Beginning in the 1990's, massive avian and fish epornitics have occurred in the Salton Sea and over 200,000 eared grebes have died [1]. The largest single epizootic occurred in 1992 when an estimated 155,000 birds, primarily eared grebes (Podiceps nigricollis), died from an undiagnosed cause. The cause of these massive grebe epornitics remains unknown, although several diseases such as avian botulism and avian cholera have been diagnosed [2]. Algal biotoxins, especially those produced by dinoflagellates, have also been listed as a possible contributing cause [3], but none of the known dinoflagellate toxins has been identified to date. Preliminary results from analyses of phytoplankton samples and eared grebe tissues collected at the sea in the early 1990's identified microcystins produced by cyanobacteria [[4, 5], internal reports to USGS]. Eared grebes winter on the Salton Sea, therefore the epornitic usually occurs annually during the late winter and early spring. Eared-grebe tissues collected from bird mortalities in the Salton Sea during 1992 – 94 had the cyanotoxin – microcystin – in concentrations high enough to cause acute toxicity. Enzyme Linked Immunosorbent Assay (ELISA) measured values of microcystins, in 25 samples of liver, gizzard and upper gastrointestinal tract showed levels of microcystin in liver as high as 700 ng/g. This is well above the known levels of microcystin in liver that could cause acute lethality (about 200 ng/g) [6]. Forty-nine water samples of phytoplankton, provided by the US Fish and Wildlife Service, collected in 1995–96, from the Salton Sea, contained levels of microcystins that ranged from negative to 2 ppb [Carmichael, internal reports to the USGS]. These levels are low and not likely to cause acute toxicity. However the toxin was associated with an organism smaller than 5 microns. These results suggest a small planktonic cell or picoplankton (i.e. the fresh/brackish/marine cyanobacterium Synechococcus) may be present in the Sea that produces microcystin. Small planktonic cyanobacteria are known to produce microcystins [7]. This background data formed the basis for the hypothesis of this project: Microcystins contribute to the eared grebe mortalities on the Salton Sea and that a significant source of organism(s) producing microcystins is to be found in the picoplankton.

The Salton Sea has salinities that vary from freshwater/brackish water at the major river outlets to hypersaline conditions in the sea proper, extreme daily summer temperatures (>40°C), and high nutrient loading (eutrophic conditions) from rivers and agricultural drainages which encourages algal blooms throughout the year [8]. These algal blooms could produce biotoxins that contribute to the massive fish and avian mortalities at the Salton Sea. In addition very little is known about algal species that survive under the existing conditions of Salton Sea. The combination of extreme summer temperatures, various saline water conditions, and nutrient loading provide a unique opportunity to investigate algal species in the Salton Sea area. Because very little is known about these algal species, and their biotoxins, there is high potential for new species and biotoxin identification that may contribute to an understanding of the high fish and avian mortalities. The purpose of this study is to collect water samples for the identification of algal species and biotoxins (cyanotoxins only), and to determine the presence of cyanotoxins in eared grebe tissues.

Microcystins have been characterized from all of the most cosmopolitan cyanobacteria genera including Anabaena, Microcystis, Oscillatoria, Planktothrix, Nostoc, Anabaenopsis and Hapalosiphon while nodularin is found in the brackish water species Nodularia spumigena. An alkaloid hepatotoxin cylindrospermopsin is produced by Cylindrospermopsis raciborskii, Umezakia natans, Raphidiopsis curvata and Aphanizomenon ovalisporum [10].

Three groups of cyanobacterial neurotoxins are known: (i) anatoxin-a and homoanatoxin-a, which mimic the effect of acetylcholine, (ii) anatoxin-a(s), which is an anticholinesterase and (iii) saxitoxins, which block nerve cell sodium channels. Anatoxin-a has been found in Anabaena, Oscillatoria and Aphanizomenon, homoanatoxin-a in Oscillatoria, anatoxin-a(s) in Anabaena, and saxitoxins in Aphanizomenon, Anabaena, Lyngbya and Cylindrospermopsis. Sixteen confirmed saxitoxins from cyanobacterial samples have been reported, some of which (the decarbamoyl-gonyautoxins) may be chemical break-down products in some species.

In marine waters, benthic cyanobacteria such as Lyngbya, Oscillatoria and Schizothrix may produce toxins causing severe dermatitis among swimmers in contact with cyanobacteria. Aplysiatoxin and debromoaplysiatoxins are protein kinase C activators and potent tumor promoters. Lyngbyatoxin A exposure has caused severe oral and gastrointestinal inflammations in humans. Trichodesmium sp. occurring in tropical seas are known to contain an as yet uncharacterized neurotoxin. Though comparatively poorly studied, cell wall components, particularly lipopolysaccharide endotoxins (LPS), from cyanobacteria may contribute to human health problems associated with exposure to mass occurrences of cyanobacteria. Cyanobacteria are also known to produce several other bioactive compounds, some of which are of medical interest, as well as compounds toxic to other cyanobacteria, bacteria, algae and zooplankton.

Selected phytoplankton samples from almost all shipments were plated onto agar plus medium and cultured for isolation of cyanobacteria that may be cyanotoxin producers. During the course of this project 100 samples were plated and approximately 150 unialgal cultures made. The isolates were dominated by the filamentous Oscillatoria and the picoplanktonic Synechococcus. Because the duration of the study was relatively short cultures tested were those that responded quickly to the particular culture conditions in our facilities. The culture types found in our study were again similar to those obtained in a study by Wood et al. [15]. Of these cultures 50 of the better growing isolates were processed and tested by ELISA for microcystin. Thirty four of the 50 cultured isolates were positive for microcystins. Values were low and ranged from detectable (0.147 μg/L) to 1 μg/L. Four of these 34 cultures were identified as Synechococcus and all four were positive for microcystins by ELISA and LC/MS. Only strain SS-1, of these four strains, was identified as Synechococcus by PCR.

