Marine water quality monitoring: A review

Marine Pollution Bulletin xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier .com/locate /marpolbul

Review

Marine water quality monitoring: A review

0025-326X/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.marpolbul.2013.09.012

⇑ Corresponding author. Tel./fax: +30 22510 36819.E-mail address: [email protected] (D. Kitsiou).

Please cite this article in press as: Karydis, M., Kitsiou, D. Marine water quality monitoring: A review. Mar. Pollut. Bull. (2013), http://dx.doi.org/1j.marpolbul.2013.09.012

Michael Karydis, Dimitra Kitsiou ⇑Department of Marine Sciences, School of the Environment, University of the Aegean, Mytilene 81100, Greece

a r t i c l e i n f o a b s t r a c t

Keywords:Marine pollutionSea protectionMarine conventionsMarine governance

Marine water quality monitoring is performed for compliance with regulatory issues, trend detection,model validation and assessment of the effectiveness of adopted policies. As the end users are managersand policy makers, the objectives should be of practical interest and the answers should reduce theuncertainty concerning environmental impact, supporting planning and decision making. Simple andclearcut answers on environmental issues require synthesis of the field information using statistics,simulation models and multiple criteria analysis (MCA). Statistics is easy to apply whereas simulationmodels enable researchers to forecast future trends as well as test different scenarios. MCA allows theco-estimation of socio-economic variables providing a compromise between scientists’ and policy mak-ers’ priorities. In addition, stakeholders and the public have the right to know and participate. This articlereviews marine water quality monitoring principles, design and data analysis procedures. A brief reviewof international conventions of regional seas is also included.

� 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 002. Marine water quality monitoring: setting objectives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 003. Marine water quality monitoring: sampling design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 004. Marine water quality monitoring variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 005. Ecological knowledge on marine water quality monitoring studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 006. Data storage and management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 007. Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

7.1. Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 007.2. Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 007.3. Multiple criteria analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

8. The integrated approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009. International conventions and EU directives with monitoring components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

9.1. Environmental governance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.2. The Mediterranean Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.3. The Black Sea. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.4. The Baltic Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.5. The North Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.6. The Wider Caribbean Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.7. The East Asian Seas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.8. The West and Central African Region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.9. The regional seas in USA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 009.10. Environmental governance in the European Union marine environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

10. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 00

0.1016/

http://dx.doi.org/10.1016/j.marpolbul.2013.09.012

mailto:[email protected]


http://www.sciencedirect.com/science/journal/0025326X

http://www.elsevier.com/locate/marpolbul



2 M. Karydis, D. Kitsiou / Marine Pollution Bulletin xxx (2013) xxx–xxx

1. Introduction

Quality assessment of the marine environment is routinely per-formed by measuring causes (pollutants) and effects (ecosystemimpact) in the sea. If this practice is based on a sound experimentaldesign and lasts over a number of years it is called ‘‘monitoring’’.Data collected during a monitoring program serve a number ofobjectives. They can be used for either compliance monitoring toensure that both pollutants and ecosystem impacts from specificactivities do not exceed standard values set by the authorities orrelevant legislation or for model validation and verification to ac-cess whether possible effects on the marine environment are with-in acceptable limits set at the onset of the activity. They can also beused for trend monitoring to identify environmental changes on along term basis. A monitoring definition given by MIT (1970) suitsrather to trend monitoring defined as the ‘‘systematic observationof parameters related to a specific problem designed to provideinformation on the characteristics of the problem that changeswith time’’. A monitoring program may fulfill one or more of theabove mentioned objectives depending on monitoring targets;the set of data collected through the program can support concep-tual and numerical modeling (future prospects), time series analy-sis (trend monitoring), statistical comparisons (impact assessment)as well as synthesis and interpretation (NRC, 1990). The differencebetween a monitoring system and individual research activities isthat a monitoring system is an integrated system producing infor-mation for environmental management (NRC, 1990). In spite of thelong experience acquired by high cost monitoring programs, theyhave been criticized that they do not provide information soundenough to form the baseline for decisions on environmental man-agement. Many historical pollution monitoring programs havefailed because they have not taken into account the key conceptthat the changes to be identified through monitoring should be sig-nificant to the marine environment and therefore be able to sup-port management needs (Segar and Stamman, 1986).

In addition to ecosystem complexity, water quality monitoringprograms suffer from variability due to design, sampling, labora-tory (analytical) errors and data manipulation practices. The num-ber of sampling sites, the number of samples per site, the numberof replicate samples and the frequency of sampling, are factors thathave to be taken into account as monitoring programs are usuallycompromising between requirements of science and budget avail-ability by the authorities. Experimental designs based on statisti-cally valid procedures would optimize the outcome; however, inmost monitoring programs the design is based rather on ‘‘judgmentsampling’’ rather than on statistically sound procedures, the latterbeing rather an exception to the rule (Erickson and Strickland,1995; Kitsiou et al., 2001).

In spite of the shortcomings mentioned above, environmentalmarine water quality monitoring forms a platform for policy mak-ing and management. Without field measurements, the creation ofregulatory issues and the application of management practices toprotect human health and marine water quality would not havebeen successful. The long term collection of data from the marineenvironment at regular intervals, their use in addressing hypothe-ses and their interpretation provides the policy maker with theobjective means to make decisions (Wolfe et al., 1987). Monitoringhas been characterized as the ‘‘central element in the rational pol-icy making’’ (de Jonge et al., 2006). Monitoring became a powerfuland decisive tool in environmental policy since the governance ofregional seas and coastal areas was supported by a number ofinternational conventions, treaties and laws over the last few dec-ades (DiMento and Hickman, 2012). Monitoring and assessmentare the fundamental components required for effective marinemanagement (Katsanevakis et al., 2011). Monitoring marine

Please cite this article in press as: Karydis, M., Kitsiou, D. Marine water qualityj.marpolbul.2013.09.012

waters also provides information on the efficiency of implementa-tion of measures decided for mitigating marine pollution and dete-rioration (Douvere and Ehler, 2011).

Initially the practice in monitoring programs in the marineenvironment was mainly focused on measuring concentrations ofhazardous compounds such as heavy metals and petroleum hydro-carbons. Currently, the information collected by marine waterquality monitoring is far more extensive and complex, includingphysical, chemical and biological variables (UNEP, 1997). In addi-tion, information on social and economic aspects is often requiredalong with the environmental information. Furthermore, data pro-cessing procedures are getting more multifaceted, including use ofstatistical methods on designing and analyzing the information(Chapman, 1996; Zuur et al., 2007; Kitsiou and Karydis, 2011),use of ecological indices for assessing ecosystem health (Magurran,2004; Karydis, 2009), methods of Multicriteria Analysis (Kitsiouet al., 2002), spatial analysis methods (Janssen, 1992; Kitsiou andKarydis, 2011), simulation models (NRC, 2003) as well as inte-grated approaches, have introduced a high level of complexity inhandling and assessing information acquired from monitoringdata.

The objective of the present work is to review the practices fol-lowed on the structure and design of water quality monitoring pro-grams as well as their potentiality and shortcomings in assessingmarine water quality. A brief review on the monitoring compo-nents of international conventions for regional seas and EuropeanUnion Directives is also presented.

2. Marine water quality monitoring: setting objectives

Clarity in the objectives is a crucial step when marine waterquality monitoring programs are designed (Table 1). The user ofthe monitoring outcome is a decision maker, who needs the infor-mation to protect human health, to make sure that there is nounacceptable impact either on ecosystems or on marine resourcesand finally make decisions concerning disposal of pollutants in themarine environment. Monitoring is successful when the results canbe used directly for effective management decisions. This assumesa two way communication between scientists and policy makers:the policy makers should realize beforehand the limitations ofmonitoring concerning the necessary information for decisionmaking, whereas the scientists should know what kind of ques-tions are of practical interest to policy makers. It is well known thatboth policy makers and the public need simple and practical an-swers to environmental problems. The most common questionsare: (a) is the water quality in an area improving or not? (b) Is fishand shellfish biomass increasing? (c) Are fish and shellfish safe toeat? (d) Are coastal waters suitable for swimming?

The ‘‘holistic approach’’ should be the final goal of every envi-ronmental manager. Monitoring processes for the policy makeraim also to enable environmental managers to set standards, touse predictive models effectively by verifying predictions or if nec-essary, readjusting the model, to ensure that there is compliancewith legislative requirements, otherwise to take necessary mea-sures and to set an early warning system in view of future prob-lems In addition, they aim to improve knowledge on structureand function of ecosystems (this target is especially feasible ifthe monitoring program is linked with relevant research projects)and to establish a better understanding about the health of themarine environment. However, not all monitoring programs aresuccessful as both managers and politicians tend to ignore threebasic principles (Segar and Stamman, 1986; de Jong, 2006;Katsanevakis et al., 2011) that there is no human activity withoutenvironmental impact, monitoring programs cannot always detectpossible environmental impact and finally a slight impact may be

monitoring: A review. Mar. Pollut. Bull. (2013), http://dx.doi.org/10.1016/



Table 1Steps for designing a water quality monitoring program; references are provided.

Step Information for designing monitoring References

Setting objectives Criteria for setting monitoring objectives Segar and Stamman (1986)Objectives based on uses of monitoring results NRC (1990)

Sampling design Sampling design Underwood (2007)Optimization of stations Kitsiou et al. (2001)Survey design McDonald et al. (1995)

Data analysis General review on data analysis Kitsiou and Karydis (2011)Data Analysis: Water Quality Assessment Chapman (1996)Analysis of Ecological Data Zuur et al. (2007)Ecological Indices Magurran (2004) Karydis and Tsirtsis (1996)Statistics: Analysis of Messy Data Milliken and Johnson (1997)Statistics: Different ANOVA designs Underwood (2007)Models: Review on Eutrophication NRC (2000)Models: Marine environment and watershed Arhonditsis et al. (2000)Models: Regime Shifts (eutrophication) Zaldivar et al. (2009)Models: Oil pollution NRC (2003)Models: Oil effects on biota French (1998)Models: Oil effects on the coastal zone Reed et al. (1989)Integrated approach: Methods of Multicriteria Analysis (MCA) Janssen (1992)

Kitsiou et al. (2002)Moriki and Karydis (1994a, 1994b)

Integrated approach: modeling and MCA Arhonditsis et al. (2000)

M. Karydis, D. Kitsiou / Marine Pollution Bulletin xxx (2013) xxx–xxx 3

acceptable and in some cases desirable (i.e. slightly eutrophicconditions).

Monitoring objectives are better defined if physical (geo-graphic) boundaries are set. Suitable boundaries can provide reli-able answers on objectives set in the technical design. In somecases, especially in site specific marine pollution monitoring, thereare questions that cannot be answered successfully as they are de-pended on a larger scale. The water quality monitoring program inthe Chesapeake Bay is a typical example (Holland et al., 1986)where the watershed area should have been taken into account.In addition to geographical boundaries, many more types ofboundaries have been identified beyond the spatial boundaries(Beanlands and Duinker, 1983): (a) administrative boundaries con-fined by political, social and economic factors (b) temporal bound-aries (c) ecological boundaries depending on physical, chemicaland biological processes and (d) technical boundaries: shortcom-ings in assessing ecological changes and impacts on ecosystemprocesses. Four dimensions have been mentioned by Walters(1986) that should be considered for setting boundaries: (a) thedifferent factors examined in the study (b) spatial scales (c) timescales and (d) the depth of analysis. Clear objectives and well setboundaries are a prerequisite for developing a sound samplingdesign.

