Identification and characterisation of the dominant Pseudo-nitzschia species (Bacillariophyceae) along the NE Spanish coast (Catalonia, NW Mediterranean)

potentially toxic species of the genus Pseudo-nitzschia were studied along the Spanish nW mediterranean coast from january 2005 to may 2006. observation in electron microscopy revealed the presence of Pseudo-nitzschia brasiliana, P. calliantha, P. delicatissima, P. fraudulenta, P. multistriata and P. pungens. Several strains were isolated from coastal waters and their clonal cultures were compared by combined techniques, including light and electron microscopy and internal transcribed spacer (itS-1, 5.8S and itS-2) rdna sequencing. cultured isolates were submitted to hpLc analysis of pigments to evaluate the possibility of taxonomic discriminations by means of a simple chemotaxonomic approach. the genus Pseudo-nitzschia showed high cell concentrations during most of the year, but the population initiated a sharp decline at all stations in the period between april and may. P. delicatissima with P. calliantha were found at the northern stations between February and april, and at the southern stations between august and november. P. brasiliana and P. multistriata were only occasionally detected in the southern region. none of the environmental variables considered was significant to explain the observed spatial and temporal distributions of Pseudo-nitzschia species in the area studied. differences in the growth rate and cell yield of the species indicate that biotic factors may play a role in the observed distributional patterns.

Palabras clave: diatomeas, Pseudo-nitzschia, proliferaciones algales nocivas, autoecología, itS, morfología, pigmentos.introduction diatoms comprise one of the species-richest phytoplankton taxonomic groups.they have a worldwide distribution, can be found in both freshwater and marine environments, and are characterised by an extraordinary morphologic diversity (round et al., 1990; mann and droop, 1996; round, 1996).it is estimated that between 1365 and 1783 different planktonic diatoms occur in marine environments (Sournia et al., 1991).these include members of the genus Pseudo-nitzschia, commonly found in phytoplankton populations, with the highest cell concentrations in coastal regions (Lundholm et al., 2002a) the genus Pseudo-nitzschia currently harbours 29 different species of needle-shaped, raphid pennate diatoms forming chains of variable length (hasle and Syvertsen, 1997;Lundholm et al., 2002a;Lundholm et al., 2003;Lundholm et al., 2005).there are 11 or 12 species of Pseudo-nitzschia (depending on the inclusion of P. pseudodelicatissima) able to produce the neurotoxin domoic acid (da) which is associated with amnesic shellfish poisoning (aSp) in humans (Bates et al., 1989;Bates and trainer, 2006) and has also been found to induce mass mortality in sea birds, fish and mammals (Bates, 2000;mos, 2001;Sierra-Beltran et al., 2005;Schaffer et al., 2006).
For the mediterranean Sea, aSp events related to Pseudo-nitzschia were first described in 2001 (amzil et al., 2001), but along the south coast of France the presence of the toxin da (attributed to P. pseudodelicatissima) was detected as early as 1998.recent publications at this point describe detection of da in greek shellfish (Kaniou-grigoriadou et al., 2005), in P. multistriata (orsini et al., 2002) and in P. calliantha (inès and asma, 2006).
in the context of the monitoring programme of the catalan Water agency (aca), performed by the institute of marine Sciences in Barcelona (icm-cSic), the distribution and abundance of Pseudo-nitzschia and other potentially harmful algae are continuously evaluated along the ne coast of Spain (catalonia).this programme, initiated in 1995, currently in-cludes 16 representative harbours, several beaches and two coastal bays (vila et al., 2001;Furones et al., 2004).Because only optical microscope analyses were done, taxonomic identifications at species level were not performed.Such discriminations are, however, important for the understanding of bloom formation, because ecological preferences may vary significantly among species, whereas not all of them are actually toxic.discrimination of species within the genus Pseudo-nitzschia currently requires the use of Sem (scanning electron microscopy) and/or tem (transmission electron microscopy) techniques for morphological characterisation.
in the present study a combination of tem and Sem was employed to elucidate, for the first time, spatial and temporal distributions of Pseudo-nitzschia species along the ne Spanish coast during one and a half years.morphologic characteristics of the detected species were carefully analysed and compared with data available in the literature.to characterise the different Pseudo-nitzschia sp.identified at the sample location and enable comparative studies of specific characteristics in the future, a culture collection was initiated using isolates from the different locations sampled.For each strain, cell yield and growth rate were determined and the samples were submitted to hpLc analysis of pigments.Special attention was given to possible differences in composition and/or relative importance of individual pigment compounds, which would allow some degree of species classification using a simple analytical tool.moreover, itS region has been shown to be very informative (coleman, 2003; amato et al., 2007) for the identification of the genus Pseudo-nitzschia. the itS-1, 5.8S, and itS-2 sequence data were used as a means of confirming the identification already performed by light and electron microscopy analyses of morphologic features.
Finally, to identify the potential biotic and abiotic variables controlling the observed spatial and temporal distributions of Pseudo-nitzschia species in the study area, statistical analysis was performed.

