. (1994) Physical disturbance and marine benthic communi-ties: life in unconsolidated sediments. Oceanography and MarineBiology Annual Review 32, 179±239.
Hily, C. (1984) Variabilit�e de la macrofaune benthique dans les milieuxhypertrophiques de la Rade de Brest. Th�ese de Doctorat dÕEtat,Univ. Bretagne Occidentale. Vol. 1, 359 pp; Vol. 2, 337 pp.
Hily, C., Le Bris, H. and Gl�emarec, M. (1986) Impacts biologiques des�emissaires urbains sur les �ecosyst�emes benthiques. Oceanis 12, 419±426.
Holme, N. A. and McIntyre, A.D. (1971) Methods for the Study ofMarine Benthos, p. 387. Blackwell, Oxford.
ISO-BMWP (1979) Assessment of the Biological Quality of Rivers by aMacroinvertebrate Score. ISO/TC147/SC5/WG6/N5, InternationalStandards Organisation, p. 18.
Kristensen, E. and Andersen, F. O. (1993) Determination of organiccarbon in marine sediments: a comparison of two CHN-analyzermethods. Journal of Experimental Marine Biology and Ecology 109,15±23.
Long, E. R., MacDonald, D. D., Smith, S. L. and Calder, F. D. (1995)Incidence of adverse biological e�ects within ranges of chemicalconcentrations in marine and estuarine sediments. EnvironmentalManagement 19, 81±97.
MAFF (1993) Analysis and interpretation of benthic community dataat sewage-sludge disposal sites. Aquatic Environmental MonitoringReport,MAFF Directorate of Fisheries Research, Lowestoft, Vol. 37,80 pp.
Majeed, S. A. (1987) Organic matter and biotic indices on the beachesof North Brittany. Marine Pollution Bulletin 18 (9), 490±495.
Mearns, A. J. and Word, J. Q. (1982) Forecasting e�ects ofsewage solids on marine benthic communities. In Ecological Stressand the New York Bight: Science and management ed. G. F. Mayer,pp. 495±512. Estuarine Research Federation, Columbia, SouthCarolina.
Mouthon, J. (1993) Un indice biologique lacustre bas�e sur lÕexamendes peuplements de mollusques. Bulletin Francßaise de la Peche et laPiscicul 331, 397±406.
OÕConnor, T. P. (1992) Recent trends in coastal environmental quality:results from the ®rst ®ve years of the NOAA Mussel Watch project.US Department of Commerce, NOAA, Rockville, 45 pp.
Pearson, T. and Rosenberg, R. (1978) Macrobenthic succession inrelation to organic enrichment and pollution of the marineenvironment. Oceanography and Marine Biology Annual Review16, 229±311.
Ritter, C. and Montagna, P. A. (1999) Seasonal hypoxia and model ofbenthic response in a Texas bay. Estuaries 22 (1), 7±20.
Roberts, R. D., Gregory, M. R. and Fosters, B. A. (1998) Developingan e�cient macrofauna monitoring index from an impact study ± Adredge spoil example. Marine Pollution Bulletin 36 (3), 231±235.
R.N.O. (1998) Surveillance du Milieu Marin. Travaux du R�eseauNational d'Observation de la Qualit�e du Milieu Marin. Edition1998. Ifremer et Minist�ere de lÕAm�enagement du Territoire et delÕEnvironnement, 52 pp.
S�aiz-Salinas, J. I. (1997) Evaluation of adverse biological e�ectsinduced by pollution in the Bilbao estuary. Environmental Pollution96 (3), 351±359.
Salen-Picard, C. (1983) Sch�emas dÕ�evolution dÕune bioc�enose macro-benthique du substrat meuble. Comptes Rendus de l'Academie desSciencies de Paris 296, 587±590.
Snelgrove, P. V. R. and Butman, C. A. (1994) Animal-sedimentrelationships revisited: cause versus e�ect. Oceanography andMarine Biology Annual Review 32, 111±177.
Stark, J. D. (1993) Performance of the macroinvertebrate communityindex: e�ects of sampling method, sample replication, water depth,current velocity and substratum on index values. New ZealandJournal of Marine and Freshwater Research 27, 463±478.
Washington, H. G. (1984) Diversity, biotic and similarity indices. Areview with special relevance to aquatic ecosystems.Water Research18, 653±694.
Weisberg, S. B., Ranasinghe, J. A., Dauer, D. M., Scha�ner, L. C.,Diaz, R. J. and Frithsen, J. B. (1997) An estuarine benthic indexof biotic integrity (B-IBI) for Chesapeake Bay. Estuaries 20 (1),149±158.
Woodiwiss, F. S. (1964). The biological system of stream classi®cationused by the Trent River Board. Chemistry Industries. 11, 443±447.