3. Marine water quality monitoring: sampling design

A monitoring sampling design like any other marine environ-mental study should define (Segar and Stamman, 1986): (a) thenumber of stations (sampling sites) (b) the station locations (c)the number of samples collected per station (d) the number of rep-licates per sample and (e) the frequency of sampling (weekly,monthly or seasonal). Properly designed monitoring programs,using statistics, provide the maximum amount of informationabout the conditions investigated for the resources used (Millikenand Johnson, 1997). The maximum amount of information resultsfrom the resolution provided by the statistical model. The resolu-tion in detecting differences depends on the variance. There aretwo sources of variation (Segar and Stamman, 1986): (a) the phys-ical variance and (b) the sampling variance. The physical variancecan be affected by factors not included in the monitoring design:water movements, fluctuations in light intensity, unexpected per-turbations by human activities and algal blooms are among the


most common sources of physical variation. The sources of vari-ance due to experimental design constraints include: spatial vari-ability, temporal variability, sample handling variability andmeasurement variability (Segar and Stamman, 1986). The physical/biological meaning of each of those variances should be wellunderstood and taken into account when the program is designed.Some of the variances seem to be the same in various monitoringprograms whereas the extent of other variances seems to be site-specific, such as in eutrophic areas (Ignatiades et al., 1992; UNEP,2003). In that case, historical data would be particularly useful asa basic understanding of the system; otherwise a preliminary sur-vey should be carried out before a program is designed.

Resolution (temporal or spatial) is an essential element in mon-itoring projects as well as impact studies. Poorly designed monitor-ing surveys, show a resolution that is either (i) inadequate or (ii)confusing. In the first case data do not show any differences (dueto an impact) whereas in the second case small changes detectedmay be insignificant and irrelevant to environmental disturbance.To achieve a required resolution in a monitoring survey, a numberof criteria should be fulfilled: (a) it must be scientifically feasible,(b) it should be supported by the sampling design, (c) there shouldbe adequate resources and (d) it should be scientifically significant.

Before the implementation of the project, optimal use of theavailable resources and attainment of maximum resolution requirea sound statistical design. Statistically designed experiments(Mason et al., 1989; Scheiner and Gurevitch, 1993) differ frommonitoring experiments designed empirically. In EPA (2000) dif-ferent types of sampling designs are presented; judgmental, simplerandom, stratified, systematic / grid, ranked set, adaptive clusterand composite sampling. In addition, guidance is provided onhow to select the most relevant sampling design depending onthe problem to assess. However, the number of field experimentsdesigned on a sound statistical background is still limited (Kitsiouet al., 2001; Legendre and Legendre, 2003; Underwood, 2007).Experimental designs for assessing biological and ecologicalimpact, aiming at maximizing resolution have been proposed byMillard and Lettenmaier (1986). The proposed monitoring designis based on the structure of an analysis of variance factorial exper-iment (Underwood, 2007) that allows the evaluation of seasonali-ty, pre/post status as well as pairing of treatment–control samplingsites. Secondary effects such as variation of the variables measuredwith depth are also included in the model. A design proposed by





Millard and Lettenmaier’s was aiming at minimizing sampling costat a given level of detection of difference; therefore, the programbudget is reduced or a higher statistical power is achieved at thesame cost (Green, 1989). The work of Millard and Lettenmaier(1986) is rather theoretical. A more applied work on selecting sam-pling sites has been given by Kitsiou et al. (2001) where an optimalsampling design has been proposed. Legendre et al. (1989) devel-oped a sampling design in a stepwise manner: (a) preliminarysampling: space intensive sampling (63 sampling stations) basedon 10 variables only; (b) final sampling design: four methods wereapplied to choose 10 out of 63 sampling sites. This resource man-agement allowed more variables to be used providing a betterunderstanding of ecological processes through modeling. The GoodPractice Guide Technical Application (GPG) document (Cefas, 2010)gives recommendation on how scientific microbiological monitor-ing programmes can be performed. As a first step, a shoreline sur-vey is recommended for detecting potential sources of marinepollution. Determination of the circulation of pollutants is also sug-gested based either on simple study of the bathymetry or on theapplication of hydrodynamic and particle track models. Conse-quently, determination of the sampling plan can be performed aswell as the number and location of representative sampling pointsand the frequency and timing of sampling.

However, there are shortcomings in monitoring designs even inmultinational programs. In the report of the OSPAR (Oslo/PARis)Commission (OSPAR, 2009), it has been discussed that the moni-toring variables used for eutrophication assessments were not al-ways sufficient due to poor representativeness in space and timeas well as to lack of data in some areas. Improvements on samplingfrequency and spatial coverage were suggested. Sampling fre-quency is another aspect in monitoring programs that is alsocostly. Mandatory sampling should be carried out on a seasonal ba-sis. However, sampling on a monthly basis is strongly recom-mended. Optimal sampling frequency should be decided upon asa compromise between varying variability and the best cost-effective strategy of the sampling program (UNEP, 2003).

4. Marine water quality monitoring variables

An important component in the governance network of regionalseas is the United Nations Environment Programme (UNEP)launched in 1974. Among 13 regional sea programs, six are admin-istered by UNEP. The program includes the Mediterranean, theBlack Sea, the Baltic Sea, Wider Caribbean, East Asian Seas (COB-SEA), South Asian Seas, North-East Pacific, North West Pacific,Red Sea and Gulf of Aden, South-East Pacific and Western Africa,covering 18 regions of the world with the participation of 149countries (DiMento and Hickman, 2012). UNEP supports monitor-ing projects to assess environmental impact. The most significantactivities that require monitoring are (UNEP, 1999): domesticwastewater, petroleum refinery effluents, urban development,chemical industry, food industry, tourism, energy production, agri-culture, tanning industry, aquaculture and shipping. The type ofactivity as well as the impact on the marine environment andthe biota indicates the type of variables to be monitored (UNEP,2003; OSPAR, 2009; Karydis and Kitsiou, 2012). The mandatoryvariables necessary as background information in a monitoringprogram are: temperature, salinity, water transparency, ortho-phosphate, total phosphorus, silicate, dissolved oxygen, chloro-phyll, total nitrogen, nitrate, ammonium and nitrite,phytoplankton (total abundance, abundance and major groups,bloom dominance). The introduction of biological parameters isalso recommended (UNEP, 2003). These include detailed analysisof phytoplankton (species composition and possibly HighPerformance Liquid Chromatography (phytoplankton pigment


studies), benthos (specific composition and population dynamicsof phyto-, meio- and macro-zoobenthos). Measurements on ben-thic communities are particularly important in transitional waterswhere benthic communities are the dominant elements (Zaldivaret al., 2008). A realistic estimation of residence time based on pre-vailing current patterns and water mass dynamics is also desirable.Methods on remote sensing for estimation of chlorophyll, sus-pended matter and turbidity are also recommended as supplemen-tary techniques (UNEP, 2003). Automatic buoys on fixed platformsare also useful in recording monitoring environmental parameters(dissolved oxygen, chlorophyll a and some of the nutrient) on acontinuous basis. If a monitoring program is aiming at evaluatingthe impact from hazardous compounds, the following substancesshould be taken into account (OSPAR, 2007; UNEP/MAP, 2011):(a) heavy metals and their compounds; mercury, cadmium, lead,zinc, copper and chromium (b) organometalic compounds, dioxinaand furans (c) oils and hydrocarbons of petroleum origin; polyaro-matic hydrocarbons (d) other organohalogen compounds such as:halogenated aliphatic hydrocarbons, halogenated aromatic hydro-carbons, chlorinated phenolic compounds and organohalogenatedpesticides (e) biocides such as pesticides, fungicides, herbicides,insecticides and slimicides.

Apart from monitoring contaminants the emphasis of monitor-ing the last decade has been on the marine biota. In addition to therequirements of the marine pollution assessment and control com-ponent of the Mediterranean Action Plan known as MED POL(DiMento and Hickman, 2012), a number of European Union Direc-tives require or encourage collection of information concerning thebiota. The Marine Strategy Framework Directive (EC, 2008) suggestsmonitoring on biological characteristics, pressures and impactsreferring to habitat types and biological features (marine commu-nity structure and dynamics, species composition, seasonal and bio-logical variability). The Water Policy Directive (EC, 2000) that dealswith coastal waters requires monitoring of biological elementsincluding composition and abundance of phytoplankton and otheraquatic flora as well as invertebrate fauna. The HELCOM Convention(HELCOM, 2008) sets ecological objectives by using indicators ori-ented towards specific targets reflecting ‘‘Good Environmental Sta-tus’’. It is obvious that collection of information on biota assumesuse of ecological knowledge, an element that requires understand-ing the potentiality and limitations of ecological tools.

5. Ecological knowledge on marine water quality monitoringstudies

In spite of the optimistic attitudes during 1950s and 1960s, eco-system theories developed for the marine environment have rarelycontributed to the understanding and solution of environmentalproblems (NRC, 1986; de Jong, 2006). MacGarvin (1995) has criti-cized the application of ecosystem theories for the marine environ-ment: ‘‘the fact that theoretical ecologists, working in far easier fieldsthan marine ecology, are now asking such searching questions of theirmethods, highlights how unreasonable it is to expect that we can pre-dict the effect of human actions upon marine ecosystems with anyaccuracy’’. On the contrary, ecological knowledge apart from eco-logical theories includes observations, facts, research methodology,data collections and synthesis of information, all of them contrib-uting to approaching a wide range of environmental problems.Ecological knowledge provides the tools for water quality assess-ment, prediction and management of environmental impacts,protection and restoration of species and ecosystems, protectionof environmental quality as well as understanding effect mecha-nisms of compounds of anthropogenic origin. Successful applica-tion of ecological knowledge eliminates the degree of uncertaintyin assessment and forecasting of future trends. The problem of




(a)(b)

Fig. 1. Total number of publications on (a) ‘‘marine water quality monitoring’’ and(b) ‘‘marine water quality monitoring’’ focusing explicitly on ‘‘policy’’ during theperiod before 1990 – to date (Source: Web of Science).


uncertainty is far more serious in ecological work due to a numberof biological sources: (a) ecosystem complexity: as the relation-ships among species are complex, changes in the presence of onespecies may have unexpected changes in the dynamics of otherspecies of the community or the assemblage (b) transfer of toxiccompounds through the food chain; a well known example is thecase with DDT. Although it was introduced in the terrestrial envi-ronment, eventually it was found in fishes and seabirds (c) non-linear response to perturbation: an unexpected outcome may beobserved if a threshold value is exceeded. For example, it is knowna long time ago that fish populations can collapse beyond a point ofoverfishing (Gallucci et al., 1996; Karagiannakos, 1995). Regimeshifts is another example of non-linearity and has been studiedin relation to eutrophication of transitional waters (Zaldivaret al., 2008) (d) natural variability: populations and communitiesvary in time and space due to intrinsic processes. In a biologicallyvariable system, a single measurement is just a ‘‘snapshot’’ of onestate among many possible states (NRC, 1986) (e) random varia-tion: driving forces on a population are more or less random. Pop-ulation dynamics can be predicted rather on a probabilistic than ona deterministic base. This is a problem when numerical models areapplied to forecast further trends (f) error of estimation: as com-munity structure is expressed in both qualitative (species compo-sition) as well as quantitative terms (number of individuals perspecies), sources of error are more significant than in scalar values.The error also depends on the difficulty to classify specimens, aswell as the rarity of some species in the sample that followbinomial distribution and (g) lack of knowledge: lack of reliabletheories or models restrict the value of data sets. These shortcom-ings should be taken into account if biological (or ecological)monitoring is planned.

