Ecological Indicators 7 (2007) 455–468

This article is also available online at:

www.elsevier.com/locate/ecolind

A comparative approach of assessing ecological status

in two coastal areas of Eastern Mediterranean
N. Simboura *, S. Reizopoulou
Hellenic Centre for Marine Research, P.O. Box 712, Mavro Lithari, GR-19013 Anavissos, Greece
Received 21 February 2006; received in revised form 8 May 2006; accepted 20 May 2006

Abstract

This work presents an assessment of the ecological quality status of two marine coastal areas in the Aegean Sea (Greece,
Eastern Mediterranean) based on the benthic macroinvertebrate quality element. S. Evvoikos and Thessaloniki gulfs, two coastal
areas subjected to slight and heavier anthropogenic pressures, were selected to test the biotic index Bentix developed for the
assessment of the ecological status. Other ecological indicators, such as the Shannon diversity index (H0 ), the species richness (S)
and the AMBI biotic index were also applied and evaluated comparatively. Faunistic data were also used to interpret results. The
resulting classification was validated with the use of physicochemical parameters and pressure information. This work also
provides an insight into the structure of the Bentix classification scheme within the scope of its use in Water Framework
Directive (WFD) implementation.
# 2006 Elsevier Ltd. All rights reserved.

Keywords: Water Framework Directive; Macroinvertebrates; Indices; Classification; Ecological status

1. Introduction intercalibrated in order to reach a common basis on the

The Ecological Quality Status (EcoQS) of coastal
waters has become a priority issue after the legislative
implementation of the European Water Framework
Directive (WFD, 2000/60/EC; EC, 2000). Water bodies
or areas with ecological quality assessed as lower than
good have to be restored to ‘‘good’’ status up to the year
2015. Various national methods for classification of
ecological status have been developed, but all have to be
* Corresponding author. Fax: +30 2291076347.
E-mail address: msim@ncmr.gr (N. Simboura).
interpretation of the ecological status (EC, 2003b).
For the biological quality element of macroinver-
tebrates recent indices and classification schemes
include the biotic indices AMBI of AZTI marine
laboratory (Borja et al., 2000) and the biotic index
Bentix (Simboura and Zenetos, 2002). Both indices
are based on the concept of indicator species and
subjectively classify species into various ecological
groups based on the international literature and
experience. In this work, the two areas of S. Evvoikos
gulf and Thessaloniki gulf were selected on the basis
of the known pressures imposed on the environment

