Introduction

Blooms of phytoplankton organisms are recurrent events in European coastal waters and worldwide (Anderson, 1989), being potentially harmful to human health, marine ecosystems and resources such as tourism, fisheries and aquaculture. The socio-economic impact of HABs (Harmful Algal Blooms) highlights the need for ecological studies addressed to the understanding of HAB dynamics and processes of bloom formation and dispersal. Summer outbreaks of the dinoflagellate Alexandrium taylorii Balech produce dense yellowish–green patches of limited extension in coastal waters of Sicily, Italy (Giacobbe & Yang, 1999; Penna et al., 2002), a central island in the Mediterranean Sea. Beyond the local phenomenon, similar events are also observed in beach areas along the Costa Brava and Balearic Islands (Spain), causing in some localities water discolouration lasting two months (Garcés et al., 1999). Dense aggregates of cells are also found near the sediment, this representing a possible way of avoiding advection from the area and prevent population losses. As a whole, the observations reported in literature suggest that the magnitude of the event is increasing in some places and that this species is in a phase of expansion, probably due to ballast water release and other causes.

Some authors have shown the morphological variability of A. taylori, the high production of temporary cysts having a significant role in the bloom dynamics (Garcés et al., 2002), and the existence of a sexual resting stage as a part of its life cycle (Delgado et al., 1997, Giacobbe & Yang, 1999). Thus, the specific identity may be easily masked when dealing with encysted stages—vegetative or sexual resting cysts.

In this study, we examined a set of biological and environmental data obtained in the last four years (2000–2003, EU Project Strategy, EVK3-CT-2001–00046) from a Tyrrhenian locality of Sicily subject to bloom events. Sequence analyses of different strains of A. taylorii isolated from various Mediterranean areas, as well as results of the application of molecular probes (by PCR assays) to preserved field samples, are also provided.

Study area

The West Bay of Vulcano (Sicily), a tourist locality of the Aeolian Islands (Tyrrhenian Sea), is a shallow, semicircular area extending approximately 300 × 100 m. The site is located at 38°25′04′′–10′′ Lat N and 14°57′14′′–23′′ Long E and covers as a whole about 7 ha. Recurrent summer blooms of A. taylorii take place in some nearshore points of the bay with evident water discolouration due to the high-biomass produced by this dinoflagellate (Penna et al., 2002). Before 2000, there had been no studies on HABs in this Tyrrhenian locality, although popular references on events of red tide date back to 1995.

Materials and methods

Sample collection and analyses

Three nearshore stations (1 m depth) located in the West Bay of Vulcano (Stn 1: 38°25′4′′ N, 14°57′14′′ E; Stn 2: 38°25′6′′ N, 14°57′19′′ E; Stn 3: 38°25′10′′ N, 14°57′23′′ E) were sampled in spring-summer 2000–2003, on 7–19 occasions each year. 100–250 ml of surface seawater samples were collected and fixed with Lugol’s solution (0.4% final concentration). Samples were taken at the same time of day (between 12 and 15 h). The bloom density was determined by cell counts of target species and other phytoplankton groups. The procedure for identification and quantification of phytoplantkon cells involved: sedimentation of a subsample in 10 or 50 ml settling chambers and counts of cells in an appropriate area (Throndsen, 1995) using an inverted Zeiss Axiovert 200 microscope equipped with epifluorescence. The identity of A. taylorii was confirmed by staining the cells with Calcofluor White M2R (Fritz & Triemer, 1985) and using a Zeiss UV filter set Fs 01.

Measures of environmental parameters such as water temperature, salinity and oxygen were taken by a mobile Multiprobe (YSI 556 MPS). Nutrient samples were filtered and frozen immediately after water collection. Concentrations of nitrate, nitrite and phosphate were determined following the Strickland and Parsons (1972) method, and ammonia according to Aminot & Chaussepied (1983).

Clonal cultures of Alexandrium taylorii were established at the Consiglio Nazionale delle Ricerche (CNR, Messina), Institut de Ciències del Mar (CSIC, Barcelona), and Instituto Español de Oceanografia (IEO, Vigo) from unfixed seawater samples taken at various sites of the Mediterranean. All marine cultures were maintained in F/2 or F/20 medium (see http://ccmp.bigelow.org/), at 18 ± 1°C and 12:12 h (light:dark) photoperiod. Illumination was provided by a photon irradiance of 100 μmol m−2 s−1.

Molecular assays

Total DNA extractions of different A. taylorii isolates and mixed phytoplankton populations from seawater samples collected in different localities of the Mediterranean basin—Vulcano (Italy), St. Carles, Catalan Sea (Spain), and Saronikos Gulf (Greece)—were performed by using DNeasy Plant Kit (Qiagen, Valencia, CA, USA), according to the manufacturer’s instructions. PCR amplification conditions of the 5.8S gene and ITS regions were performed as described in Penna & Magnani (1999). PCR amplified products of all A. taylorii isolates used in this study were directly sequenced. The sequences were deposited at EMBL Bank and given in Table 1.