Sampling
the stations sampled in the present study were selected on the basis of experiences obtained during 10 years of regular monitoring of the coast (vila, 2001) and included the period from january 2005 to may 2006.these locations (Fig. 1) cover relevant points of interest in the area, and include all those known to be affected by (recurrent) algal blooms.this coastal section is over 500 km long and many locations are densely populated.chemical conditions along the coast are strongly influenced by inflows of nutrient-enriched freshwater of continental run-off and the rivers, mainly the river ter (6-20 m 3 s -1 ) in the north, the rivers Llobregat (14-30 m 3 s -1 ) and Besòs (2.6-7.2 m 3 s -1 ) in the central part, and the large river ebro (130-630 m 3 s -1 ) in the south (idescat, 2006).
Water samples were collected (for details on the sampling frequency see Fig. 1) with a bucket, in which standard hydrographic parameters (temperature and salinity) were directly measured using WtW probe (model 315).Subsamples (150 ml) for taxonomic identification of the total phytoplankton population using inverted light microscopy (outlined below) were directly fixed with Lugol's iodine solution.additional subsamples (50 ml) for analyses of nutrients were stored on ice during transport to the laboratory and frozen (-20°c) upon arrival.nutrient samples were analysed with an autoanalyser as described in grasshoff et al., (1983).probable inorganic dissolved nutrient limitations were calculated as in justic et al. (1995).the criteria of probable nutrient limitation is as follows: p limitation (p <0.1 µm; din:p >22; Si:p >22), n limitation (din <1 µm;din:p <10; Si:din >1) and Si limitation (Si <2 µm; Si:po4 <10; Si:din <1).

Light microscopy
Subsamples (50 ml) were allowed to settle in counting chambers for 24 h and phytoplankton were subsequently enumerated in an appropriate area depending on the cell number (field or transect) using a Leica-Leitz dm-ii inverted bright field microscope with at a 200-400x magnification according to throndsen, 1995. a total of 618 samples were counted.the limit of detection of the utermöl method is 20 cells L -1 (utermöhl, 1931).identification of the species among the genus Pseudo-nitzschia in the light microscopy was attempted on the basis of the length and shape of the cells (trainer and Suddleson, 2005).

Electron microscopy
When observations with the light microscope revealed a Pseudo-nitzschia spp.population exceeding 10 4 cells L -1 , Lugol-fixed samples were subjected to Sem (63 samples) and tem (7 samples) procedures.Samples were observed at tem to distinguish the fine structure of the poroids and confirm identification.Samples with lower concentrations (<10 4 cells L -1 ) were not studied because little material was left after cleaning of the samples.
For both techniques (Sem and tem) organic material was removed from the samples with sulphuric acid and potassium permanganate with later addition of oxalic acid as described in Lundholm et al. (2002b).For Sem the remaining material was mounted on a polycarbonate filter and this was attached on stubs with colloidal silver and then sputtercoated with gold-palladium.the stubs were screened by a hitachi S-3500n microscope operating at 5 kv.For analysis with tem, drops of cleaned material were placed on Formvar-coated copper grids, dried and studied with a hitachi h800 microscope.