1470-160X/$ – see front matter # 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ecolind.2006.05.003
456 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
and estimated as covering the whole range from high
to moderate status.
The second set of study stations was selected in the
periphery of Attiki located mostly in S. Evvoikos gulf
(Marathon, Porto Rafti, Keratea) while some of the
stations around the turn of Attiki are located in
Saronikos Gulf (Anavissos, Saronida, Agia Marina).
The area, so forth referred to as S. Evvoikos gulf, is a
wide area directly open to the South Aegean Sea,
which presents the sedimentological properties of
open seas. The anthropogenic impact from land-based
sources in the gulf originates from a number of coastal
towns, ports and summer resorts, which are scattered
along the coasts of the periphery of Attiki. Other
sources of environmental disturbance in the area
include small scale fisheries.
Thessaloniki Gulf (N. Aegean Sea, Eastern
Mediterranean) (Fig. 1), is an eutrophicated enclosed
gulf, with reduced water flux, receiving high nutrient
Fig. 1. Location of the sampling sites in the two study areas.
inputs from rivers, urban and industrial runoff (Gotsis-
Skretas and Friligos, 1990; Balopoulos and Friligos,
1993; Moncheva et al., 2001), while the shellfish
farming activity reaches about 85% of the total Greek
production (Karageorgis et al., 2005).
2. Material and methods
Benthic samples were collected from 14 stations
located in S. Evvoikos gulf around Attica and from 6
stations located in the Gulf of Thessaloniki (Fig. 1).
Samples from S. Evvoikos gulf were collected during
a summer sampling trip in July 1996 using a Van Veen
0.1 m2 grab taking two replicates at each station and in
Thessaloniki gulf samples were collected in December
2002 using o Ponar 0.045 m2 grab sampling three
replicates at each station. Samples were sieved on
board with a 1 mm mesh size sieve, preserved in 4%
 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
formalin solution in water and sorted out in the
laboratory. Benthic species were identified to species
level where possible. Chemical parameters correlated,
include organic carbon content in sediment, dissolved
oxygen in the water column, oxygen in the bottom
layer, nitrates and phosphates. Chemical data were
integrated as average annual values (NCMR, 1997;
FATE PROJECT, 2005).
Benthic indices applied include the routine indices
of species richness (S), calculated per surface unit
(0.1 m2 for Evvoikos and 0.045 m2 for Thessaloniki
gulf), the Shannon–Wiener diversity index H0 (Shan-
non and Weaver, 1963), the Pielou’s evenness index J
(Pielou, 1969) and the density of individuals per
square meter (N/m2). The problem of incomparability
of the sampling surface used in the two areas was
overcome by standardizing data. The diversity and
species richness indices were calculated on the 0.1 m2
sampling surface, averaging the two replicates in the
case of S. Evvoikos gulf and pooling the three
replicates of 0.045 m2 sampling surface in the case of
Thessaloniki gulf. The biotic index Bentix was used to
estimate the ecological status of the communities. The
AMBI index (Borja et al., 2000) was also applied
using the AMBI index software (ver.3.0: http://
www.azti.es/).
The Bentix index (Simboura and Zenetos, 2002) is
a newly developed biotic index based on the relative
percentage of ‘sensitive’ (GS) and ‘tolerant’ (GT)
species in the fauna weighted analogously to derive a
single formula: Bentix = (6  %GS + 2  %GT)/100
(http://www.hcmr.gr/english_site/services/env_aspects/
bentix.html). The formula was initially developed
as Bentix = [(6  %GI + 2  (%GII + %GIII)]/100,
where GI include the sensitive and indifferent taxa,
GII the tolerant and second order opportunistic
and GIII the first order opportunistic taxa. As the
multiplying factor 2 is the same for groups GII and
GIII, these groups can be merged to simply represent
all tolerant taxa (GT) versus all sensitive taxa (GS).
The resulting classification scheme and ecological
quality ratio (EQR), which is the universal comparison
scale for different metrics is given in Table 1.
The method has been tested successfully in a
variety of Eastern Mediterranean (Greece, Cyprus and
Turkey) benthic ecosystems and in case areas subject
to organic pollution, such as the Athens or Saronikos
gulf (Simboura et al., 2005), the Izmir Bay (Doğan
457
Table 1
The Bentix index classification scheme and EQR
EcoQS Bentix values EQR
High 4.5  Bentix < 6 1
Good 3.5  Bentix < 4.5 0.75
Moderate 2.5  Bentix < 3.5 0.58
Poor 2.0  Bentix < 2.5 0.42
Bad 0 0
For naturally stressed muddy habitats only, boundary limits 4.5 and
3.5 are reduced to 4 and 3, respectively.
et al., 2004), to oil spills accidents (Zenetos et al.,
2004), to industrial solid wastes discharge area
(Simboura et al., in press) and to fish culture areas
(Simboura and Argyrou, 2006). It has been also tested
successfully in 70% of cases in the Sea of Marmara, a
basin connecting the Black Sea and the Mediterranean
Sea and receiving a large number of wastewater
discharges from land-based sources (Albayrak et al.,
in press). The index renders a five step numerical
scheme for the classification of benthic communities
into five ecological quality classes according to the
requirements of The European Water Framework
Directive (WFD, 2000/60/EC).
3. Results
3.1. Description of the benthic environment
The two gulfs under study have different char-
acteristics concerning the physical and biological
conditions as presented in Table 2. Regarding the
typological characterization of water bodies according
to the WFD guidelines (EC, 2003a), S. Evvoikos gulf
is characterized as a rocky deep exposed gulf with
depth ranging from 32 to 70 m while Thessaloniki gulf
(6–19 m) as a sedimentary shallow sheltered gulf with
the inner parts bearing characteristics of very sheltered
bays. Also, the pressures imposed on the two areas
differ in scale and type. S. Evvoikos coastal area is
slightly affected by minor pressures including touristic
activities and small scale fisheries. The benthic
communities’ composition and structure in the
periphery of Attiki (S. Evvoikos gulf, Saronikos gulf)
is described in detail in Reizopoulou and Zenetos
(2005).
458 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
Table 2
Variation of physical parameters in the areas studied
Min–max Depth Main Bottom
