Diatom species composition and indices for determining the ecological status of coastal Mediterranean Spanish lakes

Diatom indices have been used and tested mainly for assessing the ecological status of rivers and deep lakes, but there are scarce studies that determine their effectiveness in shallow lakes and in coastal Mediterranean lakes. This study evaluates the validity of several common diatom indices (SPI, BDI, CEC and TDIL) for the determination of the ecological quality of three coastal lakes (Valencia, Spain) and presents descriptions and ecological data of the main diatom species recorded. Diatom samples were collected from phytobenthos, both from epiphyton of the dominant submerged macrophytes and the sediment. The ecological status of the systems was determined according to different physico-chemical variables and was compared with the results obtained from epiphytic diatom communities. The results showed discrepancies among diatom indices and also with the state determined by the environmental variables. The effectiveness of the indices depended on the number of species assessed for each index with respect to the total species recorded and the suitability of the weight assigned to each species. The results reveal the need to gather more information about the composition and ecology of the diatoms and microalgae characteristic of coastal Mediterranean standing waters. This work contributes to their better knowledge.


INTRODUCTION
The Water Framework Directive requires to the European Union countries to monitor and control the ecological status of their water bodies, as an essential way to protect, improve and conserve the EU aquatic systems (WFD, Directive 2000/60/EC).Among the proposed bioindicators, benthic diatoms are one of the most relevant groups for the monitoring and control of the ecological status of aquatic ecosystems.These organisms are used in bioindication studies, by means of autoecological indices based on the relative abundance of each taxon, their sensitivity to environmental factors and their ecological distribution (Della Bella & Mancini, 2009;Cejudo-Figueiras, 2011).Most of the implemented protocols using diatom indices for the assessment of the water quality have been focused on rivers and deep lakes, mainly located in Central Europe (DOCE, 2008).However, there is less information about its application on shallow lakes (Kitner & Poulickova, 2003;Blanco & al., 2004;Stenger-Kovács & al., 2007), and particularly on Mediterranean shal-low lakes, even though there is evidence that their ecology differ from that reported in the temperate zone (Romo & al., 2004;Moss & al., 2004).Moreover, among the indices, there is a wide heterogeneity in their database and about the diatom species that they consider.Furthermore, their effectiveness varies depending on the ecosystem type and the eco-region (Cejudo-Figueiras, 2011;Álvarez-Blanco & al., 2011Álvarez-Blanco & al., , 2012)).Nowadays, the information about the use of diatoms indices in Mediterranean standing waters is scarce.Recently, the approach of adapting diatom indices commonly used in rivers to establish the ecological status of some inland waters in the Iberian Peninsula has been carried out (Blanco & al., 2004;Cejudo-Figueiras & al., 2010).However, similar studies of diatom indices application in coastal lakes are lacking.For example, coastal water bodies have higher salinity ranges than inland freshwaters, as well as different morphology and environmental impacts on their catchment areas.
This work tests the usefulness of several well-known diatom indices in some coastal shallow lakes.The study is a first approach to the evaluation of its suitability in the determina-Diatom species composition and indices for determining the ecological status of coastal Mediterranean Spanish lakes Beatriz Antón-Garrido 1 * , Susana Romo 1 & María José Villena 2 1 Departamento de Ecología, Edificio de Investigación, Universitat de València, Campus Burjasot, E-46100 Burjasot, Spain; beanga@alumni.uv.es;Susana.Romo@uv.es 2 Laboratorios Tecnológicos de Levante, S.L., E-46980 Paterna, Spain; mjvillen@uv.estion of the ecological status in this type of ecosystems for the implementation of the WFD.Furthermore, it reports detailed taxonomic descriptions and ecological data of the main diatom species, in order to contribute to the knowledge of the characteristic diatom flora inhabiting this type of ecosystems.