Table 1 List of A. taylorii strains, sample locations, origin, source and EMBL accession numbers from this study
Full size table

Sequence alignment was done with Clustal X software and subsequently improved by eye. The rDNA-ITS regions were aligned with other ITS sequences of Alexandrium species, listed in GenBank. Sequence alignment analyses were also carried out by including other sequence data of Alexandrium species from GenBank (Table 2). Phylogenetic relationships, based on the 5.8S rDNA-ITS regions, were inferred using the Neighbor-Joining (NJ) (Saitou & Nei, 1987). The sequence of Gymnodinium sanguineum Hirasaka [=Akashiwo sanguinea (Hirasaka) Hansen & Moestrup] (GenBank AF131075) was used as outgroup. Robustness of the phylogenetic tree generated by NJ was tested by using the non-parametric bootstrap (Felsenstein, 1985) with 10,000 pseudoreplicates. The above analysis was performed with the software package MEGA ver. 3.0.

Table 2 List of Alexandrium species 5.8S rDNA and ITS region sequences used in this study
Full size table

A. taylorii specific PCR assays were carried out on the genomic DNA of mixed phytoplankton population samples by using two species-specific designed primers (Penna et al., 2004). The PCR products were visualized on 2.0% agarose gel.

Results

Spring-summer A. taylorii densities in the last four years (2000–2003) showed exceptionally high values (Fig. 1), compared with the concentrations of phytoplankton usually found in Sicilian coastal waters. On various occasions (2001–2003), the bloom density reached a 107 order of magnitude, with maxima of 2−3×107 cells l−1 and amounts of Chla as high as 50−180 μg l−1 in June 2002 and 2003. During these blooms, the water temperature was 27.5–32°C, salinity 37.1–37.2, dissolved inorganic nitrogen 2.42–2.88 μM DIN, phosphates 2.32–5.48 μM, with a corresponding condition of oxygen supersaturation of the waters (up to 170%). The highest concentrations of DIN were due to nitrates, or alternatively to ammonia, and phosphates in some cases reached values of ≈10 μM (2001, Fig. 1) with a slight decreasing trend through the years.

Fig. 1
figure 1

Spring-summer values of (a) Alexandrium taylorii densities, salinities, and (b) nutrients in Vulcano-St.1, Sicily. Years 2000–2003

Full size image

The water discolouration lasted 1–2 months, as confirmed again in 2004 and 2005 (data not shown), gradually disappearing in July or August–September (2001). The bloom persistence and discolouration was apparently favoured by optimal, summer weather conditions. The horizontal distribution of A. taylorii cells within the West Bay (Vulcano) was characterized in all cases by the assembling of cells nearshore, in the most sheltered part of the bay, as shown in Fig. 2, although in some cases a wind-driven advection of cells to NE of the bay was observed.

Fig. 2
figure 2

Horizontal distribution of A. taylorii cells in the West Bay, Vulcano. Surfer plot of 22 July 2001. Triangles (1–3) indicate routine sampling stations. Asterisks indicate additional cruise stations

Full size image

The percentage of temporary cysts versus motile cells of A. taylorii (Fig. 3), as determined in 2001 in water samples, showed an incidence of resting vegetative stages of 7–11% over the total A. taylorii population, with the highest percentages near the sediment. These values increased to 22% in late summer when the bloom was declining.

Fig. 3
figure 3

General morphology of motile cells (fluorescence) and temporary cysts (bright field). Scale bars = 4 μm (a) and 25 μm (b)

Full size image

The NJ phylogenetic tree of A. taylorii and other Alexandrium species based on the 5.8S rDNA-ITS regions is shown in Fig. 4. In respect to the other species of Alexandrium examined to date, A. taylorii appeared to diverge. Within the genus Alexandrium, the Mediterranean A. taylorii isolates formed a homogeneous clade, distinct from the others. These isolates shared identical nucleotide sequences.

Fig. 4
figure 4

Neighbor-Joining (NJ) tree inferred from the 5.8S rDNA and ITS sequences of Alexandrium species. Sequences of A. taylorii from this study are in bold. G. sanguineum was used as an outgroup. Numbers on the major nodes represent NJ (10000 pseudo-replicates) bootstrap values. Only bootstrap values ≥ than 50% are shown

Full size image

Species-specific PCR assays of fixed field samples are shown in Fig. 5. By using species-specific primers designed for A. taylorii, PCR amplifications of the 5.8S rDNA-ITS regions yielded products of 297 base pair size for the field samples containing A. taylorii cells. Species-specific PCR amplifications for A. taylorii revealed positive results for the presence of this species in the Vulcano phytoplankton population samples in agreement with the microscopic examination.