Biological monitoring can be defined ‘‘as the systematic use ofbiological responses to evaluate changes in the environment withthe intent to use this information in a quality control program’’(Matthews et al., 1982). The observed changes are related to possi-ble human pressures that may be toxic compounds (industrialeffluents, oil spills), nutrients (sewage effluents, fertilizers) or eventhermal effluents (power stations). The emphasis in the biologicalmonitoring should be placed on biomass changes, in communitystructure changes as well as on the interactions between ecologicalgroups and abiotic factors. It is therefore obvious that a holisticanalysis of the ecosystem is required. A biological monitoring pro-gram can be designed based on the functional point of view (pri-mary productivity, respiration rate, assimilation rates, anddecomposition rates) or on structural aspects (Phytoplanktonic,zooplanktonic, benthic and bacterial community). If funding isavailable a combination of the two is possible. However, thereare two shortcomings in field observations (Cairns, 1981): (a) man-agement practices are more efficient if planned to prevent negativesituations: a simple documentation of the damage does not helpvery much and (b) a direct observation in a system does not leadto conclusions applicable to other systems: each physical systemis unique. An answer to this problem is relevant research at labo-ratory level. Toxicity testing can give answers to a number of ques-tions (Buikema et al., 1982) such as: (a) at what concentrationspecific compound is lethal to an organism (b) effects at sublethalconcentrations (c) relative toxicity of a set of compounds (d)detection of sensitive species (e) environmental conditions affect-ing toxicity (f) fate of the toxicant in the marine environment (g)short term effects of toxic compounds. If this information is avail-able the scientific team performing monitoring can understand andinterpret the behaviour of the system.

Many regulatory issues and authorities responsible forfinancing monitoring programs require the use of indicator speciesfor biological monitoring. Use of indicators increases the objectiv-ity of the information collected and decreases the budget of


monitoring. In addition, an indicator condenses extensive amountsof information into a single number. There are several advantagesusing indicators (NRC, 1986): living organisms function as contin-uous monitors; they accumulate past conditions in their body act-ing as a record of particular compounds. However, special care isneeded as: (a) some organisms alter chemicals in their tissues (b)some organisms increase the sensitivity of monitoring; as speciesaccumulate materials in their tissues they amplify weak environ-mental signals (c) organisms can provide information about com-plex mixtures of compounds; in the marine environment thereare usually many pollutants in the same site and (d) organismsprovide information on biological effects under natural conditions.The value of indicator species/community indices increases if thefundamental cause-effect relationships have been understood.

6. Data storage and management

Monitoring requires the storage of data in adequate databasesin order to be easily accessed, managed, retrieved and updated.Currently, geo-referenced databases fulfil these requirements andthey can be easily developed and managed in the framework ofGeographical Information Systems (GIS). GIS provide the meansfor handling, storing and analysing large data sets from heteroge-neous sources (Zheng et al., 2002; Fernandes and Valavanis, 2008).

Data storage and management in marine water quality moni-toring consists of several tasks such as data acquisition, qualitycontrol, data manipulation, interpretation and archiving. The inte-gration of these tasks with new user requirements and geo-information technological progress represent research challengesfor the implementation of up-to-date flexible database manage-ment systems accessible via user-friendly graphical interfaces.Such systems can represent a valuable source for sharing localinformation useful to managers and policy makers for the develop-ment of management strategies.

7. Data analysis

A wide spectrum of data analysis methods have been appliedfor the process and analysis of data sets acquired during monitor-ing surveys to assess marine water quality, including the use ofindicators, statistical methods, multiple criteria analysis, spatialmethods, modeling and integrated approaches (Kitsiou andKarydis, 2011). In this section, a short description of the methodsfocusing mainly on marine water quality monitoring that supportpolicy making is given. It is important to note that the total numberof publications on ‘‘marine water quality monitoring’’ (Fig. 1)shows a considerable increase after 1990, with the maximumnumber observed during the period 2005–2009. In addition, thenumber of publications on ‘‘marine water quality monitoring’’focusing explicitly on ‘‘policy’’ seems to increase after the year2000. There are many reasons accounting for this massive increase





in the number of papers published during the last two decades: (a)many international conventions (OSPAR, HELCOM, Barcelona Con-vention) shaped their monitoring requirements after an initialperiod where monitoring framework was rather loose. Theseconventions were revised during the 90s and their technicalappendices were amended (b) more resources became availabledue to the severity of the problems and public pressure and (c)the quality of scientific work has been improved. Scientists prefernowadays to publish monitoring work in international journals,setting interesting objectives instead of ‘‘dumping’’ them as techni-cal reports submitted to competent authorities.

7.1. Statistics

An issue which is of upmost importance in many water qualitymonitoring programs is data handling. Data handling includes thecompilation of data from various sources, quality assessment (i.e.checking whether the data comply to preset standards) and storageof data in databanks designed for easy retrieval of subsets of datafor data analysis (de Jong, 2006). Data analysis is a crucial step ina monitoring program as it will show whether the produced dataare sufficient to support the data analysis methods needed formaking correct regulatory decisions (Price, 1987). Data analysis as-sumes the existence of data; data sources can be local data collec-tions, national or international sources. Therefore, there are threelevels of data compilation (FAO, 2006): (a) primary informationor primary data; the data obtained from field measurements (b)secondary data: international data compiled from lower level data.It can be a national data base compiled by various local monitoringprojects and (c) tertiary data: these are international data compiledfrom monitoring programs of different international sources. Thesecan be data from different countries or different programs (i.e.MED POL and OSPAR). It is therefore necessary that scientists andpolicy makers should work together and agree upon the level ofenvironmental change that would discriminate between ‘‘no ef-fect’’ and ‘‘impact’’. Once these issues have been settled, the goalsof data analysis should be focused (a) to summarize the bulk ofdata (b) to test for differences and therefore for possible impact(c) to generate hypotheses (d) to forecast future trends and (e) toevaluate the degree of uncertainty concerning the conclusionsdrawn from the data (NRC, 1990). Statistical analysis is relativelyeasy to perform and can provide information that helps to charac-terize the data, assess the uncertainty associated with measure-ments and test for spatial and temporal changes (NRC, 1990).Although many statistical tests are suitable for many situations,monitoring program designers should be careful in selecting a par-ticular statistical method as it depends upon data characteristicsand the specific question being asked (Milliken and Johnson,1997). Kitsiou and Karydis (2011) in an extensive review on meth-ods for quantitative assessment of eutrophication discuss a widerange of data analysis methods that can also be applied in monitor-ing studies. In a review article on techniques of analyses of vari-ance (Underwood, 2007), the author investigated the complianceof the data used in a number of papers to the ANOVA assumptionsand found that in 151 cases these assumptions were violated. Thisis because many researchers analyzing data sets collected throughfield work, treat them as ‘‘nice data sets’’ and do not realize thatthey are actually handling ‘‘messy data sets’’ (Milliken andJohnson, 1997). Beninger et al. (2012) also review misapplicationand misuse of statistical testing and interpretation in the field ofmarine ecology.

7.2. Modeling

The use of models to evaluate either the ‘‘State-of-the-System’’or the ‘‘State-of-the-Environment’’ has an additional advantage


over statistical procedures: statistical methods inform us aboutyesterday’s news whereas modeling applications can detect andadvert tomorrow’s disasters. Parma (1998) suggested that ‘‘wecannot hope to understand the complex system we are trying tomanage unless we experiment with it’’. As a monitoring designcannot be an experiment ‘‘sensu stricto’’ the ‘‘experimentation’’can be performed as part of the modeling stage: different scenarioscan provide valuable information concerning future trends and im-pacts in the marine environment (Bergström et al., 2013).

Water quality models simulate the responses of aquatic organ-isms and ecosystems under various environmental conditions andanthropogenic pressures. Statistical models are empirical and de-rive from observations. They quantify a cause-effect relationshipwithout accounting for every component of the related process.Efforts on the development of models for assessing case-effect rela-tionships in water quality are dated since the sixties (Vollenweideret al., 1992) and continue to date (Latimer and Rego, 2010). Alinet al. (2012) applied multiple linear regression to develop empiri-cal models using oxygen, temperature, salinity and sigma theta asproxy variables to reconstruct pH, carbonate saturation states,carbonate ion concentration, dissolved inorganic carbon concen-tration and total alkalinity in the southern California CurrentSystem in order to assess the ecological impacts of ocean acidificat-ion due to coastal upwelling.

Numerical models although they have not been particularly de-signed for monitoring, they can be used for monitoring purposesand water quality management (Chapman, 1996) as they allow:(a) the identification of the most important variables of the system(b) interpretation of marine processes (c) forecasting of the im-pacts in the marine environment and (e) policy analysis. In litera-ture, various modeling approaches of one- or multi-dimensionalcharacter, simple or more complex, are presented. Numerical mod-els have been applied to understand and predict the impact fromeutrophication problems and toxic substances mainly hydrocar-bons but also metals and pesticides. Over the last decades severalmodels for marine biogeochemistry have been developed. Amongthem the European Regional Seas Ecosystem Model (ERSEM) isconsidered as a point of reference regarding modeled processesand modeling philosophy (Mateus, 2012). The Land–Ocean Interac-tions in the Coastal Zone (LOICZ), a core project of the InternationalGeosphere-Biosphere Programme (IGBP), includes modeling acrosscontinental basins for evaluating biogeochemical processes for C, Nand P; the LOICZ biogeochemical model (LBM) is based on the massbalance of water and materials (Crossland et al., 2005). As it wasfound unrealistic (Tett et al., 2003) to create a single model thatcould describe all interactions and variability of the real system,the idea of partial modeling approaches appeared in literature:models on phytoplankton dynamics (Le Gall et al., 2000) and onthe role of seaweeds in estuaries (Alvera-Azcarate et al., 2003)are examples of partial modeling. Zaldivar et al. (2009) proposeda model to predict regime shifts in shallow ecosystems and possi-bly establish threshold levels that cause the regime shifts. A goodmodel review on aspects of eutrophication has been given byNRC (2000). Some of them are suitable for monitoring purposes.

Modeling without calibration and verification processes hasusually minimal credibility; however, most of the times these pro-cesses have major requirements for datasets. Grégoire et al. (2008)developed a 1-D coupled physical – biogeochemical model to sim-ulate the ecosystem of the central Black Sea at the end of 1980swhen it was seriously affected by eutrophication and invasion bygelatinous organisms. The coupled model extended down to thesediments and was calibrated and validated using a large set ofdata, including chla (about 2400 data points), oxygen (about11800 data points), ammonium (about 6600 data points), silicate(about 10800 data points). Another significant modeling effortwas part of the Chesapeake Bay Program; three years of data were





used for model calibration (Cerco and Cole, 1993) and the cali-brated model was verified against a 30-year dataset (Cerco andCole, 1995).

A good example of modeling approach for assessing marinewater quality is given by Håkanson and Bryhn (2008) focusing oneutrophication in the Baltic Sea. Both empirically based modelsand the CoastMab validated dynamic process-based model wereapplied. CoastMab is designed to give monthly values for phospho-rus concentrations in water and sediments. Thousands of testswere performed for the calibration and validation of the modelusing empirical data. The model was also applied to test remedialscenarios and remediation strategies for the Baltic Sea. As a result,specific suggestions were made to policy makers for reducingeutrophication to desired levels.

