In this request for an LTREB grant renewal the principal investigators describe results of previously supported multi-year whole-lake food web experiment at Castle Lake, California. The investigators then outline a design for continued whole-lake manipulations. The project incorporates a continuous record of summer primary production dynamics dating back to 1959. The investigators manipulated the Castle Lake food web by varying rainbow trout (Oncorhynchus mykiss) stocking rates from zero fish per year for a three-year period (1989-1991) to 30,000 trout per year ( or three times background rates) for a following three-year period (1992-1994). Using a three- variable model they explained 77% of the interannual variability in Castle Lake primary production. This was regulated by negative relationships with total precipitation and winter snowfall and a positive one-year time lag. Whole-lake food web manipulations produced perturbations in mean annual primary production on a scale comparable to those seen for the climatic variables. During the drought, 1989-94, mean annual primar y production values were at or higher than the highest values seen since 1959. While trophic interactions among the microbial components of this and other lake food webs have been studied by other researchers, but direct comparisons of microbial dynamics in response to whole-lake climatic and food web processes have not been made. Continued LTREB support will allow the investigators to complete Phase II of their investigation (30,000 trout stocked per year), and conduct a new cycle of low and high trout stocking rate manipulations. They also will monitor microbial components of the food web (bacteria, picoplankton, mixotrophic phytoplankton, heterotrophic flagellates, and ciliates) into future whole-lake monitoring, food web manipulations. The proposed investigations will span a variety of spatial and temporal scales that will enhance understanding of whole-lake food web and microbial dynamics through longer term climate changes.
