Background


Two historically abundant, endemic fishes inhabiting the Upper Klamath Basin, Oregon, the Lost River sucker (Deltistes luxatus) and the shortnose sucker (Chasmistes brevirostris), were listed as endangered under the Federal Endangered Species Act in 1988 by the U.S. Fish and Wildlife Service following sharp declines in abundance, range reductions, and evidence that recruitment into the spawning population had decreased from historical levels (U.S. Fish and Wildlife Service, 1993; National Research Council, 2004). In addition to overharvest, habitat alteration, and the presence of nonnative fish species (U.S. Fish and Wildlife Service, 2002), the observed decline in these populations has been attributed to water quality degradation resulting from massive cyanobacterial blooms (Williams, 1988; Buettner and Scoppettone, 1990; Scoppettone and Vinyard, 1991; Perkins and others, 2000), which occur annually from June through October in Upper Klamath Lake. High rates of photosynthesis during bloom periods elevate lake water pH (9.5 and higher; Kann and Smith, 1999), and decomposition during bloom declines increases concentrations of un-ionized ammonia (> 0.5 mg/L) and dissolved nutrients (Hoilman and others, 2008; Lindenberg and others, 2009). This seasonal cycle of cyanobacterial growth and decline also causes oxygen concentrations to fluctuate from supersaturation to near anoxia (dissolved oxygen concentrations less than 1 mg/L; Wood and others, 2006), depending on the magnitude of the bloom decline (Kann and Welch, 2005). 


In addition to contributing to poor water quality conditions through photosynthetic activity and bloom decomposition, several genera of bloom-forming, freshwater cyanobacteria also produce secondary metabolites that are toxic to a wide range of aquatic organisms, including fish (reviewed in Falconer, 1999). Hepatotoxic microcystins, first isolated from Microcystis aeruginosa (Krishnamurthy and others, 1986), are the most abundant and frequently occurring of these compounds (Carmichael and others, 1986; Yu, 1989; World Health Organization, 2006; Erdner and others, 2008) and have been implicated in human, livestock, domestic animal, and wildlife illness and death in more than 20 countries worldwide and in at least 36 U.S. states, including Oregon (Carmichael, 1994; Sivonen and Jones, 1999; Graham and others, 2009). More than 80 microcystin variants have been identified (Welker and Von Döhren, 2006), which vary in toxicity by an order of magnitude (Sivonen and Jones, 1999). Microcystins are nonribosomally synthesized cyclic heptapeptides produced by strains (subspecies) of many cyanobacterial genera, including Microcystis (Krishnamurthy and others, 1986), Anabaena (Harada and others, 1991), Oscillatoria (Meriluoto and others, 1989), Gloeotrichia (Carey and others, 2007), and Pseudanabaena (Oudra and others, 2002), all of which have been identified in water samples from Upper Klamath and Agency Lakes (Kann, 1997; Kann and Asarian, Aquatic Ecosystems Sciences LLC, unpub. data, 2008 and 2009; B.H. Rosen, U.S. Geological Survey, unpub. data, 2009; Eldridge and others, unpub. data, 2010). Blooms containing these genera may contain both toxic (with microcystin synthetase, mcy, gene clusters) and nontoxic (without the mcy genes) strains, which can not be distinguished from each other by microscopy. The nitrogen (N₂)-fixing (diazotrophic, ability to convert nitrogen gas to ammonia) Aphanizomenon flos-aquae generally comprises more than 90 percent of the cyanobacteria biovolume (Kann, 1997) in Upper Klamath Lake during periods of high bloom density. Members of the Aphanizomenon genus have been shown to produce cylindrospermopsins and several neurotoxins in laboratory cultures (Carmichael, 1997; Preussel and others, 2006; Graham and others, 2008), but they have not been shown to produce microcystins, and there is currently no evidence that A. flos-aquae produces toxins in the Upper Klamath or Agency Lakes (Carmichael, 2000). Therefore, although A. flos-aquae is the dominant cyanobacterium in this area, it is not likely to be the microcystin producer here. 