7.3. Multiple criteria analysis

The assessment and trend predictions by the use of simulationmodels based on monitoring data on the quality of the marineenvironment can be quite accurate; however, it is not possiblefor the policy maker to come to a regulatory scheme on a purelyobjective basis, since social and economic aspects have also (ifmainly) to be taken into account. Many groups of interest andauthorities are involved in monitoring including: local and stateregulatory agencies, harbour and port agencies, developers, scien-tists associated with consulting firms, the residents of the area, lo-cal fishermen, local merchants and groups of professionals onrecreational activities are among the bodies with special responsi-bilities or interests in the area. Sometimes the responsibilitiesoverlap and their interests conflict. A quantitative tool that helpsthe policy maker to come to a decision based on quantitative orsemi-quantitative criteria (in both metric and non-metric form)is the multiple criteria analysis (MCA) (Janssen, 1992). MCA canincorporate environmental, social and economic values and a num-ber of scenarios can be examined. MCA provides the means forranking different scenarios known as ‘‘alternative choice possibili-ties’’ (Nijkamp and Voogd, 1986). Weights can also be assigned tothe criteria. There is a wide variety of methods (Janssen, 1992)which obey to the same principle: the pairwise comparisons ofthe values (scores) for all the alternative hypotheses and for eachcriterion. Brief descriptions of MCA methodology was examinedby Kitsiou and Karydis (2011). There is a wide range of applicationsof MCA on management issues both in terrestrial (Ananda and Her-ath, 2009) and aquatic systems (Moriki and Karydis, 1994a, 1994b;Kitsiou et al., 2002; Hajkowicz, 2007).

Fig. 2. A flow chart of policy making and implementation aiming at good marinewater quality.

8. The integrated approach

The effect of a specific policy should be continuously assessedthrough monitoring; however, only in few marine monitoring pro-grams the information provided is assimilated in integrated assess-ments due to the requirements and the complexity of such anapproach. As the final goal of a monitoring study is the establish-ment of a regulatory framework or the revision of the existing reg-ulatory framework, it is obvious that an integrated approach isneeded both at the scientific and policy level. The need for an inte-grated environmental approach has already been recognized byvarious authors (Cloern, 2001; Duarte, 2009). Nixon (1995) statedthat anthropogenic activities in coastal areas are related with socialstructure and social dynamics of the area; therefore, marine waterquality monitoring should be carried out in a broader contextincluding socio-economic aspects. The latter is of major importancesince policy makers adopt measures for environmental protectionbased on the scientists’ proposal as well as the stakeholders’ andpublic opinion. An integrated approach leading to policy making


should take into account the following: (a) information on waterquality variables assessed in the monitoring program (b) quantifi-cation of anthropogenic pressures as assessed during monitoring(c) social attitudes (d) economic aspects: activities contributing toenvironmental contamination; aquatic resources depending onwater quality: fishing, aquaculture and recreation (e) public partic-ipation groups: environmental organizations, fishermen, fish con-sumers, coastal recreationists and associated business.

The complexity of environmental marine processes, as alreadymentioned, requires a cross-disciplinary collaboration from scien-tists in order to convey their outputs to policy makers. However,it is recognized that scientific knowledge does not directly deliverpolicy. In this framework, a number of global monitoring systems,such as LOICZ, Global Ocean Observing System (GOOS), GlobalOcean Ecosystem Dynamics (GLOBEC) and Integrated MarineBiogeochemistry and Ecosystem Research (IMBER) have beendeveloped. The LOICZ approach refers to the study of global envi-ronmental change. The challenge of LOICZ is to combine the exper-tise of natural and social sciences for studying the coastal zone asone system, aiming to assess the impact of human activities onnatural systems (Crossland et al., 2005). LOICZ intends to derive in-dexes related to ecosystem change based on data collected duringroutine monitoring activities. The GOOS is a global system allowingobservations, modeling and analysis of marine and ocean variablesfor monitoring and forecasting the state of the sea taking into ac-count marine environmental conditions and social impacts. Inthe framework of GOOS, all data are accessible to the public andresearchers. The Chlorophyll Global Integrated Network (Chloro-GIN) sponsored by GOOS, supports the in situ measurement ofchlorophyll in combination with estimates from satellite data.GLOBEC, a study of Global Ocean Ecosystem Dynamics initiatedin 1990 aimed ‘‘to advance our understanding of the structure andfunctioning of the global ocean ecosystem, its major subsystems, andits response to physical forcing so that a capability can be developed





to forecast the responses of the marine ecosystem to global change’’.The Integrated Marine Biochemistry and Ecosystem Research (IM-BER) project focused on a comprehensive understanding of oceanresponses to accelerating global change and estimation of the con-sequent effects on the Earth System and human society.

The interaction between the administrative estate, stakehold-ers, public participation, regulatory issues, implementation ofmeasures and monitoring are given diagrammatically in Fig. 2.The administrative estate sets the agenda after consultation –negotiation with the stakeholders and the public. Water qualitymonitoring programs set ahead have a twofold objective: (a) to ac-cess the marine water quality in general (operational monitoring)as well as marine ecosystem quality and (b) to carry out researchfor understanding cause - effect relationships and processes (inves-tigative monitoring). The monitoring outcome provides scientificinformation useful for the development of regulatory issues; theirimplementation will give feedback to administrative estate andwill be followed by compliance monitoring to assess whether thevalues of marine water quality variables are within the preset tar-gets. The arrows show the flow of information. The right-to-knowby the public and the stakeholders is satisfied at every step of thepolicy design and implementation procedure.

9. International conventions and EU directives with monitoringcomponents

9.1. Environmental governance

Marine environmental protection and management require awhole range of activities and regimes known as ‘‘governance’’(DiMento and Hickman, 2012). These include international laws,international conventions and policies for the protection of themarine environment as well as marine monitoring programs. Inthose complex processes there is always interlink between policyand science. Science does not only provide policy makers withinvaluable quantitative ‘‘environmental information and tools’’but also as scientific authority helps policy measures to be ac-cepted by the public. Monitoring programs play a central role inproviding information for building legal frameworks as well asfor assessing the effectiveness of management practices. Most ofthe international conventions contain a monitoring component.In this section international conventions of regional seas and theirmonitoring component will be reviewed shortly.

9.2. The Mediterranean Sea

State and pressures of the marine and coastal Mediterraneanenvironment have been reviewed extensively (Haas, 1990; EEA,1999; Danovaro, 2003; Karydis and Kitsiou, 2012). The institu-tional basis for the protection of the Mediterranean Sea is the Bar-celona Convention. The ‘‘Barcelona Convention’’ was adopted in1976 by the Mediterranean countries for the protection of theMediterranean Sea against pollution. The objective of the Conven-tion is to reduce pollution in the Mediterranean Sea, protect and im-prove the marine environment in the area, thereby contributing to itssustainable development. The Mediterranean Countries and theEuropean Union adopted the Mediterranean Action Plan (MAP)with objectives to reduce pollutant inputs and improve quality oflife. The scientific and technical component of MAP is the MEDPOL program. The MED POL monitoring program has placedemphasis on oil pollution, heavy metals (mercury, cadmium, cop-per and zinc), organohalogen compounds (PCBs and DDTs), organo-tins (TBTs) and eutrophication. The aim of the MED POL monitoringprogram is threefold: (a) to assess and quantify sources of pollu-tion (b) to monitor and assess ecosystem health and (c) to monitor


the success of implementation plans and measures taken to miti-gate marine pollution in the Mediterranean Sea. In addition tothe monitoring of pollutants in the Mediterranean, MED POL mon-itoring also takes into account a number of factors needed for theunderstanding of the relationship between the socio-economicaspects of coastal zone development and the pollution problem(Jeftic, 1993).

9.3. The Black Sea

Pollution problems in the Black Sea are serious and often due tonon-point sources. Main pollution sources are industrial effluentsand sewage waste. In addition, tourism is a severe pressure overthe last twenty years. The Convention on the Protection of theBlack Sea against Pollution (the Bucharest Convention, 1992), cameinto force in 1994. The governance of the Black Sea is based on theBlack Sea Commission (BSC) supported by Convention Protocolsand Strategic Action Plans (SAPs). The Black Sea Integrated Moni-toring and Assessment Programme (BSIMAP) has been set up bythe Black Sea signatory States. These States have agreed that theBSIMAP would run on existing national monitoring programs. Theyhave also agreed to harmonize assessment, methodologies andanalytical techniques. They have also agreed to apply inter-calibra-tion exercises and to develop standard reporting formats.

9.4. The Baltic Sea

Today’s degraded status of the Baltic Sea is due to elevated con-centrations of nutrients as well as hazardous substances includingmetals, PCBs, DDT, HCB and TBT’s. The Convention on the Protec-tion of the Marine Environment of the Baltic Sea Area (HELCOM)was signed in 1992 and entered into force in 2000. The ContractingParties have to take legislative, administrative or other relevant mea-sures to prevent and eliminate pollution in order to protect ecologicalintegrity and preserve ecological balance (HELCOM, 2008). The ma-jor environmental problems monitored in the area are eutrophica-tion, hazardous substances, biodiversity and maritime activities.Eutrophication seems to have received specific attention (Boeschet al., 2000). Twenty years of the Baltic Sea protection programshave shown that scientific knowledge obtained through monitor-ing activities has significantly contributed to regulatory prioritiesconcerning the implementation of environmental management inthe area (Backer et al., 2010).

9.5. The North Sea

The Oslo and Paris Commissions responsible for the regulation,monitoring and assessment of pollution in the North Sea signed theOSPAR Convention in 1992. This Convention does not deal onlywith marine pollution but is also concerned with adverse effectsfrom human activities (Tromp and Wieriks, 1994). In addition tomonitoring programs developed to record pollutant concentrationssuch as polyaromatic hydrocarbons, tributylin and heavy metals,OSPAR Commission have adopted monitoring programs aiming atassessing the biological impact (Stagg, 1998). Among the prioritiesof the OSPAR Conventions is the monitoring of eutrophication (deJong, 2006). Ecological Quality Objectives (EcoQOs) have been setto monitor inputs and effects (Painting et al., 2005). Monitoringwinter nutrient concentrations was considered as a good EcoQO,since in coastal and estuarine waters, nutrient concentrations arehigh in winter due to run offs resulting from increased precipita-tion. Monitoring phytoplankton chlorophyll a concentrations dur-ing the growing season (March–September) was a satisfactorybiomass parameter. The use of environmental indicators in marinemonitoring to meet OSPAR Convention’s regulatory needs has beenreported by Johnson (2008).





9.6. The Wider Caribbean Region

The Wider Caribbean Region (WCR) bordered by the Antilles tothe east is a semi-enclosed area connected to the Atlantic Ocean.The WCR encompasses the Caribbean Sea, the Gulf of Mexico aswell as adjacent marine areas and bays. Twenty-eight out of 36states and territories in the area are UN member states (UNEP,2010). Due to its great area (about 15 million km2), WRC is brokeninto subsystems, most of them representing distinct ecosystementities. Contributors to the environmental impact of the area areindustry, agriculture, mining, oil pollution and tourism (CEP,2010). Ranking the environmental threats damaging the marineenvironment, tourism is the first in the list due to 100 million tour-ists per year. Eutrophication has been ranked as the second great-est problem since 6000 tons of phosphorous and 13000 tons ofnitrogen is dumped into the marine environment every year. Toprotect the marine environment of the WCR, the Cartagena Con-vention was adopted in Colombia in 1983 by more than 20 nationalsignatories and entered into force in 1986. The Convention wassupplemented by three protocols on combating oil spills, on Spe-cially Protected Areas and Wildlife Convention (SPAW) as well asthe protocol on Pollution from Land Based Sources and Activities(LBS). The LSB protocol deals with the assessment of marine pollu-tion. Article VI of the LBS Protocol is concerned with monitoringand assessment programs: ‘‘especially identify and assess patternsand trends in the environmental quality of the Convention area’’.The Integrated Planning and Institutional development for theManagement of Marine and Coastal Resources (IPID) part of theCaribbean Environment Programme deplores activities that in-clude monitoring of recreational waters.