Morphometric characteristics
cells identified as Pseudo-nitzschia were carefully examined for several characteristics under Sem, including the width and length of the valve, the density of striae, fibulae and poroids on valves, the structure of girdle bands and the pattern of perforation in the poroid hymens. in each sample, organisms were identified in a different appropriate area of the filter, from one transect to a whole filter depending on the cell abundance (minimum cell number measured= 30).Length and width of all the whole cells were always measured (table 1) and the number of measurements of fibulae and striae were variable and depended on the species.the percentages of each species obtained in Sem were applied to the cell counts.

Clonal cultures
From live field samples examined under an inverted microscope (Leica dm-ii inverted bright field microscope), cells identified as Pseudo-nitzschia spp.were isolated with a glass pasteur pipette and transferred to a tissue culture flask filled with silicate-containing f/2 or L1 medium (guillard, 1975; guillard and hargraves, 1993).these flasks were subsequently maintained at 19-21±1ºc using a 12:12 h light:dark cycle.illumination was provided by fluorescence tubes (gyrolux, Sylvania, germany), providing a photon irradiance of 100 µmol photons m -2 s -1 .With these procedures a total of 42 clonal cultures (belonging to five species identified with Sem) were established.in this study we present data on growth rate, cell yield and molecular and pigment characterisation for one representative strain of each species.

Growth rates
the growth rates of the strains in L1 medium prepared in filtered seawater with a salinity of 36 were determined.Batch cultures were grown in 50 mL polycarbonate bottles and maintained in the growing conditions described above.growth experiments were started with an exponential population of parental strains.every second day, 1 mL subsamples were fixed in Lugol's iodine and counted in a cell chamber under an optical microscope (Leica-Leitz dm-il, Leica microsystems gmbh, Wetzlar, germany) at a 200-400x magnification, and the number of cells was recorded.cell abundance was used to calculate exponential growth rates according to the method of guillard, 1973.

DNA extraction, PCR amplification and sequencing
all samples for sequencing were obtained from 15 ml of pure non-axenic cultures.dna was extracted with the dneasy® plant Kit (Qiagen) following the manufacturer's protocols, with the exception of the elution volume, which was reduced to 25 µl. the ri- bosomal dna region, including the itS-1, 5.8S and itS-2, was amplified using the primers microSSu (5'-gtgaacctgcg-gaaggatc-3') and dino e (5'-ccKSttcaYtcgccrttac-3') in a 25 µl reaction tube containing 1X polymerase buffer (invitrogen), 1.5 µm mgcl 2 , 400 µm dntp's, 2 µm of each primer, and 1 u taq dna polymerase (invitrogen).all of the amplifications were carried out in an eppendorf mastercycler using a program of 40 cycles of 30 sec at 94ºc and 30 sec at 54ºc, followed by 1 min at 72ºc with a pre-cycling incubation of 5 min at 95ºc and a post-cycling incubation of five min at 72ºc. the resulting amplicons of approximately 800 bp were cloned into the vector pcr2.1-topo(invitrogen) for growth in escherichia coli dh5α.tranformant bacterial clones were screened using pcr and the primers and conditions shown above.one positive transformant clone representative of each isolate was submitted to a private company (Sistemas genómicos, valencia, Spain) for bidirectional sequencing.the plasmid clones of the rdna were obtained in order to maintain a permanent record of these species sequences for future reference and/or other projects.the sequences obtained from these clones are those used in this study.however, direct sequencing of pcr products of some of these strains as well as others obtained during the same time period as this study have not shown any significant indels or intrastrain variation of the it-S, 5.8S, and itS-2 rdna.the sequences were analysed using a BLaSt search with no a priori knowledge of the morphometric characteristics or the possible identities of the cells.

Pigment analyses
For pigment analyse algal cells from 50 mL cultures in exponential growth (f/2 and L1) were concentrated on 25 mm gF/F filters by vacuum filtration (-25 Kpa). the filters were extracted in 3 mL 90% acetone and analysed with the hpLc method of Zapata et al., (2000).details of the sample treatment, the chromatographic setup and adaptations of the method employed were previously described by van Lenning et al., (2003).the hpLc was calibrated with authentic pigment standards obtained from dhi, denmark.concentrations of accessory pigments detected in individual cultures were normalised to chlorophyll a (chl a), the proxy for algal biomass, and their contributions to the total pigment load were calculated.results obtained for 3 different strains of the same species were averaged and the final results are presented for comparison (table 2). to follow the variation of the pigment composition in P. delicatissima over the growth period a time series was performed.every day, samples from a culture were taken and subsequently analysed until the stationary phase.