avg  S.D. (m) substrate type oxygen (ml/l)
S. Evvoikos 32–70 Muddy sands 5.24–5.43
gulf
52 5.32  0.06
Thessaloniki 6–19 Sandy muds 5.34–5.77
gulf
12 5.62  0.16
In Thessaloniki gulf, main pressures include the
urban and industrial wastes of Thessaloniki urban area
that are collected and discharged after partial
treatment in two sites of Thessaloniki gulf. The gulf
also receives agricultural wastes from a wide drainage
area transported by the rivers. Organic pollution from
urban and agricultural wastes discharged in the
enclosed Thessaloniki bay results in eutrophication
and consequently in the degradation of the benthic
element. The impact on the ecological quality is more
prominent in the enclosed bay of Thessaloniki
decreasing towards the opening of the gulf.
The benthic element of the two gulfs is bearing also
some differences pertaining to the different typology
of the two areas. The integrated communities, habitats
and water bodies classification scheme used, harmo-
nized according to the defined typology, is described
in Simboura et al. (2005) and is based on the EUNIS
system and the classical benthic bionomy of the
Mediterranean (Pérès and Picard, 1964).
The habitats and communities dominating in S.
Evvoikos gulf are those mostly related to rocky deep
exposed areas. Generally, most among the dominant
species characterize the muddy detritus bottom
communities (DE) or the muddy sands as the
polychaete species Euchone rosea, Aponuphis bre-
menti, Chaetozone sp., Nematonereis unicornis,
Lumbrineris latreilli, Monticellina dorsobranchialis,
Paralacydonia paradoxa, Pseudoleiocapitella fauveli
and Sphaerosyllis taylori. Others characterize the
coastal terrigenous muds community (VTC) or the
sandy muds as the polychaetes Laonice cirrata,
Levinsenia gracilis, Nephtys hystricis and the bivalve
Corbula gibba, while there exist some species
characteristic of coarse sands and fine gravel under
the influence of bottom currents community (SGCF)
as the polychaetes Kefersteinia cirrata and Praegeria
Dissolved Phosphates Nitrates Organic
oxygen (ml/l) PO4 (mM) NO3 (mM) carbon %
5.27–5.45 0.06–0.15 0.17–1.08 –
5.33  0.06 0.1  0.03 0.54  0.3
5.55–5.96 0.33–0.77 0.26–3.15 0.68–0.8
5.75  0.14 0.45  0.17 0.77  1.16 1.23  0.48
remota. Species characteristic of the deeper sands with
detritus (DC) community and the well sorted fine
sands community (SFBC) are also among dominants
as is the amphipod Ampelisca sarsi and the polychaete
Aricidea cerruti, respectively.
In Thessaloniki gulf, some of the above commu-
nities related also to sedimentary shallow sheltered
water bodies are also well represented by the bivalve
Hyala vitrea (VTC), and the polychaetes L. latreilli,
Melinna palmata, M. dorsobranchialis, P. fauveli
(DE). However, there exist among dominants, some
species typical of habitat or community types related
to shallower and more sheltered environments.
Among them, the polychaetes Aonides oxycephala
and Paradoneis lyra characterize the superficial
muddy sands in sheltered waters community (SVMC),
while species of the eurythermal euryaline biocoe-
nosis are also well represented as is the polychaete
Heteromastus filiformis and the amphipod Corophium
acherusicum. Also in Thessaloniki gulf and especially
at stations DA3 and TP6 are found in great abundances
amphipod and decapod species thriving on the
microhabitats formed by the rich shell debris
accumulated at these areas close to mussel culture
farms.
3.2. Comparative analysis of the descriptors
tested
The values of the benthic indices applied in the
fauna are presented in Table 3, giving also the
ecological quality status (EcoQS) assessment based on
the Bentix and the AMBI index. Table 1 of the annex
presents the five most abundant species at each station
giving also their numerical abundance (ind/m2). The
species are classified into one of the three ecological
groups of species according to the analytical Bentix
 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
Table 3
Stations characteristics and indices applied
Shadow indicate moderate ecological quality status (EQS). Diversity and species richness values refer to the sampling surface unit used at each area.
methodology (Simboura and Zenetos, 2002). GI
represents the k-strategic sensitive taxa, GII the
second order opportunists and the simply tolerant to
disturbance species and GIII the first order opportu-
nists or pioneer species adopting the r-strategy.
Stations are listed in the table in order of descending
ecological status as assessed by the Bentix index.
The application of the Bentix index in S. Evvoikos
gulf stations classified station E28 at Marathonas to
the moderate class, stations E11, E13 and E31 to the
high class (reference sites) and the rest of the stations
to the good class. Station E28 with lowest ecological
status presented the highest percentage of polychaetes
(81.2%) and the highest contribution of tolerant taxa
among Evvoikos gulf stations (M. dorsobranchialis
7.5%, Lumbrineris gracilis 7.8%, N. unicornis 6%),
although the diversity (H0 = 5.84) and species richness
(S = 87) were relatively high. Table 1 (annex) also
shows a high density of GII species as L. gracilis, M.
dorsobranchialis, L. gracilis and N. unicornis occur-
ring at this station. The EcoQS at station E31 was
459
assessed according to the modification of the border of
the metric between good and high applied to purely
muddy habitats (Simboura and Zenetos, 2002).
Actually muddy bottom habitats usually found in
very sheltered bays are naturally stressed environ-
ments as the high percentage of fines and the low
hydrological regime favor the accumulation of organic
matter and nutrients, thus normally hosting several
tolerant species. The AMBI index applied in the data
classified all stations as in good EcoQS.
At Thessaloniki gulf, station DA3 located close to the
mussel farms and TP6 in the eastern part of the gulf were
classified as in good status, while the rest of the stations,
namely TP2 close to the port and TP7, TP10, TP13 were
classified by the Bentix index as in moderate EcoQS. As
Table 1 of the annex shows, moderate class stations in
Thessaloniki gulf are dominated by GII taxa as the
polychaetes M. palmata, M. dorsobranchialis, P. fauveli
and the amphipod C. acherusicum and GIII species as
the polychaetes H. filiformis and P. lyra. AMBI again
classified all stations as being at good status.
460 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468