MATERIAL AND METHODS
The study was carried out in three interdunar, shallow and permanent lakes located in the Natural Park of Albufera and named as, Mata del Fang (MFPNS) and Lagunas del Canyar (CNP1 and CNP3) (Fig. 1).Their hydrological dynamic is characteristic of this type of Mediterranean water bodies, with seasonal fluctuations in the water level mainly during summer and water inputs from rain, groundwater and saline spray.They are surrounded by helophytic vegetation (Phragmites, Thypha, Juncus and Scirpus) and their bottom was covered by charophyte meadows, growing on a sand-silty substrate.
The ponds were sampled on 15 th October 2009.An integrated water sample from each pond was collected by sampling and mixing the water from several points across the pond and subsamples were used for water chemical analyses and other determinations.Depth, temperature, pH, conductivity and oxygen concentration in the water were determined in situ.Dissolved nitrogen and phosphorus concentrations, as well as phytoplanktonic chlorophyll a were analyzed according to standard methods (APHA, 1992).
Samples of epiphytic and benthic microalgae were also collected.Epiphyton samples were taken from the dominant macrophytes (Chara spp.).Epiphytic algae were detached from the plant by shaking in a standardized water volume that was fixed with lugol solution for further studies and taxo-Diatoms and bioindication in coastal Mediterranean lakes nomic identification (Zimba & Hopson, 1997).Benthic algae were sampled by collecting a total of 20 sediment surface cores (upper 2 cm) taken randomly all over each pond.The cores were integrated into a sample per pond and frozen for later analyses.Diatom samples were digested with hydrogen peroxide according to standard protocols (UNE-ENE 13946, 2004).Identification and counting were made from permanent preparations mounted with Naphrax ® and at least 400 valves were counted for each sample at ϫ1000 magnification.In addition, scanning electron microscope (SEM) analyses were carried out for accurate identification of some species.Morphometry of the dominant diatoms was determined by measuring length and width of a minimum of 20 individuals.The determination of the different taxa was carried following specific references : Germain, 1981;Krammer & Lange-Bertalot, 1991a, b, 1997a, b, 2000;Prygiel & Coste, 2000;CHD, 2010.The diatom indices SPI (Specific Polluosensitivity Index, CEMAGREF, 1982), BDI (Biological Diatom Index, Lenoir & Coste, 1996) and CEC (Comission for Economical Community Index, Descy & Coste, 1990) were calculated for the epiphyton by means of the software OMNIDIA version 5.1 (Lecointe & al., 1993).Only those taxa with relative abundance higher than 5% were considered in their calculation.The TDIL index was calculated according to Stenger-Kovács & al. (2007).These indices were selected for their wide use in Spain, as well as in other European countries (Blanco & al., 2008) and TDIL for being specific on shallow lakes.The ecological status of each pond was also assessed by physico-chemical metrics according to their habitat typology of permanent, coastal, interdunar lakes (BOE, 2008;CEDEX, 2010a, b).

RESULTS
The studied lakes were oligohaline, with almost saturated oxygen concentrations, neutral-alkaline pHs and water transparency reaching the bottom (Table 1).Phosphate values were low and similar in the three ponds, while nitrate was the main form of soluble inorganic nitrogen (Table 1).Ammonium and nitrite were below the detection level of the analytical methods.Mata del Fang (MFPNS) recorded the highest values of nitrate and phytoplanktonic chlorophyll a (Table 1).CNP1 had concentrations of phytoplanktonic chlorophyll a slightly higher than CNP3, even though their proximity (Fig. 1) and similar nutrient concentrations (Table 1).
The ecological status assessment of the ponds by using threshold values for pH, conductivity and mainly phytoplanktonic chlorophyll a concentration are shown in Table 2.
The diatom indices showed discrepancies with the ecological states determined by the environmental variables (Tables 2, 3).Furthermore, the quality assessment also differed between the indices, mainly depending on the number of recorded diatom species considered by each index in their calculation (Table 3).The most efficient diatom indices and those which considered higher number of the recorded diatom species were SPI and BDI (Table 3).However, com-pared to the environmental metrics, they seemed to overestimate water quality for MFPN, while underestimated that of CNP3 (Tables 2, 3).By contrast, CEC and TDIL indices were less efficient and they took into account lower percentages of the identified diatom species (Table 3).

