ICES Journal of Marine Science: Journal du Conseil Advance Access originally published online on August 1, 2008
ICES Journal of Marine Science: Journal du Conseil 2008 65(8):1442-1448; doi:10.1093/icesjms/fsn123
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
This article appears in the following ICES Journal of Marine Science issue: Marine Environmental Indicators: Utility in Meeting Regulatory Needs [View the issue table of contents]
Applicability of the trophic index TRIX in two transitional ecosystems: the Mar Menor lagoon (Spain) and the Mondego estuary (Portugal)
1 Department of Ecology and Hydrology, Faculty of Biology, University of Murcia, 30100 Murcia, Spain
2 IMAR, Institute of Marine Research, c/o Department of Zoology, Faculty of Sciences and Technology, University of Coimbra, 3004-517 Coimbra, Portugal
Correspondence to F. Salas: tel: +34 968364326; fax: +34 968363963; e-mail: fuenmar{at}um.es
Salas, F., Teixeira, H., Marcos, C., Marques, J. C., and Pérez-Ruzafa, A. 2008. Applicability of the trophic index TRIX in two transitional ecosystems: the Mar Menor lagoon (Spain) and the Mondego estuary (Portugal). – ICES Journal of Marine Science, 65: 1442–1448.Ecological indicators are commonly used to provide synoptic information about the state of ecosystems. Their main attribute is that they combine a range of environmental factors in a single value, which is thought useful for management and for making ecological-quality concepts easily understandable by the general public. In some European countries, the trophic state of marine and coastal waters is characterized using the trophic index TRIX. We used TRIX in the studies of the Mar Menor coastal lagoon (southeast Spain) and the Mondego estuary (northwest Portugal). These areas are under environmental stress, and eutrophication processes have been monitored in recent years. The results demonstrate the inefficacy of TRIX as a trustworthy tool to classify eutrophication status in estuarine waters. The causes of the malfunction of the index are discussed, and we suggest that, if the index is to be applied over wide areas, the classification criteria will have to be adapted to specific environments.
Keywords: ecological indicators, eutrophication, trophic index
Received 23 November 2007; accepted 15 May 2008; advance access publication 1 August 2008.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
For more than 30 years, nutrient enrichment has been one of the leading threats to the health of coastal ecosystems and resources (Nixon, 1995; Elmgren, 2001), and eutrophication has become a major concern in many parts of the European seas. The Urban Wastewater Treatment Directive (UWTD; EC, 1991a) defines eutrophication as the "enrichment of water by nutrients, especially compounds of nitrogen and/or phosphorus, causing an accelerated growth of algae and higher forms of plant life to produce an undesirable disturbance to the balance of organisms present in the water to the quality of the water concerned", so interpreting eutrophication as a process rather than a state. To reduce nutrient input and stop eutrophication, measures have been taken through regional, national, and supranational initiatives. The UWTD and the Nitrate Directive (EC, 1991b) stipulate that EU Member States must indicate sensitive areas and vulnerable zones for water bodies within their jurisdiction, based on criteria that focus on eutrophication. The European Environment Agency (EEA, 1999a, b, 2001) recognizes that coastal lagoons, as systems of restricted exchange, are particularly vulnerable to eutrophication, and suggests that this may apply to the estuaries and lagoons of the Iberian Peninsula.
The Water Framework Directive (WFD; EC, 2000) establishes a framework for the protection of all waters (including inland surface waters, transitional waters, coastal waters, and groundwaters) and is aimed at achieving a good-quality status in all waters by 2015. The concept of ecological status of the various waters is defined by the WFD in terms of the quality of the biological community present, as well as the quality of its hydrological and chemical characteristics. Its application requires methods that can distinguish different levels of ecological quality to classify surface waters.