The colony-forming, unicellular M. aeruginosa has been directly linked to microcystin occurrence in Upper Klamath Lake and in downstream reservoirs and was first reported in water samples from Upper Klamath Lake in July 1996 (Jacoby and Kann, 2007). The Oregon Department of Health and Health Canada used high performance liquid chromatography (HPLC) and enzyme-linked immunosorbent assays (ELISA) to show the presence of microcystins in dietary supplements produced from A. flos-aquae collected exclusively from Upper Klamath Lake (Gilroy and others, 2000; Lawrence and others, 2001). In a later study, Saker and others (2007) used multiplex polymerase chain reaction (PCR) to simultaneously detect a fragment of the microcystin-synthetase gene cluster (mcyA) and the partial 16S rRNA gene sequence specific to Microcystis in these supplements. The results of this study showed that M. aeruginosa was the most abundant microcystin-producer in the dietary supplements tested (Planktothrix sp. also was identified as a microcystin-producer) and supported previous reports that Microcystis (mostly M. aeruginosa) co-occurs with A. flos-aquae in Upper Klamath Lake (Carmichael, 2000). Microcystis is not capable of nitrogen fixation, unlike A. flos-aquae, and is dependent on ammonia and other nitrogen sources for growth. Therefore, the availability of nitrogen may contribute to the occurrence of Microcystis colonies and microcystins in Upper Klamath Lake.


In 2007, the U.S. Geological Survey continued monitoring water quality throughout Upper Klamath Lake (a project that began in 2002; Wood and others, 2006) and began a preliminary study to determine the seasonal and spatial occurrence of cyanotoxins in the lake and to determine statistical relations between microcystin concentrations and other water quality variables. In addition, the U.S. Geological Survey conducted parallel studies to determine the pathological effects of cyanotoxins on juvenile Lost River and shortnose suckers in Upper Klamath Lake (VanderKooi and others, 2010). Between July and September 2007, histopathology consistent with microcystin exposure (Malbrouck and Kestemont, 2006) was identified in 49 percent (n = 47) of age-0 suckers collected from 11 shoreline locations throughout the lake (VanderKooi and others, 2010). Multiple organ necrosis was identified in a greater percentage of individuals captured in the northern region of the lake than in the southern region. In summer 2008 (July–September), age-0 suckers (n = 103) collected from five geographic areas in the lake were examined for histopathology. Evidence for organ damage was absent in two areas and observed in less than 20 percent of fish collected in other areas. As in the previous year, fish captured in the northern region of the lake exhibited the highest occurrence of organ damage. These regional differences may be due to variation in the dose, exposure duration, and time since exposure to toxins in the lake. A gut analysis of juvenile fish collected in 2008 (n = 45) showed that all suckers observed had ingested chironomid (midge) larvae which, in turn, appeared to contain colonies of M. aeruginosa and filaments of A. flos-aquae in their digestive tracts (M. aeruginosa was found in all ingested chironomid larvae and A. flos-aquae was found in approximately 20 percent of the larvae; B.H. Rosen, U.S. Geological Survey, unpub. data, 2008). The juveniles examined had completed the ontogenetic (developmental) shift to benthic feeding, so these chironomid larvae were most likely ingested from lake sediments. Histological examinations of these fish revealed numerous gastro-intestinal lesions consistent with microcystin exposure (Malbrouck and Kestemont, 2006), which were observed regardless of whether liver necrosis also was present (VanderKooi and others, 2010). The observed histopathology did not indicate a bacterial or parasitic etiology, although infection from an undetected virus or chronic effects of ammonia toxicity could not be ruled out. These fish most likely were exposed to microcystins by ingestion and not by absorption of these compounds through their gills.