Diatom species description
A total of 63 species of diatoms belonging to 30 genera were identified for the epiphytic and benthic communities of the three sampled lakes.The identification was carried out to species level in 60 cases and to genus level in 3 cases (Table 4).In general, most of the identified species had ecological requirements related to alkaline pH, tolerance to salinity and a wide range of nutrient concentrations.It was remarkable the abundance and presence in the three study lakes of species related to brackish environments, such Halamphora cf.sydowii and those species belonging to the genus Mastogloia, as well as other species, such as Seminavis pusilla, Fragilaria gracilis and Nitzschia elegantula (Table 4).In the epiphyton, Seminavis pusilla co-dominated in abundance, together with Rhopalodia gibba var.gibba in CNP3 and with Encyonema sp. in CNP1 and MFPNS (Fig. 2).Fig. 2. Percentage of relative abundance of the main epiphytic and benthic diatoms identified in the study.Abbreviated taxa names according to Table 4.
The diatom communities in the sediment had a species composition and abundance distribution similar to that of the epiphyton in CNP1, while in CNP3 and MFPNS there were more differences, although most of the recorded species appeared in both substrates (Fig 2 , Table 4).In MFPNS the benthic microalgal community was dominated by Pseudostaurosira brevistriata, while several Halamphora species predominated in CNP3.
A catalogue of the main diatom species found with a relative abundance higher than 5% is presented below, together with a description of their taxonomic characteristics.Light (LM) and scanning electron microscope (SEM) photographs were taken from the epiphytic and benthic samples of the aquatic study systems to illustrate each taxon.The biometric ranges provided correspond to individuals measured in the samples.
Ecology and distribution: characteristic of brackish waters, widely distributed in coastal waterbodies of the Mediterranean.In this study, recorded in all the lakes (Table 4).Ecology and distribution: broad distribution, circumneutral, fresh-brackish, oligo-eutrophic-␤-mesosaprobic waters.In this study, it was observed in CNP3 and MFPNS, mainly living on the vegetation.Ecology and distribution: in this study, it was mainly observed in the epiphyton and also in the sediment of CNP1 and MFPNS (Table 4).Ecology and distribution: alkalophilic, fresh-brackish, mesotrophic oxygenated waters.In this study, it was recorded in CNP1 (epiphyton and sediment) and in low number in the epiphyton of MFPNS (Table 4).Ecology and distribution: circumneutral, fresh-brackish, eutrophic-␣-mesosaprobic waters.In this study, it was mainly present in the epiphyton of MFPNS, CNP1 and CNP3 (Table 4).
Ecology and distribution: periphytic, alkalophilic, tolerant to a wide range of conductivity, oligosaprobic-oligo-eutrophic waters.Frequently cited in the Iberian peninsula.In this study, it was abundant in the sediment of MFPNS (Table 4, Fig. 2).Ecology and distribution: epiphytic, lotic and lentic brackish waters, typically considered euryhaline.Cited in different water bodies of the Spanish Mediterranean coast.In this study, described in the three lakes (Table 4).
Ecology and distribution: benthic, alkalophilic, freshbrackish, oligo-eutrophic-␤-mesosaprobic waters.Broad distribution.In this study, it was observed in the epiphyton of CNP1 and was abundant in that of MFPNS (less present in its sediment) (Table 4).Length 10-14.6 µm, breadth 3.4-4 µm.Linear-elliptical valves with rounded apices.Filiform raphe with marked central nodules.Axial area narrow-linear, central area irregular with a central-spaced stria on both sides.Bluntly striae, radial pores distribution in the center and parallel-convergent at the ends, striae do not reach the apices.SEM showed a polar raphe fissure and apical pores along the margin.

Navicula sp. Bory de
Valves lanceolate with rounded apices and dense, fine, barely marked striae 19-21/10 µm.Axial area very narrow, small, rounded central area.Centric, filiform, straight raphe, with proximal ends slightly bent to one side and marked central nodules.Parallel striations at the ends and striae slightly shorter and radial in the center.
Ecology and distribution: in this study, it was observed in the sediment of CNP3 (Table 4).Length 26-31.1 µm, breadth 5.4-6.6 µm.Valves moderately dorsiventral, with a convex dorsal margin and a ventral margin slightly convex in the center.Rounded ends.Raphe almost central, straight with the distal fissures curved towards the ventral margin and proximal ends towards the dorsal margin, central nodules marked.Thin hyaline area along the axial axis, with irregular margins and slightly wider at the center.Marked striae, 16-17/ m (dorsal and ventral), parallel at the ends and radial in the center (alternance of short and long striae).Pores distribution visible by SEM.

Seminavis pusilla (Grunow
Ecology and distribution: temperate, epicontinental coastal waters (Gemain, 1981).In this study was abundant mainly in the epiphyton of the three study lakes (Table 4, Fig. 2).Cells with dorsiventral valves, wider dorsally in the center, sharply curved hook-shaped ends.Marginal raphe-system on the dorsal side.Ventral margin straight, dorsal margin bent in the central area, usually has a small nick in the middle marking the position of the proximal raphe endings.Parallel costae in the center and radial at the ends with two rows of pores between transapical costae.
Ecology and distribution: characteristic specie of brackish waters.In this study, it was observed in the three study lakes, being abundant in the sediment of CNP3 (Table 4, Fig. 2).12/10 µm, composed by two rows of pores, which form a ring at the valve contour (SEM).Central region smooth or slightly undulated, fine radial striae, 1-4 fultoportulae visibles under light microscope.Small peripheral spines visibles by SEM.