In some Member States, the trophic state of marine and coastal waters is being characterized through the trophic index TRIX that was proposed by Vollenweider et al. (1998). TRIX integrates Chl a, oxygen saturation, dissolved inorganic nitrogen, and dissolved inorganic phosphorus and is scaled from 0 to 10, covering a range of four trophic states (high, good, moderate, and degraded). We test the performance and robustness of TRIX for its efficiency in describing the ecological status of two coastal systems of the Iberian Peninsula, the Mondego estuary in Portugal and the Mar Menor lagoon in Spain (Figure 1).
|
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Study areas
Mar Menor lagoon (mean depth 3.6 m; surface area 135 km2) is located in southeast Spain (Figure 1b), a semi-arid region of the western Mediterranean with an average annual rainfall <300 mm and a potential evapo-transpiration close to 900 mm (López-Bermúdez et al., 1981). The lagoon is connected with the open sea at some points by channels through which water exchange takes place. The diverse human activities along its coast and the associated sources of pollution result in heterogeneity in environmental quality within the lagoon.
Until recently, there were no watercourses flowing permanently into the lagoon. Most of the more than 20 watercourses present in the watershed discharge into the southern basin of the lagoon, but their discharge is determined by a sporadic and torrential rainfall regime. The Albujon watercourse, the main collector in the drainage basin, is currently an exception, because it maintains a regular flux of water resulting from changes in agricultural practices and related phreatic rising (Pérez-Ruzafa and Aragon, 2002). During the 1990s, agriculture started to change from dry-crop farming to production systems using intensive irrigation by water diverted from the Tajo River to the Segura. Consequently, run-off from agricultural lands is causing nutrient enrichment: nitrate concentrations have increased from values <1 µmol l–1 throughout the year to occasional values of 8 µmol l–l during the period 1988–1997 (Pérez-Ruzafa et al., 2005).
Mar Menor lagoon is included on the Ramsar List of Wetlands of International Importance. It is also a Special Protected Area of Mediterranean Interest, a Special Protected Area under the EU Wild Birds Directive, and a Site of Community Importance to be integrated in the Natura 2000 Network (EU Habitats Directive). The lagoon was declared a sensitive area subject to eutrophication in June 2001 (EU Directives 91/271/EEC and 91/676/EEC).
Mondego estuary, a warm-temperate intertidal system located on the Atlantic coast of Portugal (40°08'N 8°50'W), is also included on the Ramsar List of Wetlands of International Importance. Over a stretch of 7 km close to the mouth, the estuary is divided into two channels separated by a river island (Figure 1a). The northern channel is deeper (4–10 m high tide) and constitutes the main navigation route to a commercial harbour. The southern channel, representing the study area, is shallower (2–4 m high tide) and has become almost completely silted up in upstream areas, behaving largely as a "bag", with water circulation predominantly dependent on tides and only a small input of fresh water from a tributary (Pranto River), controlled by a sluice (Marques et al., 2003).
A mixture of inputs from sewage effluents and agricultural run-off, as well as from mariculture activity, contributes to the nutrient loads entering the Mondego estuary (Martins et al., 2001; Lillebø et al., 2005). From 1998 to 2006, several mitigation measures have been taken to ameliorate the condition of the southern channel. Specifically, water circulation has been enhanced through an experimental re-establishment of the upstream connection between the two channels (allowing water to flow from the northern channel into the southern channel at high tide), and by reducing the discharges from the Pranto River sluice and diverting the nutrient-enriched fresh water to the northern channel by a sluice located upstream.
Sampling and primary analysis
Mar Menor lagoon
The lagoon was divided into five zones (Z1–Z5) reflecting the relative influences of terrestrial activities and the marine environment, and a grid of 20 sampling stations was spread evenly over these zones, so that each zone was represented by four replicates (Figure 1b): Z1 is most influenced by touristic and urban impacts; Z2 is close to the main watercourses discharging agricultural wastewater; Z3 is located in the central basin and is less influenced by anthropogenic activities; Z4 is close to the El Estacio channel that connects the lagoon to the Mediterranean Sea; and Z5 is least affected by marine waters.
The stations were sampled monthly, February–December 1997, May 2002–May 2003, and February–December 2006. Water samples were taken at an approximate depth of 1 m with a Niskin bottle or by pumping. Samples for nutrient analysis were immediately stored in the dark at 4°C. Before 2006, nitrate (N–NO3–), nitrite (N–NO2–), ammonia (N–NH4+), and phosphate (P–PO43–) were determined following the methods described by Parsons et al. (1984), and thereafter with a multiparametric analyser (MICROMAC 1000 C model). Salinity was determined with a Beckman RS 7B salinometre in 1997, and in situ with a WTW Multiline F/Set3 multiple probe in 2002/2003. In 2006/2007, salinity and dissolved oxygen were measured with an YSI 6660 sonde. Before 2006, Chl a was determined with the spectrophotometric methods reported by Parsons et al. (1984), and thereafter with the YSI 6660 sonde. Although the methods to measure the various parameters have changed over time, they should produce comparable results.