9.7. The East Asian Seas

The Seas of East Asia include the Philippine Sea, Sulu Sea, TimorSea, Celebes Sea, Arafura Sea, Banda Sea, Flores Sea, South ChinaSea, Java Sea, the Straits of Singapore, the Straits of Malacca, theOceans of Australia and the Andaman Sea (UNEP, 2008). Ecosys-tems in the East Asia are distinct and characterized by high diver-sity. Intense population density, unprecedented economic growthand rapid urbanization of East Asia are interconnected with envi-ronmental problems in the marine environment. Pressures on mar-ine water quality result from industrial wastes, domestic sewage,aquaculture practices, agricultural activities and tourism. Landbased sources are the main source of pollutants and they accountfor the 77% of the total marine pollution. The Action Plan for EastAsia of the UNEP Regional Seas Programme (RSP) was approvedin 1981. The Coordinated Body of Seas of East Asia (COBSEA) actsas a secretariat (COBSEA, 2008). The New Strategic Direction forCOBSEA, 2008–2012 places emphasis on monitoring practices.

9.8. The West and Central African Region

The West and Central African Region (WACAF) extend over8421 miles of coastline and support a population of about 320 mil-lion people (UNEP, 2002). The region is made up of 22 countries,most of them coastal and densely populated. The economy is basedon tourism, fishing and industry. Lack of environmental regulationsand intense economic growth, have adopted a form of develop-ment at the expense of environmental conditions (UNEP, 2002).Water quality in WACAF seems to be a major problem. Sizableamounts of residues resulting from mining activities are dis-charged into the marine environment. These residues includephosphate, mercury and zinc. Suspended particles are anothersource of pollution mainly due to diamond mining. Oil pollutionis also serious: in the case of the Niger Delta, there have been esti-mated as high as four million tons (Heileman, 2009). The WACAF


Action Plan has been part of the UNEP Regional Seas Programmesince 1981. The WACAF Regional Convention also known as theAbidjan Convention is accompanied by protocols addressing pollu-tion, coastal erosion and Environmental Impact Assessment (EIA).There have been regional agreements for environmental qualitystandards and monitoring protocols (Heileman, 2009).

9.9. The regional seas in USA

In USA, a number of federal programs are addressing marinepollution problems, including eutrophication, petroleum hydrocar-bons and heavy metal pollution. They encompass regulatory andpolicy issues as well as educational aspects (NRC, 2000). All theseprograms in addition to management strategies have monitoringand assessment components addressing mainly coastal conditions.The Gulf of Mexico Program (GOMP) is focusing at the restorationof coastal marine areas of the Gulf of Mexico. The Great WatersProgram, although is basically aiming at air pollution is alsoconnected with impact of atmospheric pollutants in the marineenvironment. Specific elements of the Great Water program in-clude monitoring components. The Outer Continental Shelf (OCS)has provided basic authority for the development of resources ofthe Outer Continental Shelf. As they have been concerns on possi-ble effects on the coastal areas OCS Amendments have includedmonitoring on marine environmental quality (Table 3).

9.10. Environmental governance in the European Union marineenvironment

EU Directives on the protection of coastal marine environmentalso include monitoring activities (Table 2). The Water FrameworkDirective (WFD) aims at establishing a framework for water pro-tection including coastal and transitional waters. WFD targets i.e.protection of aquatic ecosystems, promotion of sustainable useand improvement of the quality of the aquatic environment, pro-gressive reduction of pollution and protection of marine watersare among the WFD priorities. The implementation of the basicadministrative requirements of the Directive includes the develop-ment of monitoring and assessment programs. This requires theselection of suitable variables and indicators that can quantify bothwater and ecosystem quality. However, the WFD provides criteriafor designing surveillance monitoring programs of ecological andchemical status of sea waters. It also provides criteria for selectionof monitoring sites, frequency monitoring, additional monitoringrequirements for protected areas such as habitat and species pro-tection areas and standards for monitoring of quality elements.Although these guidelines have been given for WFD, they can formthe basis for designing any environmental monitoring program asthey cover all aspects of monitoring in a comprehensive way. Thecross-border aspects of implementing the WFD means that individ-ual Member States cannot design monitoring strategy in isolationfrom their neighbors (Downie and Baxter, 2004). Van Hoey et al.(2010) have attempted a realistic approach for adopting the eco-system concept, the development of benthic indicators, the defini-tion of pristine waters and human pressures to the WFDrequirements. The monitoring concept for coastal and transitionalwaters encourages the integration of water quality monitoring andbiological monitoring to basin of system oriented monitoring; in addi-tion, research on cause-effect studies has been recommended bythe same authors.

The Marine Strategy Framework Directive (MSFD) establishes aframework with main objective to maintain or achieve good envi-ronmental status in the sea. Fulfillment of this objective requiresprevention and/ or reduction of inputs in the marine environmentas well as the implementation of protection measures. Marinewater quality monitoring and regulatory issues will be the tool




Table 3Laws and European Union Directives encompassing mandatory monitoring components.

Laws/directives Monitoring components and objectives

Outer Continental Shelf Lands Act Amendments of 1978 (USA) Section 20(b): ‘‘monitor the human, marine and coastal environments. . . for thepurpose of identifying any significant changes in the quality and productivity ofsuch environments, for establishing trends in the areas studied and monitored. . .’’

EU Directive 2000/60/EC (EC, 2000) On establishing a framework for communityaction in the field of water policy

Annex V, Article 1, §3 Monitoring of ecological status and chemical status forsurface waters: ‘‘The surface water monitoring network shall be established inaccordance with. . . and provide a coherent and chemical status within each riverbasin’’

EU Directive 2006/7/EC (EC, 2006) (Concerning the management of bathing waterquality and repealing Directive 76/160/EC)

Article 1.’’This directive lays down provisions for (a) monitoring and classificationof bathing water quality. . .’’

EU Directive 2008/56/EC (EC, 2008) Establishing a framework for communityaction in the field of marine environmental policy (Marine Strategy FrameworkDirective)

Introductory Section: §21 ‘‘It is crucial for the achievements of the objectives of thisDirective to ensure the integration of conservation objectives, managementmeasures and monitoring and assessment activities. . .’’ and §21 ‘‘The next steptowards achieving good environmental status should be the establishment ofenvironmental targets and monitoring programs for ongoing assessment, enablingthe state of the marine waters concerned to be evaluated on a regular basis’’

Table 2Monitoring components in international conventions for the protection of the marine environment.

International conventions Monitoring components and objectives

United Nations Environment Program (UNEP)-MediterraneanAction Plan (MAP). UNEP/MAP (2005)

Convention for the protection of the Marine Environment and the Coastal Region of the Mediterranean.Article 12: Monitoring ‘‘The Contracting Parties shall endeavour to establish... programs at bilateral ormultilateral levels, for pollution monitoring in the Mediterranean Sea Area and shall endeavour toestablish a pollution monitoring system for that area’’

The Bucharest Convention – Convention on the Protection ofthe Black Sea Against Pollution (1992)

Article XI, §4: ‘‘The contracting parties shall inter alia establish through the Commission and, whereappropriate, in corporation with International Organizations they consider to be competent,complementary of joint monitoring programs covering all sources of pollution and shall establish apollution monitoring system for the Black Sea measuring, evaluating and analyzing the risks or effectsof marine pollution of the marine environment of the Black Sea’’

HELCOM (2008) (Convention on the Protection of the MarineEnvironment of the Baltic Sea Area, 1992)

HELCOM Convention, ANNEX III: ‘‘Criteria and measures concerning the prevention of pollution fromland based sources. Regulation 5: Monitoring and Evaluation. The contracting parties shall describe theimplementation and monitoring of measures in this Annex in their international programs. . .’’

OSPAR (2007) (Convention for the protection of the MarineEnvironment of the North-East Atlantic)

OSPAR Convention, Annex IV on the assessment of the quality of the Marine Environment. Article 2:‘‘For the purposes of the Annex, the Contributing Parties shall: (a) cooperate in carrying out monitoringprograms and submit the resulting data to the Commission’’

Cartagena Convention – Convention for the Protection of theCaribbean Sea (1983)

Land Based Sources (LBS) Protocol, ‘‘Article VI of the LBS Protocol addresses the issue of monitoring andassessment programmes. It states that such programmes may inter alia: (a) systematically identify andassess patterns and trends in the environmental quality of the Convention area and (b) assess theeffectiveness of measures taken to implement the Protocol.’’

Coordinated Body on the Seas of East Asia (COBSEA) Chapter 1, §2: ‘‘steps are urgently needed. . . and monitoring’’WACAF Area Abidjan Convention (enforced in 1984) Article 14: Scientific and Technological co-operation ‘‘. . .the contracting parties shall endeavour to

participate in international arrangements for pollution research and monitoring’’


for implementing measures for good quality of the marine environ-ment. Member States are encouraged to adopt specifications andstandardized methods for monitoring and assessment. The Direc-tive considers as a crucial objective the integration of conservationobjectives, management measures and monitoring activities. Theemphasis on monitoring is placed on an ecosystem – based ap-proach. Ferreira et al. (2011) have overviewed many indicators toaccess environmental status. Nutrient inputs, primary productionand algal biomass regulation, harmful algal blooms, resilience ofmarine ecosystem selection of indicators were among the topicsto be addressed. MSFD requires monitoring programs to be estab-lished and implemented by 2014 so that good environmentalstatus could be achieved in all European Seas by 2020. Althoughthe directive mentions physical, chemical and biological features,habitat types and health issues, it does not go into detail as faras monitoring program design is concerned. It leaves this toresearchers and consequently, there is a lot of ongoing research to-wards setting criteria, selecting variables and indicators, suitable tosatisfy the requirements of the Directive.

10. Discussion

The reasons for performing marine water quality monitoringstudies in the marine environment can be: (a) fulfillment of


requirements from legislation (NRC, 1990). This is also known ascompliance monitoring. (b) Detection of trends over a period oftime (UNEP, 1997, 2003); this type of monitoring refers mainlyto ‘‘hot spots’’ and ‘‘coastal waters’’ (UNEP/MAP, 2004) and (c) toassess the effectiveness of management actions. The first case is re-lated to the implementation of legislation, the second and the thirdare connected with policy measures and management practices.The Mediterranean Action Plan (MAP) that includes the MED POLmonitoring component, has been recognized ‘‘as the most success-ful example of international environmental collaboration’’ (Haas,1990) and continuous to be so (DiMento and Hickman, 2012).Twelve out of 17 Mediterranean countries affirm the existence offramework laws or guidelines for the protection of the marineenvironment (UNEP, 1999). The governance of the MediterraneanSea encompasses a number of interrelated components such as re-gional treaties, integrative planning, administrative support, coor-dinated research and the monitoring component (Haas, 1990). Themonitoring component, in addition to the reasons mentionedabove supports the ‘‘role of specialists in articulating state policiesunder conditions of technical uncertainty’’ (Haas, 1990). Anotherpoint that emerged over the last two decades, concerning the sig-nificance of monitoring programs and environmental impactassessments is the ‘‘right of the public to know’’. The right to infor-mation has been recognized as a basic human right shortly after





United Nation were formed (Beder, 2006): ‘‘freedom of informationis a fundamental human right and the touchstone for all freedoms towhich the United Stations is consecrated’’. The right to know is notonly based on human rights but ‘‘it is also a requirement of openand transparent government of a well functioning democracy’’ (CleanWater Fund, 2001). However, legislation of the public to have ac-cess on information concerning environmental issues followed afew decades later after the UN declaration. The European Unionin 2003 passed a directive ‘‘on Public Access to EnvironmentalInformation’’ (EC, 2003). The main objective of this Directive is‘‘to guarantee the right of access to environmental information heldby or for public authorities’’. Environmental information as definedby the Directive includes the state of the environment, the stateof human health and safety, administrative measures such asplans, programs and policies as well as reports on the assessmentof environmental quality and on implementation of environmentalmeasures. Relevant legislation was introduced in USA in 1986. TheEmergency Planning and Community Right-to-know Act (1986) al-low the authorities and the public to have access to information onhazardous chemicals as well as on chemical discharges in the mar-ine environment. International organizations like OECD (1996) haspassed a recommendation with respect to public information‘‘member countries take steps to establish, as appropriate, implementand make publicly available a pollutant release and transfer register(PRTR) system . . .PRTR systems should provide data to support theidentification and assessment of possible risks to humans and the envi-ronment by identifying sources and amounts of potentially harmfulreleases and transfers to all environmental media’’. Information tothe public is especially mentioned in the HELCOM Convention(HELCOM, 2008). Article 17 of the Convention requires from theContracting Parties to ensure that information is available to thepublic. This information includes ‘‘results of water quality andeffluent sampling carried out for the purposes of monitoring’’,‘‘water quality objectives’’ as well as allowing requirements to bemet. The European Union Directive on bathing water quality (EC,2006) refers clearly to the provision of information to the public(Article 1).

