Paul and Peter Lakes: A Liming Experiment Revisited

From 1951-1976, Peter Lake was limed periodically while neighboring Paul Lake was used as a reference ecosystem. We investigated the persistence and variability of the response of Peter Lake to liming by collecting all available historical data on the lakes and by monitoring the limnological properties of both lakes weekly from June-August, 1984. Physical and chemical changes in limed Peter Lake included increases in pH, alkalinity, dissolved inorganic carbon (DIC) concentration, transparency, oxygen content and summer heat content. These changes occurred rapidly in 1951 and have persisted with little variability until 1984. Several differences in algal and zooplankton community composition and dynamics were associated with the physical and chemical changes. Historical data and our observations indicate that the planktonic community structure of Peter Lake has been more variable than that of unlimed Paul Lake. INTRODUCTION In 1951, Arthur D. Hasler began the first whole-lake ecosystem experiment to use a "control" or reference ecosystem (Likens, 1985). Hasler divided the two basins of an hourglass-shaped lake with an earthen dike. He used rotenone to remove the native fish from the lakes and then stocked them with trout. The upstream basin, Paul Lake, was used as a reference ecosystem to determine the effects of lime additions on fish productivity in downsteam Peter Lake. Results of the study were summarized in several publications (Johnson and Hasler, 1954; Stross and Hasler, 1960; Stross et al., 1961). Periodic liming of Peter Lake continued until 1976. According to Likens (1985), this pioneering experiment was a model for more recent whole ecosystem experiments at the Hubbard Brook Experimental Forest and the Canadian Experimental Lakes Area. Liming is one management technique used to mitigate the adverse effects of acid precipitation in lakes (Bengtssan and Dickson, 1980; Driscoll et al., 1982; Fraser and Britt, 1983). Consequently, there have been many recent studies investigating the effects of lime additions on aquatic ecosystems. Most of these studies have lasted only a few years. The experiment in Paul and Peter lakes is the longest liming project known to us and, therefore, provides an interesting opportunity to study the long-term effects of liming. This paper has two goals. First, we collected and reviewed the historical data on these lakes, much of which was unpublished, to examine them for long-term limnological trends. Second, we analyzed the limnological properties of Paul and Peter lakes in 1984, 8 years after the last liming, to evaluate any persisting differences between the lakes that might be due to the liming of Peter Lake. METHODS Peter and Paul lakes are located in section 36, T45N R42W, Gogebic Co., Michigan. They are seepage kettle lakes surrounded on three sides by moraine ridges. A connecting culvert allows water to flow from Paul Lake into Peter Lake when water levels are high. Surface area of Peter Lake is 2.4 ha, and its maximum depth is 19.3 m with an average depth of 8.3 m. Surface area of Paul Lake is 1.2 ha with a maximum depth of 12.2 m and an average depth of 5.0 m. Both lakes have some bog mat development. In 1984, both lakes contained similar densities of largemouth bass (Micropterus salmoides) and no other vertebrate planktivores (J. R. Hodgson, D. M. Lodge, J. F Kitchell, pers. comm.). Invertebrate planktivores include Chaoborus punctipennis and C. flavicans (von Ende, 1979). The lakes were sampled weekly from 6 June to 22 August 1984. Permanent sam-