Generally, as demonstrated by the application of

the Bentix index in S. Evvoikos gulf, the area is
characterized mostly as in good EcoQS reaching to
high status, while in Thessaloniki gulf the EcoQS is
mostly moderate reaching to good. It is noteworthy
that stations with high and good EcoQS are populated
mostly by GI group of species, while stations
designated as in moderate EcoQS are mostly
characterized by GII and GIII species (annex).
Fig. 2 shows the range of the Bentix index over the
stations, covering the classes from moderate to high.
Fig. 2. Range of the Bentix index over the stations studied. Fig. 3 shows the variation of the diversity (Fig. 3a)
and species richness index (Fig. 3b) against the axis of
increasing values of the Bentix index. The correlation
of the Bentix with diversity is statistically significant

Fig. 3. Covariance of Bentix with diversity (a) and species richness (b) indices.
 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
Table 4
Simple regression of the Bentix index with chemical parameters in Thermaikos gulf
Chemical parameter P-value Bentix
Bottom oxygen (ml/l) 0.1335
Dissolved oxygen (ml/l) 0.058
Phosphates PO4 (mM) 0.000
Nitrates NO3 (mM) 0.0035
(P = 0.06 < 0.1), while species richness is not
correlated with the Bentix (P = 0.4). This is inter-
preted by the fact that Shannon diversity encompasses
two components of the community structure: evenness
and species richness. The evenness or equitability
component of diversity is sensing the occurrence of
opportunist species but generally diversity is much
more influenced by species richness. For example,
diversity and species richness may overestimate the
communities’ health state in cases of transitional
zones where a high number of opportunistic and
sensitive species may co-exist according to the model
of communities’ succession (Pearson and Rosenberg,
1978).
The species richness component of diversity is also
greatly influenced by many natural or methodological
factors as commented later in the discussion.
It is noteworthy, that station E30 with maximal
value of the diversity index (H0 = 6.21) is character-
ized as in good but not high EcoQS. As seen in Table 1
of the annex, among the five most abundant species
dominating station E30 are listed only GII and GIII
group species indicating some disturbance that results
into this deviation from optimal EcoQS. Similarly,
stations TP13 and E28 classified as in moderate
EcoQS and populated by opportunistic species
(annex) present high diversity values (H0 = 5.8) while
species richness is maximal (S = 138) in TP13 station.
3.3. Analysis of the sources of variation of the two
biotic indices
To validate the EcoQS results, the Bentix and
AMBI indices were correlated with the available
chemical parameters. When all data from S. Evvoikos
and Thessaloniki gulfs were correlated with the
abiotic parameters, no correlation was proved sig-
nificant for both Bentix and AMBI indices. This is
expected since the different typology of the gulfs
largely shapes the special physico-chemical conditions
461
Statistical significance P-value AMBI Statistical significance
0.046 <0.1
<0.1 0.084 <0.1
<0.01 0.4
<0.01 0.44
characterizing each area. S. Evvoikos gulf, much
deeper than Thessaloniki gulf, and seasonally stratified,
presents lower mean annual oxygen values compared to
Thessaloniki gulf which is more eutrophicated but
shallower and with fresh water inputs. It is noteworthy,
that during the winter period (period of sampling in
Thessaloniki gulf), conditions of vertical mixing of the
temperature salinity and density fields in the gulf and
bay of Thessaloniki have been recorded (NCMR,
1996). Thus, correlations were tested separately for
each gulf. Table 4 presents the correlations among the
biotic indices and chemical parameters in Thessaloniki
gulf.
Organic carbon content was not correlated with the
biotic indices, due to the limited number of measure-
ments available, but it is indicative that at station DA3
with highest EcoQS according to both indices, the
sediment organic carbon content was lowest (0.68%).
Dissolved oxygen showed a statistically significant
correlation with the Bentix while also bottom oxygen
showed a weak positive correlation. However,
phosphates and nitrates showed a significant positive
correlation with the Bentix, which is not expected as
with increasing EcoQS, the nutrients of the environ-
ment should normally decrease. This situation is
linked to the particular conditions characterizing the
marine environment at stations DA3 (mussel culture
area) and TP6 of Thessaloniki gulf assessed as in good
EcoQS.
This situation is considered as a case where
moderate chemical status is paired with good
ecological status as reflected in the macroinverte-
brates. The good quality status of these patches of the
gulf (stations DA3 and TP6) is shaped by the high
hydrodynamic regime of these sites reflected also in
the maximal values of dissolved and bottom oxygen
(5.75–5.96 ml/l) among stations plotted. This well-
oxygenated condition is also reflected in the composi-
tion of the substrate consisting of stones, shells and
shell debris. Shells drifted or accumulated from
462 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
mussel farming activity in the area, create a special
micro-habitat favoring the establishment of a rich and
diverse benthic fauna dominated by hard and soft
bottom species, among them many GI group
crustaceans developing on the shells surface and
reaching densities as high as 1353 ind/m2 (Astacilla
sp.) or 1407 ind/m2 (Decapoda sp.)—see Table 1 of
annex. In that nutrient rich environment, where
nutrient values are maximal among stations studied
(PO4: 0.5–0.7 mM and NO3: 0.8–3.15 mM), the water
is well oxygenated (5.75–5.96 ml/l) and the macro-
invertebrate element is in good ecological condition.
The correlations of AMBI with chemical para-
meters in Thessaloniki gulf showed a significant
correlation with dissolved and bottom oxygen and no
correlation with phosphates and nitrates. The good
correlation of AMBI with bottom oxygen levels lies on
the fact that station TP7 with lowest EcoQS according
to the AMBI showed the lowest levels of dissolved and