Mondego estuary
Five subtidal sampling stations (SA1–SA5) in the southern channel (Figure 1a) were surveyed monthly, January 2003–November 2006 (except for June, August, and October 2004). During high tide, salinity and dissolved oxygen (%, mg l–1) were measured in situ in surface waters, and water samples were collected for analyses of phytoplankton biomass and dissolved nutrients. Transparency was measured by a Secchi disk (depth in metres). Concentrations of dissolved nutrients (ammonia, nitrate, nitrite, and phosphorus) were measured in the laboratory following the standard procedures of APHA (1995) and Strickland and Parsons (1972). Phytoplankton biomass was calculated from the concentration of Chl a (Strickland and Parsons, 1972).
Two zones were distinguished, a downstream zone (ZD including stations SA1–SA3) and an upstream zone (ZU including stations SA4 and SA5). These zones differ in the extent to which they demonstrate symptoms of eutrophication (ZD being less affected than ZU; Rodrigues, 2004).
Data analysis
TRIX values were calculated for each sample according to the algorithm of Vollenweider et al. (1998), then averaged by zone. According to Giovanardi and Vollenweider (2004) and Penna et al. (2004), values ranging from 0 to 4 correspond to high-quality, 4–5 to good, 5–6 to moderate, and 6–10 to degraded conditions.
TRIX results were compared with the EEA classification in four categories for the annual means of nitrogen (nitrate + nitrite) and phosphorus (phosphate) separately, based on their relative concentrations (µmol l–1), as proposed by Crouzet et al. (1999): good, fair, poor, and bad (Table 1).
|
Mean TRIX, nutrient concentrations, and Chl a were subjected to a three-factor ANOVA considering year (Mar Menor: 1997, 2002/2003, 2006; Mondego: 2003–2006), season (winter, spring, summer, and autumn), and zone (Mar Menor: Z1–Z5; Mondego: ZD and ZU). To understand what could be driving the TRIX value, relationships between the index and parameters measured in each case study were tested using the Pearson correlation coefficients.
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Mar Menor
The TRIX values obtained were always lower than 4, indicating a high-quality condition (Figure 2a), and hardly varied among zones except during the latter part of 2002 and in 2003. Maximum values were obtained in spring 1997, summer 2002, and winter/spring 2006. Broadly speaking, TRIX appears to have declined over the entire period, suggesting improving conditions.
|
Figure 2b–e shows the time-series for Chl a, phosphorus, nitrate, and nitrite. Chl a values were <0.6 µg l–1 during most of 2006, indicating oligotrophic conditions, after reaching a maximum of 1.9 µg l–1 in May 2002 (Figure 2b). Trends appeared to be quite similar across zones. Maximum nitrate concentrations were observed in winter/spring 2006 (indicating bad conditions on the EEA scale), but thereafter concentrations were reduced to virtually zero (good). Nitrite concentrations remained almost consistently <0.2 µmol l–1 and did not contribute much to total dissolved N. Phosphate concentrations reached values higher than 1.1 µmol l–1 (bad) at all stations during February–April 1997, but at other times, demonstrated mostly good conditions according to the EEA scale. TRIX values, as well as all nutrient concentrations, revealed significant year*season interactions (Table 2), whereas the only other term contributing significantly to the variance was zone for nitrite.
|
Nitrate, nitrite, phosphate, and Chl a concentrations all demonstrated significant positive relationships with TRIX (Table 3). There were also significant positive correlations between phosphate and salinity, phosphate and nitrite, Chl a and nitrite, and a significant negative correlation between nitrate and salinity.
|
Mondego estuary
The TRIX results (Figure 3a) indicate that the eutrophication status in the south channel varied between good (minimum value 4.3) and degraded (maximum value 6.8). The values were generally higher in the upstream zone (ZU). Overall, the south arm of the estuary could be characterized as having a moderate to degraded water quality (83% of the samples), typical for moderate to high productive systems (Giovanardi and Vollenweider, 2004).