As mentioned in a previous section numerous state laws (NRC,2000), European Union Directives (EEC, 1979, 1991) and interna-tional conventions require monitoring of marine water qualityparameters. However, the effectiveness of a monitoring programdepends a great deal on the availability of financial and human re-sources. Financial constraints fairly often result into lack of conti-nuity which is a serious defect as far as the reliability of theinformation is concerned, especially in the trend monitoring. Wolfeet al. (1987) have stressed the implications from limited financesfor science and management; lack of funding does not allowhypothesis formation and testing, a necessary tool for document-ing ecological variability regimes (Likens, 1983) and interpretingenvironmental trends (Coull, 1985). The importance of structuringthe design of data collection within a hypothesis testing frame-work has long been recognized (Livingston, 1987; Wolfe et al.,1987; Kingston, 2002). Cuts in funding can also have a negative ef-fect on work planning. Reduction in the number of sampling sites,frequency of sampling, variables measured and number of repli-cates can increase the degree of uncertainty in interpretation anddecision making. Environmental impact is sometimes intangibleand therefore difficult to detect so as to document the damage.Higher uncertainty means that more preventive measures shouldbe taken. The financial benefit of the funding authority is revertedinto a much higher cost during the phase of implementation: ‘‘themore uncertain the threat, the greater the degree of precautionrequired’’ (Deville and Harding, 1997).

As many assessments on environmental impact are based onstatistical methodology, the possibility for Type I and Type II errorsdue to limited field information increases. If the conclusion is that


impact has been detected and this conclusion happens to be wrong(Type I error), any measures limiting economic activities and thesubsequent social hardship are unnecessary (Mapstone, 1995);these measures can have an impact on both the economy andthe society. Loss of competiveness is the first impact to appear inthe area. Measures to protect the environment increase the costof production if this extra cost is taken unilaterally. In other words,if the environmental cost does not apply to all states, makes theproducts less competitive. This reduces the income in the areaand also leads to an increase of unemployment. As environmentallaws and standards are implemented, many firms have eventuallyto close down. However, this argument is controversial as new jobsare often created. The other end of statistical error is the ‘‘no im-pact’’ assessment (Type II or error b). This means that a seriousenvironmental degradation has occurred leading to deteriorationof marine ecosystem extinction of species or even collapse of fish-eries. According to Fairweather (1989) a Type II error is even moreserious because environmental remediation is more costly and animpact on the economic cost and employment is going to incur.The possibility of error Type I and II increases if the assessmentis based on biological information due to seasonality and built-invariability of the system.

Ecological indices that are fairly often used in water qualitymonitoring impact studies due to their simplicity and objectivitymay induce both types of errors. Attempts to condense the bulkof monitoring data has led the researchers either to use ecologicalindices or to develop new indices suitable for detecting and quan-tifying ecosystem responses (Matthews et al., 1982). These indicesare suitable and therefore applicable if: (a) cover wide geographicareas (b) are sensitive and reproducible (c) are suitable for longterm monitoring and (d) are efficient in discriminating betweenseasonal fluctuations and stress effects (Hooper, 1969). There aretwo types of community response indices (Matthews et al., 1982;Karydis, 2009): (a) functional indices; they represent indices basedon rates. An example is production indices (i.e. ratio between pri-mary productivity and net microbial productivity) and (b) indicesbased on structural information. They include a wide range of indi-ces mathematically derived (Magurran, 2004). Indices expressingvariability as well as various trophic indices are also included.However, it is doubtful whether these indices can express theintegrity of biological systems. Monitoring the ecosystem integrityis also a requirement of recent EC Directives on the marine envi-ronment (EC, 2000, 2008).

It has been widely accepted (Janssen, 1992) that social, eco-nomic, ecological and environmental aspects should be taken intoaccount and included in policy making. Indicators seem to be anappropriate tool for condensing ecological information. Ecologicalindicators stand for an array of variables that designate as a wholemarine environmental quality of a certain area (Turnhout et al.,2007). Recent regulatory issues such as the Mediterranean ActionPlan (EEA, 1999), the Water Frame Directive (EC, 2000) and theMarine Strategy Directive (EC, 2008) directly refer to the ecologicalstatus. This is because it is not the pollutants per se we are inter-ested in, but their impact on the environment. Hajer (1995) hasstated that ‘‘pollution as such was not a problem; the real issue wasto guarantee a certain environmental quality’’. Besides, in cases ofmultiple pollution areas such as hot spot areas, the effect fromthe combination of many pollutants on ecosystem quality cannotbe assessed by measuring the concentration of each pollutant. Eco-logical indicators describe environmental or ecological qualitywith objectivity and connect this information with policy mea-sures. In spite of their suitability in decision support systems thereis a possible shortcoming: as they are the result of condensationand simplification of information from the marine environment,they may fail to express ecosystem complexity and heterogeneity.It is therefore necessary to use in parallel standard ecological





knowledge from community ecology for validation of the indica-tors. Ecological indicators are placed in the science-policy interfacementioned in a previous section. If scientists are responsible fordeveloping ecological indicators, policy makers are responsiblefor adopting them in legislation and management.

During marine water quality monitoring the application of sta-tistical procedures is sometimes preferable compared to the appli-cation of numerical models. This is due to the simplicity of thestatistical procedures compared to modeling and their direct appli-cation via specialized software of low cost leading to reliable re-sults. Furthermore, statistics can be used for determination ofthreshold values for harmful events in the marine environment(Stolte and Graneli, 2006), for scaling of variables expressing waterquality (Spatharis and Tsirtsis, 2010) and for site discriminationbased on selected variables (Primpas et al., 2008). On the otherhand, modeling requires higher expertise of the researchers andis more demanding process regarding computer capacity and avail-ability of datasets. In addition, modelers should prove that theircomputations are reliable and model results are realistic, beforetheir practical use from policy makers.

The degree of uncertainty in impact assessments in a waterquality monitoring program may be due to one or more of the fol-lowing reasons (Yearley, 1991; Beder, 2006): (a) Pragmatic uncer-tainty; researchers are sometimes asked by policy makers to cometo a conclusion without enough data or data with little relevance tothe problem. (b) Theoretical uncertainty; ecological science is char-acterized by qualitative information as well as by a high degree ofheterogeneity in space and time. As a consequence the bulk of datais never adequate and very often more than one interpretation canbe given for each question. In addition to disagreements on theinterpretation of data, disagreements on methodology is a usualsource of argument among scientists designing and running mon-itoring studies. Uncertainty about causal connections is also com-mon. (c) Complexity of the natural systems. Consequences ofecosystem complexity is the variability of ecological processesand the ‘‘interdeterminacy’’ that is a doubt about the correctnessof the scientific questions asked or poor understanding of the socialcontext in an activity that may have an impact on the environment(Deville and Harding, 1997). Scientific uncertainty is getting verysignificant if the relevance of science to policy decision is high. Itgives the opportunity to groups with vested interests to supporttheir views by highlighting particular points and interpreting themfor their own benefit.

The right of groups of interest to participate in shapingdecisions and policies establishes a mutual relation betweenmonitoring programs – data interpretation on one hand and datainterpretation – policy making on the other hand (Fig. 2). Duringthe Rio Conference in 1992, Agenda 21 emphasizes the need forpublic participation: ‘‘one of the fundamental prerequisites for theachievement of sustainable development is broad participation indecision making. Furthermore, in the more specific context of environ-ment and development, the need for new forms of participation hasemerged. This includes the need for individuals, groups and organiza-tions to participate in environmental impact assessment proceduresand to know about and participate in decisions, particularly thosewhich potentially affect the communities in which they live and work’’(Chapter 23.2). The Marine Strategy Framework Directive (EC,2008) encourages Member States to ‘‘ensure that all interested par-ties are given early and effective opportunities to participate in theimplementation of this Directive (Article 19). The term ‘‘interestedparties’’ includes existing Management Bodies, Scientific AdvisoryBodies and Regional Advisory Bodies. However, a barrier to theparticipatory process arises when stakeholders believe that theirsuggestions are not seriously considered, or they have not a mean-ingful impact on the policy making process. On the other hand,very often policy makers consider stakeholders’ opinion subjective


and unsystematic, thus problematic. The range of stakeholders var-ies depending on their interest, their perception about the state ofthe marine environment and their aspect about the requirementsof policy making. The level of stakeholder involvement should de-pend largely on the political and legal requirements for participa-tion. Stakeholder analysis provides the means to determine whoof the range of stakeholders, how and when should be involvedwithin the process of marine water quality policy making.

This interfacing between science and policy is sometimes un-clear not only for jurisdictional reasons mentioned before but alsobecause between the two lie the administrative and professionalestate as well as the public (Guston, 2000). The administrativeblock uses interdisciplinary information to produce rules and lawsusually under the oversight of the political estate and the pressurefrom the professional estate and the public. Policy making is influ-enced by two different forces (a) the scientific expertise and (b) thepublic: broad participation of the public based on democratic prin-ciples may induce criteria for good decision making and finding thecorrect solution. There is one more perspective less effective whichis the use of science in policy making as symbolic (Keller, 2009).

The close relationship between the collection, processing andanalysis of marine water quality data sets and the decision-makingprocess linked to policy making reveal the need for integrated ap-proaches. In this framework, use of integrated methods combiningeconomic, social, environmental and ecological priorities such asmultiple criteria analysis methods, can help all parties involvedto rationalize their views and come to a policy making relying onsound science. This process can be further supported by the devel-opment of dynamic user-friendly systems such as webGIS applica-tions (Kulawiak et al., 2010). In this framework, GIS mapping canbe updated periodically to reflect local changes of the study areaand explore the potential impact of alternative policy scenarioson the areas under study. As a final step, this information isdisseminated and shared among scientists, stakeholders and thepublic facilitating their further collaboration by ongoing feedbackand discussion.

References

Alvera-Azcarate, A., Ferreira, J.G., Nunes, J.P., 2003. Modeling eutrophication inmesotidal and macrotidal estuaries. The role of intertidal seaweeds. Estuar.Coast. Shelf S. 57, 715–724.

Alin, S.R., Feely, R.A., Dickson, A.G., Hernández-Ayón, J.M., Juranek, L.W., Ohman,M.D., Goericke, R., 2012. Robust empirical relationships for estimating thecarbonate system in the southern California Current System and application toCalCOFI hydrographic cruise data (2005–2011). J. Geophys. Res.–Oceans117(C5), doi:10.1029/2011JC007511.