bottom oxygen. On the other hand, the lowest EcoQS
station according to the Bentix is station TP10 with
second lowest oxygen levels. This difference in the
rank of EcoQS assigned by the two indices is largely
attributed to differences in the scoring of the
polychaete species L. latreilli found in significant
density (14%) at station TP10. The species is
considered as an indicator of instability and zones
of transitional pollution (Bellan, 1985; Pearson and
Rosenberg, 1978) and is scored as a tolerant, while
AMBI methodology scores it as an indifferent to
pollution species.
AMBI showed no statistically significant relation
with the nutrients, although station DA3 with highest
values of nutrients (PO4: 0.7 mM and NO3: 3.15 mM)
was assigned to the highest EcoQS according to AMBI
as with the Bentix.
In S. Evvoikos gulf the correlations among the
biotic and abiotic factors tested, were not significant
and this is attributed to the lack of significant pressure
gradients as shows the low level of variation of the
chemical parameters compared to Thessaloniki gulf
(Table 2). The lack of strong gradients is also indicated
by the biotic indices assessments showing an
accumulation of sites in the good class.
Attempting a direct comparison among the two
biotic indices tested, Fig. 4 shows the relation
among the two indices’ ecological quality ratio
(EQR) which is linear and statistically significant
Fig. 4. Correlation of AMBI and Bentix EQRs. H, high class; G,
good class; M, moderate class; P, poor class; B bad class.
(P = 0.0007). The lines in the plot define the ranges
of each class set by the boundary limits of each
method.
The two indices are linearly correlated but the
EcoQS assessment rendered by each method often
differs. This difference is due to: (a) the differential
weight each index puts in the various ecological
groups and (b) in the different boundary limits and
EQRs set among classes for the two indices as Fig. 4
shows.
In the good class range, the two indices give a
common assessment. However, the AMBI classifica-
tion method gives a wider ‘‘good’’ class range
compared with the Bentix, classifying most moderate
and often high class sites according to the Bentix to the
‘‘good’’ class. Generally, moderate and high class
ranges in the AMBI are more compressed compared to
the Bentix scheme.
Over the whole range of stations studied, a clear
gradient of EcoQS was demonstrated by the Bentix,
ranging from high to moderate status (Fig. 2). The
AMBI shows a good correlation with the oxygen
levels but renders a uniform assessment classifying all
stations to the good class.
The relative percentage contribution of the two
ecological groups of species along a gradient of
increasing Bentix values represents a biological
interpretation of the metric’s functioning in relation
to the normative definitions of WFD as already
described elsewhere (Simboura et al., 2005).
 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
Fig. 5. Bentix values plotted against the percentages of the two ecological groups. M, moderate class; G, good class; H, high class.
Fig. 5 shows the plotting of the percentage of
tolerant and sensitive taxa along the gradient of
increasing Bentix index. Based on this plot, it is
evident that the cross-line point of the two lines corres-
ponding to the value of the Bentix = 4, represents
the class center of ‘‘good’’ where literally the two
ecological groups of tolerant and sensitive share the
fauna by 50% each.
At the high to good class boundary (Bentix = 4.5),
the percentage of the sensitive taxa drops to less than
60% of the fauna and the percentage of the tolerant
taxa accounts for more than 40%. At the good to
moderate class boundary (Bentix = 3.5), the percen-
tage of tolerant species becomes over 60% (roughly
2/3 of the fauna) and the sensitive taxa less than 40%
(1/3 of the fauna).
4. Discussion
A significant effort of the scientific community has
been devoted lately on the development, testing and
evaluation of various univariate and multivariate
indices for use in the ecological status assessment
according to the requirements of WFD. Among such
newly developed indices are included the AZTI
marine Biotic Index (AMBI) (Borja et al., 2000), the
Factorial Analysis combining the AMBI, richness and
diversity (Borja et al., 2003), the Benthic Quality
Index (BQI) of Rosenberg et al., 2004 and the Bentix
index (Simboura and Zenetos, 2002). Different
metrics may render different results, as often indices
463
are designed to fit best to the specific environments for
which they were developed. However, it is essential to
justify the ecological relevance of a given metric by
using pressure data, demonstrate the gradients and
comparing with other metrics (EC, 2003b). The
compliance of the ECoQS assessment is validated by
the good correlation of the Bentix with oxygen levels
and the pressure information. The ecological status of
the benthic communities is mostly related to the
chemical conditions of the bottom sediments and
may not be affected by the chemical quality of the
water column as showed the inconsistency among
ecological and chemical status in Thessaloniki gulf.
Besides, the benthic communities may be affected
by sources of disturbance other than chemical
pollution.
Both indices are well correlated to each other and
the oxygen levels, but the final EcoQS differs. The
AMBI, although has a good correlation with dissolved
and bottom oxygen gives a uniform assessment
designating all stations to the good class. Bentix,
correlated more to dissolved oxygen levels and less so
to bottom oxygen, discriminated among two classes:
good and high in S. Evvoikos gulf and good and
moderate in Thessaloniki gulf.
Thessaloniki gulf is an eutrophicated area sub-
jected to organic pollution from urban wastes and is
expected to show stretches of moderate EcoQS. As
mentioned in the results, the difference in the two
methods lies to the different weighting of each
ecological group in the formula and the different
scaling of boundary limits among classes.
464 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
In the AMBI method, firstly tested and designed for
the Atlantic benthic fauna where naturally there exist
some species with very high densities covering