DISCUSSION
The diatom indices SPI, BDI and CEC are widely used to assess water quality, mainly in rivers, both in Europe and the Iberian Peninsula, and only recently their suitability for lentic aquatic systems has been tested (Blanco & al., 2004;Cejudo-Figueiras & al., 2010).The TDIL index was specifically made to be applied in shallow lakes of Central Europe (Hungary) (Stenger-Kovács & al., 2007).However, coastal water bodies have higher salinity ranges and a different morphology and watershed impacts.In the present study, several problems were detected when these indices were tested for coastal shallow lakes.The main pitfall was the already limited knowledge gathered about the diatom flora and their ecological requirements in coastal Mediterranean standing waters (Trobajo, 2005;Cantoral-Uriza & Aboal, 2008).In our study, the effectiveness of the indices depended on the number of species assessed for each index with respect to the total species recorded and the suitability of the weight assigned to each species.For instance, several of the dominant species in this study, which have been also reported in other Mediterranean water bodies (e.g.Mastogloia spp., Tomàs, 1982), were omitted by the database of the indices tested.In addition, one of the most representative species in our study, Seminavis pusilla, was only included by the SPI and TDIL indices.This could also explain the observed discrepancy between indices, as a result of the different listed and weight given to diatom species by every single index (Table 3).The fact that taxonomy challenged the identification of some species, such as Encyonema sp., which was dominant in CNP1 and MFPNS, constitutes an additional problem to cope with the finest taxonomic accuracy required by the indices.Some of these problems have been also described by Blanco & al. (2004), when they used the SPI and BDI indices for the water quality evaluation of some inland shallow lakes of the Northwest Iberian Peninsula.Nevertheless, these authors found a good correlation between these indices and the lakes trophic status, based on the distribution of diatom abundances and trophic preferences.Therefore, to use more extensively diatom indices in coastal lentic systems, further studies should be carried out on the description and autoecology of their characteristic diatom flora.
The ecological requirements of the algal species may vary depending on the eco-region and thus the use of any diatom index must be calibrated for different ecosystems (Álvarez & al., 2011, 2012).The design of new and specific indices for different eco-zones is a new approach (e.g.DDI index, Álvarez- Blanco et al., 2012).This alternative approach, even though, is time and investment consuming might be more accurate for the ecological evaluation of coastal water bodies of the Mediterranean zone.
It is remarkable that a large number of the recorded diatom species in this study (38 in total, highlighted in bold in Table 4) have been also cited by other authors in coastal Mediterranean Iberian environments (lagoons, marshes, ponds, river mouths and streams) (Tomàs, 1982;Aboal, 1989;Trobajo, 2005;Cantoral-Uriza & Aboal, 2008;Rovira & al., 2009).This set of species could be a starting point for further studies, in order to define their role as bioindicators in coastal water bodies and for the design of new diatom indices.Therefore, more studies on the ecology and algal flora are needed to characterize, evaluate and implement the WFD in these ecosystems.
In general, the diatom species found in this study had a wide tolerance range to several physico-chemical variables, which agree with the dynamic nature of coastal water bodies, adapted, for instance, to water level and environmental changes (Trobajo, 2005;Della Bella & Mancini, 2009).The limnological variables evaluated in the present study complemented the diagnosis given for the diatom indices and resulted especially helpful to discriminate between possible water quality status (Tables 2, 3).This result agrees with that described by other authors that argued about the use of several metrics for the correct water quality evaluation of shallow lakes (Kitner & Poulícková, 2003;Moss & al., 2003).
In conclusion, our results suggest that diatom bioindication could be a useful tool for the determination of the ecological status of coastal shallow lakes, together with environmental metrics.However, it would be necessary modify or create new indices based specifically on the autecology and distribution of diatom species in the gradient of environmental conditions characteristic of the Mediterranean eco-zone.Therefore, it is important to gather more information about the composition and ecology of the microflora and diatoms of these ecosystems.

Table 1 .
Morphometric and physico-chemical characteristics of the study lakes.

Table 2 .
Ecological status assessment of the study lakes based on some environmental variables according to CEDEX (2010 a, b).

Table 3 .
Diatom indices SPI, BDI, CEC and TDIL with indication of the percentage of identified epiphytic diatoms species considered by each index and the ecological status assessment for the study lakes.

Table 4 .
List of diatom species identified in the study.Asterisk indicates diatom species used by the indices databases (SPI, BDI and TDIL).Relative abundance (percentage) of each species in the epiphyton and sediment.Diatom species cited in other coastal Mediterranean environments are highlighted in bold (see discussion section