|
The average Chl a concentrations over the entire period were 3.3 and 4.7 µg l–1 for downstream and upstream stations, respectively. Records above those mean concentrations (maxima of 7.8 and 17.9 µg l–1, respectively) occurred only during the growing season (March–September; Figure 3b). In both zones, the nitrate + nitrite concentrations combined remained for most of the study period within the limits considered good (<6.5 µmol l–1) according to EEA classification (Figure 3d and e), and this was also true for phosphate (<0.5 µmol l–1). The ANOVA revealed that zones explained a significant part of the variance in TRIX values, Chl a, and all nutrients, whereas year and season also contributed significantly to the variance of all nutrients except phosphate. Season was a significant factor for Chl a as well, and the interaction term year*season was significant for phosphate (Table 2).
The oxygen concentration (Figure 3h) remained virtually always within the high-quality standard set by the WFD at 5.7 and 7 mg l–1 for marine waters and fresh water, respectively, which represent the limits above which most biological requirements are satisfied (Best et al., 2007). The mean concentrations found were 9.2 and 8.5 mg l–1 in ZD and ZU, respectively. However, some records in 2003 corresponded to conditions of oxygen deficiency (<6 mg l–1) according to OSPAR (2005).
TRIX results were positively correlated with nutrients in the system (nitrate: r = 0.4; nitrite: r = 0.4; ammonia: r = 0.5; phosphate: r = 0.6; all p < 0.01) and negatively with transparency (r = –0.5; p < 0.01), but the index did not demonstrate any significant relation with Chl a (Table 3).
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
In comparing nutrient concentrations in Mar Menor over a 9-year period (1988–1997), Pérez-Ruzafa et al. (2002) found a considerable increase in nitrate concentration and a reduction in phosphate concentration, which they related to the increase in irrigated lands, in the amount of agricultural fertilizers used, and in wastewater treatment plants. However, the TRIX values obtained here would rather indicate a situation where eutrophication is not a concern. This result is apparently the result of the low Chl a values, particularly in 2006. Several studies (Pérez-Ruzafa et al., 2002, 2004) have demonstrated that top–down control exerted by fish larvae and large gelatinous zooplankton in the lagoon is responsible for these low Chl a concentrations.
In Mondego estuary, Chl a did not constitute a key factor, and the TRIX was driven by the nutrients. The index did reveal that the system is under stress and also identified its upstream zone as the area most affected by eutrophication. In contrast, the EEA classification based on nutrient values would indicate a good condition. Moreover, Chl a values were below the values proposed as reference conditions for Atlantic coastal waters in the UK and Republic of Ireland (10 µg l–1 offshore and 15 µg l–1 nearshore, respectively; Devlin et al., 2007), although in general, the primary production in an estuary may be expected to be higher than in offshore waters. This should be even more so for this particular system, because it is under the influence of coastal upwelling of nutrient-rich waters (Loureiro et al., 2006). The degraded condition, according to TRIX, is also contrary to signs of improvement in the system’s ecological integrity, as evidenced by a decline in the occurrence of green macroalgae blooms and the recovery of Zostera noltii beds, as well as by an increase in secondary production (Dolbeth et al., 2007; Lillebø et al., 2007).
The multimetric TRIX offers the advantage of combining routinely collected measurements of several environmental variables (Giovanardi and Vollenweider, 2004), and also provides information that is readily understandable by, and meaningful to, various stakeholders. However, our results lead us to suggest that its interpretation for the transitional waters investigated is not straightforward. A specific drawback may be that TRIX is derived from the combination of four parameters that are not totally independent and, therefore, might be overemphasizing particular features.