percentages as high as 40–60%, each one of the five
ecological groups (sensitive, indifferent, tolerant, first
order opportunists and second order opportunists),
takes a different weight in the formula. Thus, a given
species with high dominance is assigned to one group
among the five so that the overall effect of that species
is leveled and the condition is also evaluated from the
other groups’ contribution.
The Bentix index was developed in the Medi-
terranean ecoregion where the benthic fauna is
usually very diverse and evenly distributed with no
one species naturally dominating over 10%. Also
under slight disturbance conditions there might be
the situation where a high diversity of sensitive and
opportunistic species may co-exist giving the
impression of a high quality status. Given the
naturally even distribution of the fauna, the groups
are weighted equally in the Bentix formula taking
into account simply only the ratio of the groups’
occurrence in the fauna: the probability of one
species randomly picked up from the fauna to belong
to a ‘‘tolerant’’ over a ‘‘sensitive’’ group is 3:1. Also
the indifferent group is counted with the sensitive.
This approach combined with an equal scaling of the
ranges of good and moderate classes (2.5–3.5, 3.5–
4.5, respectively, see also Table 1) results to
successfully designate the community health in the
Mediterranean benthic ecosystem where, naturally a
high number of species are evenly distributed over
the benthic population.
AMBI may demonstrate good class in the
Mediterranean sites where Bentix has assessed as
high class, as in the case of S. Evvoikos gulf. This is
partly attributed to the fact that in the AMBI method
the indifferent group species are weighted separately,
while Bentix counts them with the sensitive. Also the
high class in the AMBI (0–1.2) is more compressed
compared to the Bentix high class (4.5–6).
Another difference is that AMBI demonstrates
good class in some Mediterranean sites where
Bentix has assigned a moderate class, as in the case
of Thessaloniki gulf. One reason for this is that
tolerant species are weighted separately in the
AMBI giving different significance to each ‘‘toler-
ant’’ group of species. In the Bentix method, all
tolerant species are weighted equally versus the
sensitive ones. Besides, the scaling of the distances
among classes is different in the two methods. The
Bentix sets equal distances for the moderate (2.5–
3.5) and good (3.5–4.5) classes, while the AMBI
renders a wider good class (1.2–3.3) compared to
the moderate (3.3–4.3) and high (0–1.2).
The WFD guidance document (EC, 2003a) states
that ‘‘methods combining composition, abundance
and sensitivity may be the most promising’’.
However, univariate indices, such as the Shannon
diversity index, species richness and multimetric
indices encompassing those indices, have been
criticized as being dependent on too many factors.
Seasonal natural variability, habitat type, sample size
and sampling methodology influence diversity and
species richness, which are therefore generally
recommended to be used with caution as ecological
classification device (Reiss and Kröncke, 2005; Salas
et al., 2006). Actually, and as evidenced by the results
of this and other studies in this Mediterranean area
(Simboura and Zenetos, 2002), diversity index (H0 )
has a tendency to overestimate the condition of the
communities as it may suppress and not highlight the
dominance of opportunistic species in cases where
the species richness is relatively high as in transi-
tional zones corresponding with the ecotone point of
the communities’ evolutional model in response to
environmental pressure (Pearson and Rosenberg,
1978). In other not strictly marine habitats, such as
estuaries, lagoons or low salinity coastal areas, where
the dominance of certain species is not related to any
type of disturbance, being favored by abundant food
resources, the Shannon diversity may have a
tendency to underestimate EcoQS (Labrune et al.,
2006; Reizopoulou and Nicolaidou, 2004; Salas
et al., 2006).
Biotic indices, such as AMBI and BENTIX, based
on the ecological grouping of species, are generally
considered as a promising approach for ecological
quality assessment in order to avoid drawbacks due to
the seasonal variability of the benthic communities
and dependence from other factors (Reiss and
Kröncke, 2005; Salas et al., 2006).
The WFD requires that a given classification
scheme should be interpreted according to the
normative definitions of the directive. According to
the model described in Fig. 4, the good class center of
 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
the Bentix method corresponds to both groups sharing
‘‘equally’’ 50% of the fauna. The good class according
to the WFD corresponds with only slight deviation
from undisturbed conditions and most sensitive taxa
must be present. Our results indicate that at good class
the sensitive species account to no less than 40%. This
may not be the absolute majority of the fauna but we
consider it as a satisfactory percentage for good class,
given that tolerant species cover all grades of
resistance to disturbance (Simboura and Zenetos,
2002). The moderate class according to this model
corresponds with sensitive species dropping below
40% and tolerant species increasing to more than 60%.
This actually complies with the definition of moderate
class as corresponding with moderate deviation from
normal where most of the sensitive taxa should be
absent (less than 40% present).
Conclusively, testing of the Bentix index in these
coastal areas, typical of the Eastern Mediterranean
465
oligotrophic environment, discriminated among high,
good and moderate ecological quality status and the
EcoQS assessment was validated by abiotic para-
meters and pressure information.
Acknowledgements
The means and framework for acquiring the data
used in this work were provided through HCMR
regional (NCMR, 1997) and EU project: FATE
PROJECT, 2005: contract EVK3-CT01-00055. The
helpful comments of the referees are also gratefully
acknowledged.
Appendix A
Table A.1.
 466
Table A.1
Five most abundant species (ind/m2) in the studied stations and Bentix index values in decreasing order