Fig. 5
figure 5

A. taylorii specific PCR assays of mixed phytoplankton population samples from different Mediterranean localities. (a) template plasmid containing A. taylorii 5.8S rDNA-ITS regions at different concentrations: lane 1, 10 ng; lane 2, 1 ng; lane 3, 0.1 ng. (b) Saronikos Gulf (Greece) field sample: lane 1, phytoplankton template DNA (1 ng); lane 2, template plasmid containing A. taylorii 5.8S rDNA-ITS regions (1 ng); lane 3, negative control (no template DNA). (c) St. Carles field sample, Catalan Sea (Spain): lane 1, phytoplankton template DNA (1 ng); lane 2, negative control (no template DNA); lane 3, template plasmid containing A. taylorii 5.8S rDNA-ITS regions (1 ng). (d) Vulcano West Bay field samples, Tyrrhenian Sea (Italy), 15th June 2003: lane 1, Station1 phytoplankton template DNA (1 ng); lane 2, Station 3 phytoplankton template DNA (1 ng); lane 3, Station2 phytoplankton template DNA (1 ng); lane 4, Station Capo Testa Grossa phytoplankton template DNA (1 ng); lane 5, template plasmid containing A. taylorii 5.8S rDNA-ITS regions (1 ng). (e) Vulcano field sample, West Bay Station1: lane 1, phytoplankton template DNA (10 ng), 16th June 2003; lane 2, phytoplankton template DNA (0.1 ng), 16th June 2003; lane 3, phytoplankton template DNA (0.01 ng), 16th June 2003; lane 4, phytoplankton template DNA (10 ng), 17th June 2003; lane 5, phytoplankton template DNA (0.1 ng), 17th June 2003; lane 6, phytoplankton template DNA (0.01 ng), 17th June 2003. M, size standards; arrow indicates reactions that produced a single species-specific PCR product

Full size image

Discussion

Without understimating the setbacks to the tourism industry, the ecological impact of A. taylorii blooms may be important considering the high levels of biomass produced, as in the case of the West Bay, Vulcano (up to 3 × 107 cells l−1) and coincident conditions of oxygen supersaturation of the waters, that may threaten marine life. Trophic trends in the Tyrrhenian target site indicate high amounts of nutrients linked to the increased anthropogenic activity in summer (see also Penna et al., 2001), although we observed an apparent shift of the marked eutrophic conditions towards a slighter eutrophy. This could be due to a recent improvement of the sewage treatment. Previous studies in this area suggested that nutrients play a relevant role in the development and maintenance of A. taylorii blooms, the high concentrations in summer of inorganic nitrogen and phosphorus being probably due to the hotel sewage, when the anthropogenic pressure increases in relation to the tourism (Penna et al., 2002). Then, climatic conditions such as high water temperature, calm sea and weak tide, seem to favour the bloom development. Garcés et al. (2002) suggested that under short-term environmental perturbations, for instance storms or swells, A. taylorii encysts to withstand the unfavourable conditions and restores the population after a few days of calm weather, thus functioning as a population stock. In addition, encystment would also reduce the population losses from the area, also through cell aggregation in clusters of cysts. This, together with the physical characteristics (relatively low water renewal) and favourable wind conditions, may favour population growth and persistence (Garcés et al., 1999, Basterretxea et al., 2005).

Many factors, natural or anthropogenic, are known to contribute to the spreading worldwide of HAB species. Such phenomena are amplified by human activities, such as the wrong management of the coastal area leading to increased amounts of nutrients in the waters and eutrophy, lack of seawage treatments, use of fertilizers in agriculture, the sea discharge of ship’s ballast water favouring the introduction of non-indigenous organisms, the so-called “alien species”. Cell dispersion can also be favoured by anthropogenic means, as plastics (Masò et al., 2003) or movement of recreational boats in summer. Therefore, areas previously free by harmful species can later become subject to HABs, if such species find in the new environment conditions favouring cell growth.

In the specific case of A. taylorii, we obtained further evidence for its spreading in the Mediterranean region, from the western side (the Catalan coast of Spain) to the eastern area (Greece).

Aiming at the improvement of the HAB detection tools, species-specific primers for A. taylorii were applied in this study in parallel to conventional techniques. This allowed the detection of the target organism also when it occurred as resting stage and the exact species identity was more difficult to be determined. The application of the PCR analysis to field seawater samples, evidenced the presence of A. taylorii in all mixed phytoplankton population samples collected in different sampling points in the Vulcano West Bay, with exception of the Station 3, where the cell abundance was often extremely low. Further, the species-specific primer application to the PCR analyses resulted also suitable for the Greek and Spanish field samples containing the target species. These results on the species-specific detection of A. taylorii by PCR assay were always in agreement with the microscopic examination.

The phylogenetic analyses of geographically distinct isolates of A. taylorii (Italy, Spain, Greece) indicated the existence of a unique, homogeneous Mediterranean population, constituted by isolates with identical ITS-5.8S rDNA sequences.

In this study, the confirmation of the specific identity through the coupling of microscopic and molecular analyses provided new insights into the biogeography of A. taylorii, showing the extension of its geographic range from the western Mediterranean to the eastern area. Future researches will give priority to the advancement in both control and prevention strategies as safeguard of the coastal marine environment and public health, and development of advanced technologies for the rapid detection and mitigation of HAB events through laboratory experiments and in situ application.