Statistical analysis
to identify potentially variables controlling the temporal and spatial dynamics of Pseudo-nitzschia sp on the catalan coast one matrix was created (temperature, salinity, freshwater content (FWc), chlorophyll a, dissolved inorganic nutrients and Pseudonitzschia species abundance, n=44).the species P. brasiliana appeared only once in the study period, so the sample was excluded for the statistical analysis.
prior to all analysis abiotic data were transformed v' = log10 (v + 1) and biotic data were transformed v' = (v+1).as no variable showed a normal distribution (Kolmogorov-Smirnov and Shapiro-Wilk) only non-parametric statistical analyses were applied.the following analyses were performed: one-way analy- sis of similarities (anoSim) with the corresponding pairwise tests if needed, cluster analysis using the group average method and working with euclidean or Bray curtis distances with the corresponding similarity profile (SimproF) to create objectively-defined groups, multi-dimensional scaling (mdS), BeSt (biota and environment matching test), and similarity percentages (Simper) in order to determine crucial variables.all these analyses were performed using the primer 6 statistical package (clark and Warwick, 2001).additionally, tables using Spearman rank order correlations among all variables, transformed v' = log (v+1), were created using the StatiStica 6.1 statistical package (StatSoft, 2003).reSuLtS
considering the cell length, two groups can be distinguished: smaller than 60 µm, which includes P. brasiliana, P. delicatissima and P. multistriata; and larger than 60 µm, which includes P. calliantha, P. fraudulenta and P. pungens.moreover, the width of cells was considered to be an important morphological characteristic (hasle et al., 1996;trainer and Suddleson, 2005).Based on this character Pseudonitzschia species were subdivided into two groups: the "P.delicatissima group" with narrow valves (width ≤3 µm) and the "P.seriata group" with wide valves (width ≥3 µm). the first group includes the three species P. brasiliana, P. calliantha and P. deli-catissima.the second group includes P. fraudulenta and P. pungens.P. multistriata has a range of widths that overlaps the 3 µm threshold value (table 1).
the analyses of the itS-1, 5.8S and itS-2 region of the rdna from our strains were compared with representative strains of genBank and maximum identity was estimated using a Blast analysis (table 3).

Growth rates and cell yield
of the five species, P. delicatissima showed the highest growth rates and cell yield recorded in culture (1.61 d -1 ). the lowest growth rate was obtained from P. fraudulenta and the lowest cell yield from P. pungens (table 3).