N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468

N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468 467
468 N. Simboura, S. Reizopoulou / Ecological Indicators 7 (2007) 455–468
References
Albayrak, S., Balkis, H., Zenetos, A., Kurun, A., Kubanç, C.
Ecological quality status of coastal benthic ecosystems in the
Sea of Marmara. Mar. Pollut. Bull., in press.
Balopoulos, E.T., Friligos, N.C., 1993. Water circulation and eutro-
phication in the North-West Aegean Sea: Thermaikos Gulf. J.
Environ. Sci. Health A 28, 1311–1329.
Bellan, G., 1985. Effects of pollution and man-made modifications
on marine benthic communities in the Mediterranean: a review.
In: Moraitou-Apostolopoulou, M., Kiortsis, V. (Eds.), Medi-
terranean Marine Ecosystems. NATO.
Borja, A., Franco, J., Perez, V., 2000. A marine biotic index to
establish the ecological quality of soft-bottom benthos within
European estuarine and coastal environments. Mar. Pollut. Bull.
40 (12), 1100–1114.
Borja, A., Franco, J., Muxica, I., 2003. Classification tools for
marine ecological quality assessment: the usefulness of macro-
benthic communities in an area affected by a submarine outfall.
ICES CM 2003/J:02, pp. 1–10.
Doğan, A., Önen, M., Ergen, Z., Katagan, T., Çinar, M.E., 2004.
Ecological quality assessment in izmir Bay: Using the Bentix
Index. The Regional Workshop on Marine sciences & Natural
Resources, 25–26 May 2004, Lattakia, Syria.
EC, 2000. Directive of the European parliament and of the Council
2000/60/EC establishing a framework for community action in
the field of Water Policy. PE-CONS 3639/1/00.
EC, 2003a. Guidance on typology, reference conditions and classifica-
tion systems for transitional and coastal waters. Produced by: CIS
Working Group 2.4. (Coast), Common Implementation Strategy of
the Water Framework Directive, European Commission, p. 116.
EC, 2003b. Towards a guidance on establishment of the intercali-
bration network and the process on the intercalibration exercise.
Produced by: CIS Working Group 2.5. (Intercalibration), Com-
mon Implementation Strategy of the Water Framework Direc-
tive, European Commission, p. 54.
FATE PROJECT, 2005. Transfer and Fate of Harmful Algal Bloom
(HAB) toxins in European Marine Waters (1 January 2002–31
December 2004). In: E. Graneli (coord.), Final Scientific and
Management Report. Contract EVK3-CT01-00055, part of the
EC-EUROHAB cluster.
Gotsis-Skretas, O., Friligos, N., 1990. Contribution to eutrophica-
tion and phytoplankton ecology in the Thermaikos Gulf. Tha-
lassographica 13, 1–12.
Karageorgis, A.P., Skourtos, M.S., Kapsimalis, V., Kontogianni,
A.D., Skoulikidis, N.Th., Pagou, K., Nikolaidis, N.P., Drako-
poulou, P., Zanou, B., Karamanos, H., Levkov, Z., Anagnostou,
Ch., 2005. An Integrated Approach To Watershed Management
within the DPSIR Framework: Axios River Catchment and
Thermaikos Gulf. Reg. Environ. Change 5, 138–160.
Labrune, C., Amouroux, J.M., Sarda, R., Dutrieux, E., Thorin, S.,
Rosenberg, R., Grémare, A., 2006. Characterization of the
ecological quality of the coastal Gulf of Lions (NW Mediterra-
nean). A comparative approach based on three biotic indices.
Mar. Pollut. Bull. 52 (1), 34–47.
Moncheva, S., Gotsis-Skretas, O., Pagou, K., Krastev, A., 2001.