Coelho et al. (2007), based on results obtained after applying the index in Foz de Almargem lagoon (Portugal), suggest that specific classification criteria should be developed to improve its performance in the assessment of water quality in lagoons. Because the thresholds have been developed for Mediterranean waters, their application might well be restricted to these waters, and the classification may have to be adapted for use in other geographical areas and systems. Our results support this view; of course, this would reduce the general utility of TRIX. The basic steps involved in adjusting the criteria involve the selection of a dataset that encompasses degraded and non-degraded reference sites in each major habitat type with information on associated pressure indicators. Obviously, this implies extensive knowledge of these habitats, as well as the availability of a large dataset to validate the criteria. So far, these constraints discourage a general application of TRIX in a European-wide context.
|
Acknowledgements |
|---|
The study was funded by "Respuesta de la red trófica de una laguna costera a las entradas de nutrientes" (EUTROCOST), "Efeitos dos caudais duciaquícolas sobre as comunidades de invertebrados macrobentónicos, na perspectiva da avaliação da qualidade ecológica dos estuários" (EFICAS; POCI/MAR/61324/2004), and "Sustainable management of Mediterranean coastal fresh and transitional water bodies" (WADI; INCO-CT-2005-015226) research projects. It was also supported by the FCT (Portuguese National Board of Scientific Research) through one grant (SFRH/BD/24430/2005). The Instituto da Água (Portugal) provided institutional support. We express special thanks to all who assisted us during field and laboratory work.
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
APHA. Standard Methods for the Examination of Water and Wastewater. (1995) Baltimore, MD: Ed. by American Public Health Association, American Water Works Association and Water Environment Federation. United Book Press.
Best M. A., Wither A. W., Coates S. Dissolved oxygen as a physico-chemical supporting element in the Water Framework Directive. Marine Pollution Bulletin (2007) 55:53–64.[CrossRef][Web of Science][Medline]
Coelho S., Gamito S., Pérez-Ruzafa A. Trophic state of Foz de Almargem coastal lagoon (Algarve, South Portugal) based on the water quality and the phytoplankton community. Estuarine, Coastal and Shelf Science (2007) 71:218–231.[CrossRef]
Crouzet P., Leonard J., Nixon S., Rees Y., Parr W., Laffon L., Bøgestrand J., et al. Nutrients in European ecosystems. In: Environmental Assessment Report, 4.—Thyssen N., ed. (1999) European Environment Agency. 82. http://reports.eea.eu.int/.
Devlin M., Best M., Coates D., Bresnan E., O’Boyle S., Park R., Silke J., et al. Establishing boundary classes for the classification of UK marine waters using phytoplankton communities. Marine Pollution Bulletin (2007) 55:91–103.[CrossRef][Web of Science][Medline]
Dolbeth M., Cardoso P. G., Ferreira S. M., Verdelhos T., Raffaelli D., Pardal M. A. Anthropogenic and natural disturbance effects on a macrobenthic estuarine community over a 10-year period. Marine Pollution Bulletin (2007) 54:576–585.[CrossRef][Web of Science][Medline]
EC. Urban Wastewater Treatment Directive 91/271/EEC. Official Journal of the European Communities (1991) a. L135/40–52, 30 May 1991.
EC. Nitrate Directive 91/686/EEC. Official Journal of the European Communities (1991) b. OJ L375, 31 December 1991.
EC. Water Framework Directive 2000/60/EC. Official Journal of the European Communities (2000) L327/1eL327/72, 22 December 2000.
EEA. Nutrients in European Ecosystems. Environmental Assessment Report, 4. European Environment Agency. In: Office for official publications of the European Communities (1999) a. 155.
EEA. Environmental signals 2000. Environmental Assessment Report, 6. European Environment Agency. In: Office for official publications of the European Communities (1999) b. 108.
EEA. Eutrophication in Europe’s coastal waters. Topic Report, 7. In: European Environment Agency, Copenhagen (2001) 86.
Elmgren R. Understanding human impact on the Baltic ecosystem: changing views in recent decades. Ambio (2001) 30:222–231.
Giovanardi F., Vollenweider R. A. Trophic conditions of marine coastal waters: experience in applying the Trophic Index TRIX to two areas of the Adriatic and Tyrrhenian seas. Journal of Limnology (2004) 63:199–218.