Pigment composition of cultures
most pigments identified in the culture collection were common to all five Pseudo-nitzschia species considered, regardless of the strain, growth medium or physiological stage.chl a comprised on average 39.3% of the total pigment content (tpc) identified in the cultures.common accessory compounds were also estimated (table 2).trace levels of a chl c 2like Pavlova gyrans-type pigment (0.3% tpc) and mgdvp pigment (0.1% tpc) should probably be included in the range of common Pseudo-nitzschia pigments, but they were not always within detection levels.chl c 1 , the monovinyl analogue of chl c 2 , was only detected in cultures of P. fraudulenta (3.8% tpc).the chl a degradation compound chlorophyllide a was detected in all cultures, but was always exceptionally abundant in extracts of P. delicatissima and P. pungens (20.5-22.1 and 11.6-24.8% tpc in F2 and L1 media, respectively).a second chlorophyllide a form (chlide a-2; 1.7% tpc) with a retention time similar to chl c 1 was detected in P. delicatissima grown in L1 medium.Synthesis of chl c 1 comprised the only significant difference in pigment compositions detected, and its possible coelution with chlide a-2 underlined the need for on-line spectral data regarding identification purposes.in aging cultures several additional compounds came within detection levels (Fig. 4), including a chl c 3 derivative, a phaeophorbide, an unknown carotenoid, a lycopene derivative and a range of chl a derivatives (like allomers).these compounds were not detected during the exponential phase, and none of them was typically associated with a specific species.the carotenoid diatoxanthin is usually considered to be a common compound of the light-harvesting system in diatoms, but during the present study this carotenoid was not positively identified in any extract analysed.results obtained with the P. delicatissima pigment time-series (Fig. 5) revealed that degradation of chl a occurred at all times, but with an increasing tendency with age of the culture.the sum of the accessory pigments identified in the Pseudo-nitzschia culture collection (not including the chl a degradation products) comprised on average 50.2%tpc, with minor variations between species and culture medium (table 2).the relative importance of chl a + chlide a, however, only showed minor variations between the 9 days of sampling (Fig. 5), with a mean value of 43.3% tpc, similar to the value obtained for the total culture collection analysed (table 2). in all five species the accessory pigments were strongly dominated by fucoxanthin, and the relative importance of this specific carotenoid oscillated within a narrow range, from 26.7% tpc in P. fraudulenta to 30.7% tpc in P. delicatissima. the relative importance of fucoxanthin always yielded the lowest values in cultures characterised by the lowest degradation of chl a. chl c 2 , the second most important pigment, showed relatively large variations between species, ranging from 7.9% tpc in P. fraudulenta to 16.5% tpc in P. pungens, but the values did not change greatly according to the growth media.Large variations were observed even for chl c 3 (0.89-9.13% tpc), whereas the far less important pigments (<3% tpc) diadinoxanthin, β,β-carotene and chl c 2 -P.gyrans , showed only minor variations.during the P. delicatissima pigment time-series the relative importance of the accessory pigments oscillated within a very narrow window, showing no clear variations according to the culture age.

Environmental variables accompanying high cell densities of Pseudo-nitzschia species
during the sampling period, surface water temperature values fell within the range commonly reported for the coastal surface waters of the nW mediterranean, ranging from 8.2 to 28.9ºc.the periods of cold waters (8 to 16ºc) were january to march 2005 and november 2005 to april 2006, and the warm period was april to october 2005. in general, during the warm period at the south station the temperature was higher than in the northern area (Fig. 6a).Salinity ranged from 18 to 38 in the south, especially at Station 22, where lower salinity was recorded due to the influence of the water discharges of the irrigation channel in this area (Fig. 6B).
the highest din concentration was found at Station 8 during the study period with values above 100 µm, whereas at the south station values of approxi-   mately 20 µm were found (Fig. 6c).po 4 showed a concentration above 0.5 µm sporadically, but in the period from September to november 2005 there was an increase at all the sample stations (Fig. 6d).
Both Sio 4 and din showed a high concentration at Station 8 and at the south stations, especially Station 22, where the freshwater supply normally occurs (Fig. 6e). the din:p ratio was affected by the high concentration of din at Station 8, showing both similar pattern.moreover, in the south high values were measured (Fig. 6F).the din/Si ratio showed high values in the south region from january to april 2006.probable Si limitation was found mainly at Station 10, and from may to September at station 5, 13 and 14 principaly (Fig. 6g).probable din limitation was found only twice and p limitation three times.during the study period, chl a showed a wide range of values, with a maximum of 91.8 mg L -1 in july and a minimum of 0.05 mg L -1 in September (data not shown).