LC/MS analyses of selected cultures was also used to confirm the presence of microcystin production. Since the picoplankton have potential to be a significant source of microcystin year round in the Salton Sea the presence and type of microcystin in these cultures was analyzed. Figure 1 and 2 shows that microcystin-YR and LR, respectfully, were present in strain SS-1 of Synechococcus while strain O-1 of Oscillatoria contained microcystin-LR (Fig. 3).

Cyanobacterial 16S rRNA gene sequences available from GenBank and those examined in the study were aligned using the multiple sequence alignment tools in CLUSTAL W version 1.7 [16]. This was followed by conversion to a distance matrix. The distance matrix was used to reconstruct a phylogenetic tree (Fig. 4) by the neighbor-joining (NJ) algorithm of CLUSTAL W version 1.7, with multiple substitutions corrected and positions with gaps excluded. The seed number for random number generation and the number of bootstrap trials were set to 111 and 1000, respectively.

Access number. Partial 16S rRNA gene sequences of the strains used in this study can be obtained in Genbank using access number: DQ455751. Results of the sample culturing, microcystin analyses and generation of the dendrogram show that microcystin producing Synechococcus was a member of the Salton Sea phytoplankton during the course of this study. From the dendrogram results and comparison with other GenBank gene sequences we found that the Salton Sea Synechococcus strain tested is closer to marine strains of Synechococcus than to freshwater strains of this genus. This is significant since it shows that marine cyanobacteria can produce Microcystins, a finding not demonstrated until this study. This has implications for the possible presence of microcystins in other marine environments and may make microcystin a marine HAB as well as a freshwater HAB toxin.

Microcystin types in selected liver samples were analyzed by LC/MS. Mass peaks indicative for microcystin-LR and YR were identified although the sensitivity was not low enough to confirm their presence by LC/MS-MS.

The cyanotoxin group microcystins should be considered as a possible contributing cause of grebe morbidities and mortalities in the Salton Sea and efforts to remediate the Salton Sea water quality should take this into account. Our study did not include controlled exposure studies to grebes and this is an obvious need for future research in order to determine the role of cyanotoxins in these toxicities. We also did not investigate the possible routes by which microcystins vector into grebe livers and intestines-another important study that should be undertaken. Finally the finding that all cultured strains of the picoplankton Synechococcus produces microcystin is important in itself but that in addition the PCR based genetics places it within the marine Synechococcus cluster is even more significant. Clearly more work is needed to extend this finding and clarify if marine Synechococcus producing microcystin are widespread in the Salton Sea and in other marine systems. Salinity levels may be a significant factor in selecting for the microcystin producers. In a study on the Swan River Australia, Orr et al. [17] demonstrated that laboratory cultures of microcystin producing Microcystis aeruginosa were more tolerant to high salt concentrations and were preferentially selected for, as salt levels were increased in the culture media.

Efforts to reduce salinity and improve water quality in the Salton Sea should consider that the genera of cyanobacteria currently in the sea are not high producers of microcystins. If salinities are lowered without also managing levels of key nutrients such as nitrogen and phosphorus other genera of cyanobacteria that produce higher levels (acute lethal) of microcystins may be selected for. These genera could include Microcystis and Anabaena. Waterblooms of these genera are typically more intense and would present a higher risk from cyanotoxin (microcystins, anatoxins, cylindrospermopsin and saxitoxins) exposures.

An investigation to determine the role of cyanotoxins in grebe mortalities and morbidities on the Salton Sea was started in November 1999 and ended in April 2001. Water sampling dates varied but samples were received from every month except May, June and July of 2000. Twenty shipments were received over this 18-month period. Fifteen of the shipments contained water and phytoplankton samples while six contained grebe tissue samples (one shipment contained both water and tissue samples). Water samples containing cyanobacteria at numbers giving a visible color to the water samples (approx 10⁵ to 10⁷ cells/ml) were lyophilized and analyzed for microcystins by ELISA. Approximately 85% of 247 samples were positive for microcystins. Concentrations of microcystins were typically less than 100 μg/gdw. This concentration is generally less than the needed to cause acute lethal toxicity from ingestion of bloom samples by mammals.

Throughout the sampling period the majority of water samples were dominated by the filamentous genus Oscillatoria and the picoplanktic genus Synechococcus. Isolation, culture and ELISA testing for microcystin of 50 strain isolates found that microcystins were produced by all strains – although at low levels (<1 μg/gdw). The genera producing measurable levels of microcystin included mainly Synechococcus and Oscillatoria. Other positive microcystin genera included Gloeocapsa, Phormidium, Aphanothece and Lyngbya.

PCR 16S rRNA analyses was done on one strain of Synechococcus, SS-1. Results confirmed that this strain was Synechococcus and comparison with a marine cluster of Synechococcus supports a conclusion that this Synechococcus strain is more similar to marine than to freshwater members of this genus. All four Synechococcus isolates were shown by ELISA and LC/MS to produce the microcystins-MCYST-LR and MCYST-YR.

Microcystins were detected by ELISA in the majority of liver and intestine samples from grebes collected due to morbidity or mortality. The concentrations of microcystins were not always high enough to account for acute lethal toxicity (< 50 ng/g) but a significant number of liver samples did contain microcystin levels high enough to account for acute lethal toxicities (>50 ng/g).

Although the concentration of microcystins in the water samples and in strain isolates of cyanobacteria were not high enough to account for acute lethal toxicity, levels in grebe livers often did suggest acute or acute lethal toxicity could occur. This suggests that microcystins accumulated, by an as yet unknown vector(s), in grebe tissues to levels that could be lethal.