Phytoplankton blooms in Black Sea and Mediterranean ecosys-
tems subjected to anthropogenic eutrophication: similarities and
differences. Estuar. Coast. Shelf Sci. 53, 281–295.
NCMR, 1997. Study of the marine ecosystem of E-SE Attica from
the Bay of Marathon to the Bay of Anavissos. In: E. Christou
(Ed.), 377 pp. (in Greek).
NCMR, 1996. Study of the marine ecosystem of Thermaikos gulf.
In: Ch. Anagnostou (Ed.), 431 pp. (in Greek).
Pearson, T.H., Rosenberg, R., 1978. Macrobenthic succession in
relation to organic enrichment and pollution of the marine
environment. Oceanogr. Mar. Biol. Annu. Rev. 16, 229–311.
Pérès, J.M., Picard, J., 1964. Nouveau manuel de bionomie benthi-
que de la Mer Méditerranée. Rec. Trav. Stn. Mar. Endoume 31
(47), 137.
Pielou, E.C., 1969.In: An Introduction to Mathematical Ecology.
Wiley-Interscience, London.
Reiss, H., Kröncke, I., 2005. Seasonal variability of benthic indices:
an approach to test the applicability of different indices for
ecosystem quality assessment. Mar. Pollut. Bull. 65, 253–274.
Reizopoulou, S., Nicolaidou, A., 2004. Benthic diversity of coastal
brackish-water lagoons in western Greece. Aquat. Conserv: Mar.
Freshwater Ecosyst. 14, S93–S102.
Reizopoulou, S., Zenetos, A., 2005. Summer resorts and ecological
quality. A preliminary study of an eastern Mediterranean coastal
ecosystem: summer resorts and benthic ecosystems. Mediterr.
Mar. Sci. 41 (6/1), 41–50.
Rosenberg, R., Blomqvist, M., Nilsson, H.C., Dimming, A., 2004.
Marine quality assessment by use of a benthic species-abun-
dance distributions: a proposed new protocol within the Eur-
opean Union Water Framework Directive. Mar. Pollut. Bull. 49
(9–10), 728–739.
Salas, F., Patrı́cio, J., Marcos, C., Pardal, M.A., Pérez-Ruzafa, A.,
Marques, J.C., 2006. Are taxonomic distinctness measures
compliant to other ecological indicators in assessing ecological
status? Mar. Pollut. Bull. 52, 162–174.
Shannon, C.E., Weaver, W., 1963.In: The Mathematical Theory of
Communication. University of Illinois Press, Urbana, USA, p. 117.
Simboura, N., Zenetos, A., 2002. Benthic indicators to use in
ecological quality classification of Mediterranean soft bottom
marine ecosystems, including a new biotic index. Mediterr. Mar.
Sci. 3/2, 77–111.
Simboura, N., Panayotidis, P., Papathanassiou, E., 2005. A synthesis
of the biological quality elements for the implementation of the
European Water Framework Directive in the Mediterranean
Ecoregion: the case of Saronikos Gulf. Ecol. Indicators 5,
253–266.
Simboura, N., Papathanassiou, E., Sakellariou, D. The use of a biotic
index (Bentix) in assessing long term effects of dumping coarse
metalliferous waste on soft bottom benthic communities. Ecol.
Indicators, in press.
Simboura, N., Argyrou, M., 2006. Implementation of Water Frame-
work Directive in Cyprus: testing of the Bentix Index in Limassol
Bay. Eighth Hell. Symp. Oceanogr. Fish. Thessaloniki, 4–8 June
2006, Book of Abstracts, p. 77 (in Greek with English abstract).
Zenetos, A., Chadjianestis, I., Lantzouni, M., Simboura, M., Skli-
vagou, E., Arvanitakis, G., 2004. The Eurobulker oil spill: mid-
term changes of some ecosystem indicators. Mar. Pollut. Bull. 48
(1–2), 12–131.