Cyanotoxin concentrations may vary widely between sites and depths within a single lake or during a sampling season and between years. Such spatiotemporal variation generally results from indirect environmental influences on the presence and abundance of toxigenic strains (Kurmayer and others, 2003; Kurmayer and Christiansen, 2009), and from the direct effects of the environment on cellular rates of microcystin production, which are determined by cyanobacterial community structure, growth stage, and nutrient dynamics (Reynolds, 1998; Jacoby and others, 2000). Changes in cellular microcystin production rates and the abundance of toxigenic M. aeruginosa strains have been documented on a seasonal basis in reservoirs downstream of Upper Klamath Lake (Kann and Corum, 2009, 2010; Bozarth and others, 2010). Determining which environmental factors, direct or indirect, have the greatest effect on naturally occurring communities of mixed cyanobacteria populations is difficult. However, studies have shown that microcystin production is favored by factors that regulate population growth (Orr and Jones, 1998), including limitation by phosphorus (Trimbee and Prepas, 1987; Jacoby and others, 2000), nitrogen (Jones and Jones, 2002; Downing and others, 2005; Moisander and others, 2009), and light (Wicks and Thiel, 1990; Wiedner and others, 2003), among others. Although microcystins are contained primarily within the cells that produce them, dissolved microcystins can be detected during early stages of bloom development (Sedmak and Elersek, 2006) because cellular excretion and lysis occur continually throughout the growth cycle (Hughes and others, 1958). However, the highest concentrations of dissolved microcystins commonly are detected during bloom senescence and decomposition (Park and others, 1998). Therefore, in the present study, an elevated concentration of dissolved microcystins was considered indicative of a decline in the M. aeruginosa population.


Study Area


Upper Klamath Lake, located within the Klamath Graben structural valley at the base of the Cascade Mountains (eastern slope) in south-central Oregon, is a naturally occurring, large and shallow water body with a surface area of 232 km²and an average depth of 2.8 m (fig. 1). More than 90 percent of the lake is less than 4 m deep, but the western shoreline between Eagle Ridge and Buck Island (crossing the entrance to Howard Bay) reaches depths of up to 15 m. The Williamson River, which enters the lake from the north, contributes, on average, approximately 46 percent of the total inflowing water to the lake annually (Johnson, 1985). The lake’s drainage basin is 9,415 km² and is composed of phosphorus-rich, volcanic soils. The Upper Klamath Lake system has been eutrophic since at least the mid-1800s (the earliest known records), but major changes in land use and hydrology of the watershed and lake over the past century, including forest clear-cutting, cattle grazing in upstream flood plains, degradation of riparian corridors, and the conversion of neighboring wetlands to flood-irrigated pasture and agricultural fields (Klamath Tribes, 1994), have intensified eutrophication and increased the biomass of cyanobacteria (primarily of a single species, A. flos-aquae) that bloom during the summer and autumn (Bortelson and Fretwell, 1993; Bradbury and others, 2004; Eilers and others, 2004). Although the lake occurs naturally, it is artificially controlled. Upper Klamath Lake has been the primary water source for the Klamath Project, which has supplied water to agricultural areas within the Upper Klamath Basin since the completion of the Link River Dam at the southern outlet of the lake in 1921 (Bureau of Reclamation, 2000). This has allowed regulation of lake water levels and volume that has resulted in more extreme fluctuations in annual surface water elevation (as much as 1 m between early spring and late summer), although, in recent years, the overall maximum and minimum elevations have been similar to pre-dam conditions.


Purpose and Scope


The purpose of this report is to characterize and compare cyanotoxin concentrations in water samples collected from Upper Klamath Lake during the 2007 through 2009 field seasons. The physical and chemical characteristics of study sites are summarized to determine if relations exist among the spatial and temporal distributions of microcystins and other environmental variables in Upper Klamath Lake, including dissolved and total nutrients; chlorophyll a; particulate carbon, nitrogen, and phosphorus; water temperature; pH; dissolved oxygen; wind speed; site depth; and water column stability (relative thermal resistance to mixing, RTRM). Nutrient, chlorophyll a, continuous monitor, and meteorological data obtained through the long-term monitoring program are included here to provide context for microcystin analyses, but are described in Kannarr and others (2010). Results of this study will contribute to understanding the environmental influences on the occurrence of cyanotoxins, specifically microcystins, in Upper Klamath Lake, which is critical for effective lake management and understanding the causes of apparent high mortality rates in juvenile Lost River and shortnose suckers.


First posted May 30, 2012