Ananda, J., Herath, G., 2009. A critical review of multi-criteria decision makingmethods with special reference to forest management and planning. Ecol. Econ.68 (2535), 2538.

Arhonditsis, G., Tsirtsis, G., Angelidis, M.O., Karydis, M., 2000. Quantification of theeffects of nonpoint nutrient sources to coastal marine eutrophication:applications to a semi-enclosed gulf in the Mediterranean Sea. Ecol. Model.129, 209–227.

Backer, H., Leppanen, J.M., Brusendorff, A.C., et al., 2010. HELCOM Baltic Sea ActionPlan – A regional programme of measures for the marine environment based onthe Ecosystem Approach. Mar. Pollut. Bull. 60, 642–649.

Beanlands, G.E., Duinker, P.N., 1983. An ecological framework for environmentalimpact assessment in Canada. Halifax, Nova Scotia: Dalhousie UniversityInstitute for Resource and Environmental Studies and Federal EnvironmentalAssessment Review Office.

Beder, S., 2006. Environmental Principles and Policies. Earthscan, London.Beninger, P.G., Boldina, I., Katsanevakis, S., 2012. Stregthening statistical usage in

marine ecology. J. Exper. Mar. Biol. Ecol. 426, 97–108.Bergström, U., Sundblad, G., Downie, A.-L., Snickars, M., Boström, C., Lindegarth, M.,

2013. Evaluating eutrophication management scenarios in the Baltic Sea usingspecies distribution modelling. J. Appl. Ecol. 50, 680–690.

Boesch, D., Hecky, R., O’Melia, C., Schindler, D., Seitzinger, S., 2000. Eutrophication ofSwedish Waters. Final Report, Stockholm.

Buikema Jr., A.L., Niederlehner, B.R., Cairns Jr., J., 1982. Biological Monitoring. PartIV: Toxicity testing. Water Res. 16, 239–262.

Cairns Jr., J., 1981. Biological Monitoring: future needs. Water Res. 15, 941–952.


http://refhub.elsevier.com/S0025-326X(13)00548-1/h0005

























Cefas, 2010. Microbiological Monitoring of Bivalve Mollusc Harvesting Areas. Guideto Good Practice: Technical Application. EU Reference Laboratory forMonitoring Bacteriological and Viral Contamination of Bivalve Molluscs.

CEP, 2010. Marine Coastal Issues. Caribbean Environment Programme, <http://www.cep.unep.org> (retrieved).

Cerco, C.F., Cole, T., 1993. Three-dimensional eutrophication model of ChesapeakeBay. J. Environ. Eng. 119 (6), 1006–1025.

Cerco, C.F., Cole, T., 1995. User’s guide to the CE-QUAL-ICM. Technical Report EL-95-1. U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.

Chapman, D., 1996. Water Quality Assessments, second ed. E&EN Spon, London.Clean Water Fund, Minerals Policy Centre, OMB Watch, UK Public Research Interest

Group and Working Group on Community Right to Know, 2001. Ignorance istoxic bliss: the secret war on our right-to-know, Terra Knowledge Network.

Cloern, J.E., 2001. Our evolving conceptual model of the coastal eutrophicationproblem. Mar. Ecol. Prog. Ser. 210, 223–254.

COBSEA, 2008. New Strategic Direction for Cobsea (2008–2012). 23pp.Coull, B.C., 1985. The use of long-term biological data to generate testable

hypotheses. Estuaries 8, 84–92.Crossland, C.J., Kremer, H.H., Lindeboom, H.J., Crossland, J.I.M., Le Tissier, M.D.A.,

2005. Coastal Fluxes in the Anthropocene. Springer, Berlin.Danovaro, R., 2003. Pollution threats in the Meditarranean Sea: an overview. Chem.

Ecol. 19, 15–32.de Jong, F., 2006. Marine Eutrophication in Perspective. On the Relevance of Ecology

for Environmental Policy. Springer, Berlin.de Jonge, V.N., Elliot, M., Brauser, V.S., 2006. Marine Monitoring: its shortcomings

and mismatch with the EU Water Framework Directive’s objectives. Mar. Pollut.Bull. 53, 5–19.

Deville, A., Harding, R., 1997. Applying the Precautionary Principle. Federation Press,Sydney.

DiMento, J.F.C., Hickman, A.J., 2012. Environmental Governance of the Great Seas:Law and Effect. Edward Elgar, Cheltenham, UK.

Douvere, F., Ehler, C.N., 2011. The importance of monitoring and evaluation inadaptive maritime spatial planning. J. Coast. Conserv. 15, 305–311.

Downie, A.J., Baxter, J.M., 2004. The Water Framework Directive: driving thedevelopment of a co-ordinated marine monitoring system. AquaticConservation: Mar. Freshwater Ecosyst. 14, S69–S79.

Duarte, C.M., 2009. Coastal eutrophication research: a new awareness.Hydrobiologia 629, 263–269.

EC, 2000. Directive of the European Parliament and the Council 2000/60/ECestablishing a framework for community action in the field of Water Policy.Official J. Europ. Commun., Bruss. L327, 1–72.

EC, 2003. Directive 2003/4/EC of the European Parliament and of the Council of 28January 2003 on Public Access to Environmental Information, EuropeanCommission, Luxembourg.

EC, 2006. Directive 2006/7/EC of the European Parliament and the Council of 15February 2006 concerning the management of bathing water quality andrepealing Directive 76/160/EEC. Official J. Europ. Commun. L64, 37–51.

EC, 2008. Directive 2008/56/EC of the European Parliament and of the Council of 17June 2008 on Establishing a Framework for Community Action in the Field ofMarine Environmental Policy (Marine Strategy Framework Directive). L164/19.

EEA, 1999. State and Pressures of the Marine and Coastal MediterraneanEnvironment. European Environment Agency, Copenhagen.

EEC, 1979. Directive 79/923/EEC of the European Parliament and the Council on theQuality Required of Shellfish Waters, vol. L377, pp. 1–9.

EEC, 1991. Dir Council Directive concerning Urban Waste Water Treatment (91/271/EEC). Official J. Europ. Commun. L135, 40–52.

EPA, 2000. Guidance for choosing a sampling design for environmental datacollection. Office of Environmental Information, US Environmental ProtectionAgency.

Erickson, W.P., Strickland, M.D., 1995. Survey design, statistical analysis and basisfor statistical inferences in coastal habitat injury assessment: exxon Valdez oilspill. In: Wells, P.G., Butler, J.N., Hughes, J.S. (Eds.), Exxon Valdez Oil Spill: Fateand effects in Alaskan Waters. ASTM Publication, Philadelphia.

Fairweather, P.G., 1989. Statistical power and design requirements forenvironmental monitoring. Austr. J. Mar. Freshwater Res. 42, 555–567.

FAO, 2006. Water monitoring. Mapping existing Global Systems & Initiatives.Report prepared by FAO on behalf of the UN-Water Task Force on Monitoring.Stockholm (21.08.06).

Fernandes, R.N., Valavanis, V.D., 2008. A GIS-based tool for storage, selection andvisualization of time series 4D marine datasets. Hydrobiologia 612, 297–300.

Ferreira, J.G., Andersen, J.H., Borja, A., et al., 2011. Overview of eutrophicationindicators to assess environmental status within the European Marine StrategyFramework Directive. Estuar. Coast. Shelf S. 93, 117–131.

French, D.P., 1998. Estimates of Injuries to Marine Communities Resulting from theNorth Cape oil spill, based on Modeling of Fates and Effects. Report to NOOADamage Assessment Center, Silver Spring, MD.

Gallucci, V.F., Saila, S.B., Gustafson, D.J., Rothschild, B.J., 1996. Stock Assessment:quantitative methods and applications for small-scale fisheries. LewisPublichers, CRC Press, Boca Raton, 527 pp.

Green, R.H., 1989. Power analysis and practical strategies for environmentalmonitoring. Environ. Res. 50, 195–205.

Grégoire, M., Raick, C., Soetaert, K., 2008. Numerical modeling of the central BlackSea ecosystem functioning during the eutrophication phase. Prog. Oceanogr. 76,286–333.

Guston, D.H., 2000. Between Politics and Science: Assuring the Integrity andProductivity of Research. Cambridge University Press, Cambridge.


Haas, P.M., 1990. Saving the Mediterranean. The Politics of InternationalEnvironmental Cooperation. Columbia University Press, NY, 303 pp.

Hajer, M.A., 1995. The Politics of Environmental Discourse: EcologicalModernization and the Policy Process. Oxford University Press, Oxford.

Hajkowicz, S., 2007. A comparison of multiple criteria analysis and unlaidapproaches to environmental decision making. Environ. Sci. Policy 10, 177–184.

Håkanson, L., Bryhn, A.C., 2008. Eutrophication in the Baltic Sea. EnvironmentalScience and Engineering. Springer-Verlag, Berlin.

Heileman, S., 2009. I-2 Guinea Current: LME #28’. Large Marine Ecosystems of theWorld. NOOA. <http://www.lme.nooa.gov/> (accessed).

HELCOM, 2008. Convention on the Protection of the Marine Environment of theBaltic Area. Helsinki Convention, Helsinki, Finland.

Holland, A.F., Shaughnessy, A.T., Hiegel, M.H., Scott, L.C., 1986. Long term benthicmonitoring for the Maryland portion of Chesapeake Bay: July 1984–December1985, vol. 1. Prepared for Maryland Departments of Health and Mental Hygieneand of Natural Resources by Martin Marietta Environmental Systems, Columbia,Maryland.

Hooper, F.F., 1969. Eutrophication indices and their relation to other indices ofecosystem change. In: Eutrophication: Causes, Consequences, Correctives.National Academy of Sciences, Washington, DC, pp. P225–P235.

Ignatiades, L., Vounatsou, P., Karydis, M., 1992. A possible method for evaluatingoligotrophy and eutrophication based on nutrient concentration scales. Mar.Pollut. Bull. 24 (5), 238–243.

Janssen, R., 1992. Multiobjective Decision Support for Environmental Management.Kluwer, Amsterdam.

Jeftic, L., 1993. Long-term programme for pollution monitoring and research in theMediterranean (MED POL). Water Sci. Tech. 27 (7–8), 345–352.

Johnson, D., 2008. Environmental Indicators: their utility in meeting the OSPARConvention’s regulatory needs. ICES J. Mar. Sci. 65, 1387–1391.

Karagiannakos, A., 1995. Fisheries Management in the European Union. AshgatePubl. Co., Vermont, 286 pp.

Karydis, M., 2009. Eutrophication assessment of coastal waters based on indicators:a literature review. GNEST J. 11 (4), 373–390.

Karydis, M., Tsirtsis, G., 1996. Ecological Indices: a biometric approach for assessingeutrophication levels in the marine environment. Sci. Tot. Environ. 186, 209–219.

Karydis, M., Kitsiou, D., 2012. Eutrophication and environmental policy in theMediterranean Sea: a review. Environ. Monit. Assess. 184, 4931–4984.

Katsanevakis, S. et al., 2011. Ecosystem-based marine spatial management: Reviewof concepts, policies, tools, and critical issues. Ocean Coast. Manage. 54, 807–820.

Keller, A.N., 2009. Science in Environmental Policy: The Politics of Objective Advice.The MIT Press, Cambridge, Massachusetts.

Kingston, P.F., 2002. Long-term environmental impact of oil spills. Spill Sci. Techn.Bull. 7 (1–2), 53–61.

Kitsiou, D., Tsirtsis, G., Karydis, M., 2001. Developing an optimal sampling design. Acase study in a coastal marine ecosystem. Environ. Monit. Assess. 71, 1–12.