Pseudo-nitzschia spp. distribution patterns
Pseudo-nitzschia spp.were present along all the coast, with high cell abundance in winter-early spring (january to may) 2005, and in early 2006.moreover, in 2005, high cell abundance was observed at the southern stations from late summer to autumn (September to november 2005).periods of no cell presence were restricted to the warmest months (june and july 2005) on the entire coast and december to january 2006 at the northern stations (Fig. 6a). the distribution patterns of the most abundant species along the coast (the 23 stations sampled were arranged from n to S during the 17 months of study) is presented in Figure 7.
Pseudo-nitzschia brasiliana was only found at Station 13 in September 2005, with a maximum cell abundance of 2.58 x10 5 cells L -1 (Fig. 7a).
P. calliantha was mainly found from january to april at the northern stations in both 2005 and 2006. in the south it was present sporadically at Station 20 in january.at Station 22 a persistent bloom of this species was present from august to november 2005 and from january to march 2006 (Fig. 7B).maximum cell abundance (>10 5 cells L -1 ) was found at Stations 4, 16, 22 and 23.
P. delicatissima was present during the cold months at the northern stations in 2005. in the south it was present at Station 22, but with lower abundance than P. calliantha. in 2006 it was abundant in the south from december to may (Fig. 7c).maximum cell abundance (>10 5 cell L -1 ) was found at Stations 4, 16, 18, 19 and 23.
P. fraudulenta was present in 2005 in the north from February to april with relative low densities (>10 3 cell L -1 ). in 2006, the species appeared sporadically in the same months in the south (Fig. 7d). in this study, the maximum cell abundance of the species was 2.61 x 10 5 cells L -1 and it was mainly mixed with P. calliantha.
P. pungens was present in the north from February to may in 2005 and sporadically at other stations (Fig. 7F).its cell abundance was no higher than 10 5 cells L -1 .

