In fall 1997, the toxic cyanobacterium Microcystis aeruginosa was documented in Lake Sammamish (western Washington, U. S. A.) for the first time. Cyanobacterial activity and environmental conditions that may promote toxic cyanobacteria were investigated during summer and fall 1999. Development of toxic Microcystis was hypothesized to be due to runoff of nutrients from the watershed (external loading hypothesis) or from vertical migration of dormant cyanobacteria from the nutrient-rich sediments into the water column (cyanobacterial migration hypothesis). Microcystins were detected using an enzyme-linked immunosorbent assay during late August and early September 1999 despite low cyanobacterial abundance. Microcystin concentrations ranged between 0.19-3.8 mug l(-1) throughout the lake and at all depths with the exception of the boat launch where concentrations reached 43 mug l(-1). Comparison of the conditions associated with the toxic episodes in 1997 and 1999 indicate that Microcystis is associated with a stable water column, increased surface total phosphorus concentrations (> 10 mug l(-1)), surface temperatures greater than 22 degreesC, high total nitrogen to phosphorus ratios (> 30), and increased water column transparency (up to similar to5.5 m). Migration of the cyanobacteria, Microcystis and Anabaena, occurred in both the deep and shallow portions of the lake. Microcystis dominated (89-99%) the migrating cyanobacteria with greater migration from the shallow station. External loading of nutrients due to the large rainfall preceding the 1997 toxic episode may have provided the nutrients needed to fuel that bloom. However, toxic Microcystis occurred in 1999 despite the lack of rain and subsequent external runoff. The migration of Microcystis from the nutrient-rich sediments may have been the inoculum for the toxic population detected in 1999.
