research by a large group of collaborators. Their research continued until the early 1940s.

Following a hiatus, A. D. Hasler led a second major period of research with an emphasis on experimental studies (Hasler, 1964). This research, involving several large-scale experiments, was conducted around Madison and in the Northern Highland area. Particularly important was the neutralization of a naturally acidic lake by using a now classic design in which the basin was divided into similar manipulated and unmanipulated sections (Hasler, 1964). Researchers involved in several more recent whole-ecosystem manipulations trace their inspiration to these early large-scale experiments by Hasler and his coworkers (e.g., Likens, 1985c; Schindler, 1988; Carpenter and Kitchell, 1994).

Research in Wisconsin continues with major ongoing efforts in the Madison area, primarily on Lake Mendota, and in the Northern Highland. By evaluating current conditions relative to the range of conditions in the past, recent work on Mendota has illustrated the utility of long-term data in understanding system processes (Brock, 1985; Kitchell, 1992). This work also provides another example of the potential gains when nontraditional fields are combined in research projects, in this case fisheries biology and limnology (Vanni et al., 1990). Substantial shifts in the lake's water clarity were ultimately attributable to an unusual die-off of one of its dominant fish. Lake Mendota has also been the site of a major joint collaboration between lake managers and basic scientists (Kitchell, 1992).

Other efforts in Wisconsin include a substantial expansion of projects at the Trout Lake Station in the Northern Highland. The station has served as the staging facility for one of the sites in the Long-Term Ecological Research program (Magnuson et al., 1984) and for a whole-lake acidification experiment (Brezonik et al., 1993). LTER work here has provided another example of the utility of combining disciplines, in this case ground water geology and limnology, by demonstrating how a relatively small portion of the lake's hydrologic inputs could play a critical role in its overall nutrient budget because of the high concentration of minerals in ground water (Hurley et al., 1985).

Regional evaluations of long-term records illustrate how variability, often perceived as an impediment to system understanding, can provide useful insights into the processes controlling system features (Kratz et al., 1987). Related projects have further demonstrated how evaluating patterns in data collected during the same period from sets of lakes within a region can indicate how different system properties are controlled by forces that operate on fundamentally different spatial scales (Kratz et al., 1991). Additional northern work has continued, using a large-scale experimental approach to evaluate the effects of lake acidification (Brezonik et al., 1993) and the role of food web structure on basic lake processes (Carpenter and Kitchell, 1993). The latter project has been conducted at