Statistical analysis
no significant differences between samples from the north and south stations were noted in the anoSim tests either with biotic and abiotic variables or between seasons.however, using only data of abundances of Pseudo-nitzschia species and performing a cluster with its corresponding SimproF test (p<0.05),three groups of samples were differentiated as is shown in the mdS test (Fig. 8a): one (a) characterised by a high abundance of P. delicatissima, which includes all samples from Stations 01, 03, 17 and 21; one (b) characterised by high abundances of P. delicatissima and P. calliantha, which includes all samples from Stations 07, 10, 12 and 18; and one (c) characterised by high abundances of P. calliantha, which includes all samples from Stations 02 and 05 (table 4).means and standard deviation (Sd) of abundances of each group are shown in table 5. in contrast, using only data of abiotic variables and performing both a cluster and a SimproF test (p<0.05),two groups of samples were identified, as shown in mdS (Fig. 8B): one (a) characterised by a low concentration of no 3 and Sio 4 , which includes all samples from Stations 01, 02, 03, 05, 07, 10, 11, 12, 15, 16, 18, 21 and 22; and one (b) characterised by a high concentration of no 3 and Sio 4 , which includes all samples from Station 17 (table 4).means and Sd of abundances of each group are shown in table 5. abiotic and biotic groups did not match and this result was confirmed by BeSt analysis, which demonstrated that no abiotic variable or combina-tion of abiotic variables could accurately explain the distribution of Pseudo-nitzschia species in the area of study.this result was again corroborated by the correlation table (table 6), which showed a low degree of association between abiotic variables and abundances of Pseudo-nitzschia spp.nevertheless, several species of Pseudo-nitzschia showed significant correlations with some abiotic variables: P. delicatissima with temperature (positive correlation), salinity (negative correlation), freshwater content (positive correlation) and no 2 (positive correlation); and P. fraudulenta with salinity (positive correlation), freshwater content (negative correlation) and no 2 (negative correlation).Furthermore, P. delicatissima was negatively correlated with P. fraudulenta, and in relation to abiotic variables, temperature was negatively correlated with salinity and positively correlated with freshwater content, Sio 4 was positively correlated with no 3 , and po 4 was positively correlated with nh 4 .diScuSSion improvements in taxonomic identification and differentiation of Pseudo-nitzschia under the optical microscope have been a continuous challenge for many studies carried out during the last decade.detailed examination of such organisms requires electron microscopic examination of ultrastructure and analysis of genetic variations.We studied strains of different Pseudo-nitzschia species isolated from the Spanish coast (nW mediterranean) from 2005 to 2006. the strains were characterised using morphological, molecular, and physiological methodologies in order to understand the species present in the field.moreover, field samples were observed with both optical and electron microscopy to identify the ecological variables under which these species proliferate in high densities.
the morphometric measurements of the Pseudonitzschia species found on the catalan coast fit with those reported in previous studies on P. brasiliana, P. multistriata and P. pungens.however, some data differ for P. calliantha, P. delicatissima and P. fraudulenta.We reported length values of up to 123.1, 78 and 132.5 µm, while other works reported maximum lengths of 98, 71 and 98 µm, respec-tively. the maximum length values were recorded during periods of high cell abundance.it has been hypothesised that during a bloom, sexual reproduction occurs in order to restore the size of the species (drebes, 1977;Bates and davidovich, 2002;mann, 2002).this suggests that in high cell abundance of Pseudo-nitzschia sp cells are larger because sexual reproduction has occurred.
another important morphological aspect is the width of the cells that are used in defining species.Following the literature, the "P.seriata group" included wide species (width ≥3 µm) (hasle and Syvertsen, 1997).From this study, P. multistriata had a width range that overlapped the 3 µm threshold value.Since other morphological characters differentiate P. multistriata, we agree with (orsini et al., 2002), who suggest that the width of the cell is an ambiguous character for grouping species.
the itS-1, 5.8S, and itS-2 sequence data, via a BLaSt search, were used to strengthen the identification already performed by light and electron microscopy analyses of morphologic features of the species cultured.in each case the first species listed in the BLaSt results was that identified by morphological analysis.it is worth noting that many such BLaSt searches provided results which showed sequences of near complete identity but sometimes with two to four different species names interspersed.this indicates some degree of disagreement between morphology and sequence identity (which may be either due to incorrect morphological identification or sequences which were submitted without any careful examination by electron microscopy) or the possibility of pseudo-species.For this reason we point to the need for the morphometric analysis accompanying the sequencing in order to avoid the erroneous identification of species in genBank.
in hpLc results obtained for natural field samples or unialgal cultures we usually observe that concentrations of chl a comprise about 50% of the total pigment load (tpc) detected.however, in cultured isolates of Pseudo-nitzschia analysed in the present study the relative contribution of chl a was far less.this was attributed to a marked degradation of chl a to chlide a. the sum of chl a and its degradation product chlide a accounted, on average, for 49.9% tpc. the highest degradation was observed in extracts of P. pungens and P. delicatissima grown in L1 medium, regardless of the culture age.Since hpLc pigment samples were extracted in acetone, which (unlike methanol) hinders activity of chloro-  none of the environmental variables studied seemed to play an important role in either the spatial or the temporal distribution of Pseudo-nitzschia spp.(BeSt and mdS analysis).however, even if there is no statistical relationship between abiotic and biotic variables, samples could be objectively grouped depending on: a) the P. delicatissima and P. calliantha abundances, and b) the no 3 and Sio 4 concentrations.the distinction between the groups based only on these two species of Pseudo-nitzschia spp is obvious, as both species are the most abundant in the catalan coast.the differences between groups in the no 3 and Sio 4 concentrations responds to the fact that it is possible to distinguish stations with high freshwater influence-and consequently a high concentration of both inorganic nutrients-from stations with poor inflows of freshwaters and low nutrient concentrations.in addition, all correlations obtained in this study between abiotic variables and abundances of Pseudonitzschia spp.were low, confirming the lack of relationship between biotic and abiotic variables, which was also demonstrated by BeSt analysis.these results are not in agreement with those described by penna et al. (2006), which showed a negative relation between Pseudo-nitzschia spp.and phosphate.
although none of the specific abiotic variables measured explain either spatial or temporal distributions of the species in our coastal area, some general trends can be described and thus compared with previous knowledge of each species.P. brasiliana was found in September with a water temperature of 24.9ºc.this agrees with the fact that the species has been observed mainly in warmer waters for e.g.Brazil, the gulf of panama, the gulf of mexico, the gulf of california, vietnam, indonesia, thailand and South Korea.the species has recently been described in mediterranean waters (Quijano-Scheggia et al., 2005).
P. calliantha has been found in denmark, norway, the north atlantic, Scotland, Kiel Bay, ría de vigo in Spain, the Black Sea, the adriatic Sea, northern canada, the gulf of mexico, the Bermudas, chile, vietnam and australia: geographically widespread observations which indicate a fairly cosmopolitan distribution (Lundholm et al., 2003).recently, a toxic clone was found in tunisia (inès and asma, 2006). in the bay of Banyuls-Sur-mer, nW mediterranean Sea, P. calliantha was found associated with warmer temperatures and relatively nutrient-rich waters (Quiroga, 2006).on the other hand, caroppo et al. (2005) found that P. calliantha was negatively correlated with water temperature and positively correlated with nutrient availability.no significant relationship between temperature or dissolved inorganic nutrients and the abundance of P. calliantha was found during the study period.
P. delicatissima is common, occasionally as the predominant diatom species, in the north atlantic (hasle and Syvertsen, 1997). in the gulf of naples it was found to produce regular blooms in late spring, with less recurrent peaks in late summer (orsini et al., 2004). caroppo et al. (2005) found that the dynamics of these species appears to be significantly correlated with the environmental features.in the present study we found only a correlation with no 2 and P. delicatissima show high cell abundance under very different abiotic conditions.moreover, this species showed the highest growth rate and cell yield in lab conditions.Both facts suggest that the species could be described as one of the r-selected phytoplankton species, which have optimised their fitness for conditions with ample resources and a high growth rate.in fact, when phytoplankton were ranked after their growth rates, one of the highest values was recorded for diatoms (Stolte and garcés, 2006).
P. fraudulenta is a cosmopolitan species (hasle and Syvertsen, 1997) in the mediterranean Sea that is more abundant during spring but never attains a density higher than 10 5 cells L -1 (Zingone et al., 2006).on the catalan coast, this species appears mixed with other Pseudo-nitzschia species such us P. pungens.moreover, it showed the lowest growth rate in the lab, suggesting that it took a long time to reach high cell densities.
P. multistriata appears in spring in the mediterranean Sea (Zingone et al., 2006).however, in this study the species showed high cell abundance in summer as reported from inlets of southern japan, where it forms blooms in summer (hasle and Syvertsen, 1997).the species has been reported as toxic (orsini et al., 2002) but our cultures showed no toxicity (Franco, pers. comm.).

