The summer phytoplankton population in the Arcos reservoir (Spain), an oligotrophic/mesotrophic waterbody, was dominated by the cyanobacterium Aphanizomenon ovalisporum, accompanied by Planktothrix cf. agardhii and smaller proportions of other cyanobacterial species. The bloom remained for approximately 1 month in the summer of 2004. The Arcos reservoir had moderate conductivity, up to 1756 mS cm À1, and relatively high nutrient concentrations. Planktonic biomass collected during the bloom period was found to be toxic by the Artemia salina bioassay. Analytical procedures based on high-performance liquid chromatography (HPLC) indicated the absence of microcystins and anatoxin-a but clearly demonstrated the presence of cylindrospermopsin (CYN), suggesting that CYN from Aphanizomenon ovalisporum was the compound responsible for the toxicity in the bioassay. Measurements indicate that up to 9.4 mg CYN l À1 were present in the sestonic fraction. This is the first report of high concentrations of this cyanotoxin in European waters and the first known instance of a massive bloom of the cyanobacterium Aphanizomenon ovalisporum in Spain.

The occurrence and concentration of microcystins were analysed in sestonic samples from seven water reservoirs in the Madrid region (Spain) between July and November in two consecutive years (2002 and 2003). The data collected indicate that microcystins were present on several occasions in both years in all the reservoirs studied. The months of maximum risk for cyanotoxin occurrence were July, September and October, when Microcystis typically dominated the phytoplankton community. Four of the seven reservoirs exhibited conspicuous blooms, three of which were toxic. Of the samples in 2003, 45% and 70%, respectively, contained microcystins (Mc). In four of the reservoirs, microcystin concentrations were higher than the WHO recommended limit for drinking water (1 mg Mc l . From the data obtained we can conclude that M. aeruginosa is the main producer of microcystins in freshwaters from the Madrid region.

The presence of anatoxin-a (ANA) in Mediterranean region freshwaters has been reported only recently, and this work presents the first survey for ANA on a national scale in such waters. Fourteen reservoirs were sampled over a distance of 550 km from northeast to southwest Spain. Genera of cyanobacteria with ANA-producing members in other countries were present in all of the Spanish reservoirs, but this toxin was detected in only one reservoir. The maximum ANA concentration detected was 0.31 lg AE L )1 , and the most probable producer was Anabaena flos-aquae (Lyngb.) Bréb. ex Bornet et Flahault. These findings suggest that in spite of the abundant cyanobacterial populations, ANA may be of lower occurrence in the Spanish reservoirs investigated than in other European freshwaters, and different strains of potential ANA producers occur in different ecoregions.