Surface samples of the 2007 Microcystis bloom occurring in Copco Reservoir on the Klamath River inNorthern California were analyzed genetically by sequencing clone libraries made with amplicons at three loci: the internal transcribed spacer of the rRNA operon (ITS), cpcBA, and mcyA. Samples were taken between June and October, during which time two cell count peaks occurred, in mid-July and early September. The ITS and cpcBA loci could be classified into four or five allele groups, which provided a convenient means for describing the Microcystis population and its changes over time. Each group was numerically dominated by a single, highly represented sequence. Other members of each group varied by changes at 1 to 3 nucleotide positions, while groups were separated by up to 30 nucleotide differences. As deduced by a partial sampling of the clone libraries, there were marked population turnovers during the season, indicated by changes in allele composition at both the ITS and cpcBA loci. Different ITS and cpcBA genotypes appeared to be dominant at the two population peaks. Toxicity (amount of microcystin per cell) and toxigenic potential (mcyB copy number) were lower during the second peak, and the mcyB copy number fell further as the bloom declined.Toxic freshwater cyanobacterial blooms, commonly caused by Microcystis, are of current concern in many parts of the world because of their effects on drinking water, water-based recreation, and watershed ecology (5, 7). Microcystis cells are able to produce microcystin, a nonribosomally synthesized cyclic heptapeptide hepatotoxin with potent inhibitory activity against mammalian protein phosphatases (27) whose synthesis is directed by the 55-kb mcy gene cluster (25). The Microcystis genus exhibits worldwide occurrence, although the extent of genetic differentiation between or within geographical regions is currently uncertain due to a relatively sparse database, in spite of a growing number of studies (1,2,9,11,26,28,29).Only a few studies to date have used gene-specific tools to investigate the changes in the Microcystis population structure throughout the development of a bloom season. In some instances, there has been little indication of major population changes. Thus, the proportion of toxigenic (mcyB ϩ ) Microcystis was stable over the course of two consecutive bloom seasons in Lake Wannsee (Berlin, Germany) (17). The internal transcribed spacer of the rRNA operon (ITS) genotype, as assessed by denaturing gradient gel electrophoresis (DGGE) and sequencing, was also stable in Lake Volkerak (Netherlands) during 2001 (15). In contrast, studies of other lakes have observed changes in the Microcystis genotypes and in the proportion of potentially toxigenic cells during a bloom season (3,15,21,31,32). A better understanding of the population changes that occur during the development of toxic blooms is important in understanding their ecology and in assessing whether it might be feasible to manage Microcystis blooms in order to minimize toxicity.