acKnoWLedgementS
We are grateful for the technical help provided during tem analyses at the university of Barcelona, r. ventosa for the nutrient analyses, a. reñe is thanked for field data and p. Franco for toxin analyses.this study was supported by the agència catalana de agua (aca) contract, the ec-funded research project Seed, goce-ct-2005-003875, and the cra (centre de referencia en aquicultura, cirit-generalitat de catalunya, Spain).S. Quijano's work was supported by a promep grant, colima university (mexico).e. garcés's work was supported by a ramon y cajal contract of the Spanish ministry of education and Science. the work of K. andree was supported by the project irta-2005-00109-00-00, inia (instituto nacional de investigación y tecnología agraria y alimentaria) reFerenceS

Fig. 4 .
Fig. 4. -representative absorbance chromatogram obtained for an extract of Pseudo-nitzschia multistriata culture, simultaneously containing most pigments and degradation products detected in the culture collection employed.

Fig. 5 .
Fig. 5. -temporal variations in the relative contributions of accessory compounds to the total pigment content observed during growth of P. delicatissima in L1 medium.

Fig. 6 .
Fig.6.-temporal-spatial distribution of; a) Pseudo-nitzschia spp and temperature; B) salinity.the dots on the graphs are the sampling stations c) din µm; d) po 4 µm; e) Sio 4 µm; F) ratio din/po 4 (note that there is no limitation at any point); g) ration din/Sio 4 (q indicated places were there are limitations of Sio 4 according to justic et al, 1995).

Table 2 .
-mean relative contributions of chl a (n.c.).degradation products and accessory pigments to the total pigment load (%) determined for a culture collection of five different Pseudo-nitzschia species grown in F2 and L1 culture media.vertical arrangement of accessory pigment is based on descending mean relative importance.
*only considering samples in which it was identified; nt: not detected.

Table 3 .
-culture strains of different species from nW mediterranean.Strain code, genBank number, growth rate and cell yield of each strain.reference strains were used from the genBank: source, accession number and maximum identity estimated using a Blast analysis.

Table 5 .
-Pseudo-nitzschia abundance (mean ± standard deviations) of each biotic group and abiotic variables (mean ± standard deviations) of each abiotic group.Table6.-Spearman correlations between abiotic variables and Pseudo-nitzschia spp.abundances.Bold correlations are significant at p<0.01.

Table 4 .
-Summary of critical variables and stations for each abiotic and biotic groups.Station only present in that group in bold.