Kitsiou, D., Coccossis, H., Karydis, M., 2002. Multi-dimensional evaluation andranking of coastal areas using GIS and multiple criteria choice methods. Sci. Tot.Environ. 284, 1–17.

Kitsiou, D., Karydis, M., 2011. Coastal marine eutrophication assessment: a reviewon data analysis. Environ. Inter. 37, 778–801.

Kulawiak, M., Prospathopoulos, A., Perivoliotis, L., Łuba, M., Kioroglou, S.,Stepnowski, A., 2010. Interactive visualization of marine pollutionmonitoring and forecasting data via a Web-based GIS. Comput. Geosci. 36,1069–1080.

Latimer, J.S., Rego, S.A., 2010. Empirical relationship between eelgrass extent andpredicted watershed-derived nitrogen loading for shallow New Englandestuaries. Estuar. Coast. Shelf S. 90, 231–240.

Le Gall, A.C., Hydes, D.J., Kelly-Gerreyn, B.A., Slinn, D.J., 2000. Development of a 2Dhorizontal biogeochemical model for the Irish Sea DYMONIS. ICES J. Mar. Sci. 57,1050–1059.

Legendre, P., Trousselier, M., Jarry, V., Fortin, M.-J., 1989. Design of simultaneoussampling of ecological variables: from concepts to numerical solutions. Oikos55, 30–42.

Legendre, P., Legendre, L., 2003. Numerical Ecology, second ed. Elsevier,Amsterdam.

Likens, G.E., 1983. A priority of ecological research. Bull. Ecol. Soc. Am. 64, 234–243.Livingston, R.J., 1987. Field sampling in estuaries: the relationship of scale to

variability. Estuaries 10, 194–207.MacGarvin, M., 1995. The implications of the precautionary principle for biological

monitoring. Helgolander Meeresunters 49, 647–662.Magurran, A.E., 2004. Measuring Biological Diversity. Blackwell Science Ltd.,

London.Mapstone, B.D., 1995. Scalable decision rules for environmental impact studies;

effect size, Type I and Type II errors. Ecol. Appl. 5 (2), 401–410.Mason, R.L., Gunst, R.F., Hess, J.L., 1989. Statistical Design and Analysis of

Experiments. J. Wiley and Sons, New York.Mateus, M., 2012. A process-oriented model of pelagic biogeochemistry for marine

systems. Part I: Model description. J. Mar. Syst. 94 (2012), S78–S89.Matthews, R.A., Buikema, A.L., Cairns Jr., J., Rodgers Jr., J.H., 1982. Biological

monitoring. Part II: Receiving system functional methods, relationships andindices. Water Res. 16, 129–139.

McDonald, L.L., Erickson, W.P., Stricklan, M.D., 1995. Survey Design, StatisticalAnalysis and Basis for Statistical Inferences in Coastal Habitat InjuryAssessment: Exxon Valdez Oil Spill. In: Wells, P.G., Butler, J.N., Hughes, J.S.


http://www.cep.unep.org

http://www.cep.unep.org



































































http://www.lme.nooa.gov/









































































(Eds.), Exxon Valdez Oil Spill: Fata and Effects in Alaskan Waters. AmericanSociety for Testing and Materials (ASTM STP 1219), Philadelphia.

Millard, S.P., Lettenmaier, D.P., 1986. Optimal design of biological samplingprograms using analysis of variance. Estuar. Coast. Shelf S. 22, 637–656.

Milliken, G.A., Johnson, D.E., 1997. Analysis of messy data. Volume I: DesignedExperiments. Chapman & Hall, London.

MIT, 1970. Man’s impact on the global environment. Report of the study of CriticalEnvironment Problems (SCEP): assessments and recommendations for action.Massachusetts Institute of Technology, Cambridge, Mass. and London: MITPress.

Moriki, A., Karydis, M., 1994a. Application of multi-criteria choice methods inassessing eutrophication. Environ. Monit. Assess. 33, 1–18.

Moriki, A., Karydis, M., 1994b. Multicriteria evaluation in coastal management. J.Coast. Manage. 12 (1), 171–178.

Nijkamp, P., Voogd, H., 1986. A survey of qualitative multiple criteria choicemethods. In: Nijkamp, P., Leitner, H., Wringley, N. (Eds.), Measuring theUnmeasurable. Kluwer Nijhoff, Dordrtecht, pp. 425–447.

Nixon, S.W., 1995. Coastal marine eutrophication: a definition, social causes andfuture concerns. Ophelia 41, 199–219.

NRC, 1986. Ecological Knowledge and Environmental Problem Solving: Conceptsand Case Studies. National Academy of Sciences, National Academy Press,Washington, DC.

NRC, 1990. Managing Troubled Waters. The Role of Marine EnvironmentalMonitoring. National Academy of Sciences. National Academy Press,Washington, DC.

NRC, 2000. Clean Coastal Waters: Understanding and Reducing the Effects ofNutrient Pollution. National Academy of Sciences, National Academy Press,Washington, DC.

NRC, 2003. Oil in the Sea III: Inputs, Fates and Effects. National Academy of Sciences,National Academy Press, Washington, DC.

OECD, 1996. Recommendation of the Council on Implementing Pollutant Releaseand Transfer Registers. Organization of Economic Co-operation andDevelopment, Paris, pp. 21.

OSPAR, 2007. Convention of the Protection of the Marine Environment of the North-East Atlantic. The OSPAR Commission, London.

OSPAR, 2009. Eutrophication status of the OSPAR Maritime Area. Second OSPARIntegrated Report. The OSPAR Commission. Eutrophication Series. London.

Painting, S.J., Devlin, M.J., Rogers, S.I., Mills, D.K., Parker, E.R., Rees, H.L., 2005.Assessing the suitability of OSPAR EcoQOs for eutrophication vs ICES criteria forEngland and Wales. Mar. Pollut. Bull. 50, 1569–1584.

Parma, A., 1998. What can adaptive management do for our fish, forests, food andbiodiverisity? Integr. Biol. 1, 16–26.

Price, B., 1987. Environmental monitoring statistics and the law: room forimprovement. Am. Stat. 41 (4), 257–259.

Primpas, I., Karydis, M., Tsirtsis, G., 2008. Assessment of clustering algorithms indiscriminating eutrophic levels in coastal waters. GNEST J. 10, 359–365.

Reed, M., Gundlach, E., Kanay, T., 1989. A coastal zone oil spill model: developmentand sensitivity studies. Oil Chem. Pollut. 5 (6), 411–449.

Scheiner, S.M., Gurevitch, J. (Eds.), 1993. Design and Analysis of EcologicalExperiments. Chapman and Hall, New York.

Segar, D.A., Stamman, E., 1986. Fundamentals of marine pollution monitoringprogram design. Mar. Pollut. Bull. 17 (5), 194–200.

Spatharis, S., Tsirtsis, G., 2010. Ecological quality scales based on phytoplankton forthe implementation of Water Framework Directive in the EasternMediterranean. Ecol. Indicat. 10, 840–847.


Stagg, R.M., 1998. The development of an international programme for monitoringthe biological effects of contaminants in the OSPAR Convention Area. Mar.Environ. Res. 46 (1–5), 307–313.

Stolte, W., Graneli, E., 2006. Threshold Nutrient Levels for Harmful Algal Events in someEuropean Coastal Waters. THRESHOLDS: Thresholds of Environmental Sustainability,D3.1.2, Sixth Framework Programme, <http://www.thresholds-eu.org/>.

Tett, P., Gilpin, L., Svedensen, H., et al., 2003. Eutrophication and some Europeanwaters of restricted exchange. Cont. Shelf Res. 23, 1635–1671.

Tromp, D., Wieriks, K., 1994. The OSPAR Convention: 25 years of North SeaProtection. Mar. Pollut. Bull. 29, 622–626.

Turnhout, E., Hisschemoller, M., Eijsackers, H., 2007. Ecological indicators: betweenthe two fires of science and policy. Ecol. Indicat. 7, 215–228.

Underwood, A.J., 2007. Experiments in Ecology. Their Logical Design and Interpolationusing Analysis of Variance. Cambridge University Press, Cambridge.

UNEP, 1997. Monitoring for Compliance Control. Athens: UNEP(OCA)/MED WG.127/Inf. 4.UNEP, 1999. Review of Activities Carried out as Part of MED POL Phase III.

UNEP(OCA)/MED WG.161/Inf. 3, Athens.UNEP, 2002. African Environmental Outlook – Past, Present and Future Perspectives.

Web Publication: Grid-Arendal. <http://www.grida.no/publications/other/aeo/?src=/aeo/>.

UNEP, 2003. Eutrophication Monitoring and Strategy of MED POL. Athens:UNEP(DEC)/MED WG.231/14.

UNEP, 2008. New Strategic Direction for COBSEA (2008–2012). COBSEA Secreteriat,United Nations Environment Programme.

UNEP, 2010. West and Central African Regional Seas Programme, <http://www.unep.org/regionalseas/wacaf> (retrieved).

UNEP/MAP, 2004. Report of the Review Meeting of MED POL-Phase III MonitoringActivities. UNEP(DEC)/MED WG.243/4, Athens.

UNEP/MAP, 2005. Convention for the Protection of the Marine Environment and theCoastal Region of the Mediterranean and its protocols. United NationsEnvironment Programme/Mediterranean Action Plan, Athens.

UNEP/MAP, 2011. Development of Assessment Criteria for Hazardous Substances inthe Mediterranean. UNEP(DEPI)/MED WG. 365/Inf. 8, Athens.

Van Hoey, G., Borja, A., Birchenough, S., et al., 2010. The use of benthic indicators inEurope: from the water framework directive to the marine strategy frameworkdirective. Mar. Pollut. Bull. 60 (12), 2187–2196.

Vollenweider, R.A., Marchetti, R., Viviani, R. (Eds.), 1992. Marine CoastalEutrophication. Elsevier, London.

Walters, C.J., 1986. Adaptive Management of Renewable Resources. Macmillan, NewYork.

Wolfe, D.A., Champ, M.A., Flemer, D.A., Mearns, A.J., 1987. Long-term biological datasets: their role in research, monitoring and management of estuarine andcoastal marine systems. Estuaries 10 (3), 181–193.

Yearley, S., 1991. The Green Case: A Sociology of Environmental Issues, Argumentsand Politics. Harper-Collins, London.

Zaldivar, J.M., Cardoso, A.C., Viaroli, P., et al., 2008. Eutrophication in transitionalwaters: an overview. Transit. Waters Monit. 1, 1–78.

Zaldivar, J.M., Bacelar, F.S., Dueri, S., Marinov, D., Viaroli, P., Hernandez-Garcia, E.,2009. Modeling approach to regime shifts of primary production in shallowcoastal ecosystems. Ecol. Model. 220, 3100–3110.

Zheng, X., Pierce, G.J., Reid, D.G., Jolliffe, I.T., 2002. Does the North Atlantic currentaffect spatial distribution of whiting? Testing environmental hypotheses usingstatistical and GIS techniques. Ices J. Mar. Sci. 59, 239–253.

Zuur, A.F., Ieno, E.N., Smith, G.M., 2007. Analyzing Ecological Data. Springer, NewYork.




















































http://www.thresholds-eu.org/









http://www.grida.no/publications/other/aeo/?src=/aeo/

http://www.grida.no/publications/other/aeo/?src=/aeo/

http://www.unep.org/regionalseas/wacaf

http://www.unep.org/regionalseas/wacaf

























Date post:	15-Dec-2016
Category:	Documents
Upload:	dimitra
View:	218 times
Download:	1 times

Marine water quality monitoring: A review

Documents