Lillebø A. I., Neto J. M., Martins I., Verdelhos T., Leston S., Cardoso P. G., Ferreira S., et al. Management of a shallow temperate estuary to control eutrophication: the effect of hydrodynamics on the system’s nutrient loading. Estuarine, Coastal and Shelf Science (2005) 65:697–707.[CrossRef]
Lillebø A. I., Teixeira H., Pardal M. A., Marques J. C. Applying quality status criteria to a temperate estuary before and after the mitigation measures to reduce eutrophication symptoms. Estuarine, Coastal and Shelf Science (2007) 72:177–187.[CrossRef]
López-Bermúdez F., Ramirez L., Martin Agar P. Analisis integral del medio natural en la planificación territorial: el ejemplo del Mar Menor. Murcia(VII) (1981) 18:11–20.
Loureiro S., Newton A., Icely J. Boundary conditions for the European Water Framework Directive in the Ria Formosa Lagoon, Portugal (physico-chemical and phytoplankton quality elements). Estuarine, Coastal and Shelf Science (2006) 67:382–398.[CrossRef]
Martins I., Pardal M. A., Lillebø A. I., Flindt M. R., Marques J. C. Hydrodynamics as a major factor controlling the occurrence of green macroalgae blooms in a eutrophic estuary: a case study. Estuarine, Coastal and Shelf Science (2001) 52:165–177.[CrossRef]
Marques J. C., Nielsen S. N., Pardal M. A., Jørgensen S. E. Impact of eutrophication and river management within a framework of ecosystem theories. Ecological Modelling (2003) 166:147–168.[CrossRef][Web of Science]
Nixon S. W. Coastal marine eutrophication: a definition, social causes and future concerns. Ophelia (1995) 41:199–219.[Web of Science]
OSPAR. Synergies in assessment and monitoring between OSPAR and the European Union. In: Analysis of synergies in assessment and monitoring of hazardous substances, eutrophication, radioactive substances and offshore industry in the North-East Atlantic. (2005) OSPAR Publication, 230.
Parsons T. R., Maita Y., Lalli C. M. A Manual of Chemical and Biological Methods for Sea-Water Analysis. (1984) Oxford: Pergamon Press. 173.
Penna N., Capellacci S., Ricci F. The influence of the Po River discharge on phytoplankton bloom dynamics along the coastline of Pesaro (Italy) in the Adriatic Sea. Marine Pollution Bulletin (2004) 48:321–326.[CrossRef][Web of Science][Medline]
Pérez-Ruzafa A., Aragon R. Implicaciones de la gestion y el uso de las aguas subterraneas en el funcionamiento de la red trofica de una laguna costera. In: Conflictos entre el desarrollo de las aguas subterraneas y la conservacion de los humedales: litoral mediterraneo—Fornes J. M., Llamas M. R., eds. (2002) Madrid: Fundacion Marcelino Bot
n-Ediciones Mundi-Prensa. 215–245.
Pérez-Ruzafa A., Gilabert J., Gutierrez J. M., Fernandez A. I., Marcos C., Sabah S. Evidence of a planktonic food web response to changes in nutrient input dynamics in the Mar Menor coastal lagoon, Spain. Hydrobiologia (2002) 475/476:350–369.
Pérez-Ruzafa A., Marcos C., Gilabert J. The ecology of the Mar Menor coastal lagoon: a fast-changing ecosystem under human pressure. In: Coastal Lagoons: Ecosystem Processes and Modeling for Sustainable Use and Development—Gönenç I. E., Wolflin J. P., eds. (2005) Boca Raton, FL: CRC Press. 392–422.
Pérez-Ruzafa A., Quispe-Becerra J. I., Garca-Charton J. A., Marcos C. Composition, structure and distribution of the ichthyoplankton in a Mediterranean coastal lagoon. Journal of Fish Biology (2004) 64:202–218.[CrossRef][Web of Science]
Rodrigues E. Variação das comunidades macrobentónicas subtidais do estuário do Mondego ao longo de mais de uma década. (2004) Master thesis, University of Coimbra, Portugal.
Strickland J. D. H., Parsons T. R. A Practical Handbook of Seawater Analysis. In: Bulletin of the Fisheries Research Board of Canada (1972) 167, 2nd edn. 311.
Vollenweider R. A., Giovanardi F., Montanari G., Rinaldi A. Characterization of the trophic conditions of marine coastal waters with special reference to the NW Adriatic Sea: proposal for a trophic scale, turbidity and generalized water quality index. Environmetrics (1998) 9:329–357.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||