Temporal coherence in the limnology of a suite of lakes in Wisconsin, U.S.A
1. Temporal coherence between pairs of lakes over 7 years was measured for thirty-seven limnological variables in seven lakes at the North Temperate Lakes Long Term Ecological Research site in Northern Wisconsin. This analysis tested. First, whether lakes more similar in exposure to the atmosphere were more temporally coherent than lakes which differed more in exposure and, second, whether temporal coherence in lakes progressively decreased from variables more directly influenced by climatic factors such as thermal and hydrologic properties, to those chemical and biological properties which may be less directly influenced by climatic factors. 2. The lakes were a heterogeneous set in respect to exposure to climatic factors as estimated by the ratio of‘lake area/mean depth’and by other morphometric features, and they also differed in their position in the landscape, fertility and fish assemblages. Limnological variables formed a progression from those expected to respond directly to climatic factors to those which would not. They ranged from water level and temperatures to chemical variables such as pH, calcium concentrations and total dissolved phosphorus to biological variables such as chlorophyll concentrations, invertebrate and fish abundances. 3. Coherence was estimated by the correlation between lake pairs for each of the different variables. The mean correlation and the percentage of strong correlations were calculated for each lake pair across all variables, and for each variable across all lake pairs, and both measures of coherence gave similar answers to the questions posed above. 4. Temporal coherence between takes was higher for lakes similar in their exposure to climatic factors; mean correlation (r̄) being +0.3 to +0.7 for these lakes and \textless+0.3 for lakes not similar in exposure. None of the lake pairs had high coherence across all variables. 5. Temporal coherence between lakes was greater for limnological variables directly influenced by climatic factors than for variables either indirectly affected by climate or complexly influenced by other types of factors. Water-level variables had a coherence near 1, r̄=0.9. All biological variables had low coherence, some near r̄=0.0. Chemical variables more likely to be influenced directly by climatic factors appeared to be more coherent than those more influenced by hydrology or biology. Most silica and phosphorus variables had coherences less than r̄=0.15. 6. Coherence was not as strongly related to similarity in landscape position as it was to similarity in exposure to climatic factors, and was not jelated to the proximity of the lake pairs or to their similarity in fertility. 7. A conceptual model was presented to explain how climatic signals are filtered by the lake’s exposure to climatic factors and by terrestrially mediated and in-lakc processes to reduce the coherence of lake pairs owing to time lags, frequency shifts and complex interference pattems. 8. Coherence is an important property to evaluate because it influences how broadly we can extrapolate results from a lake or set of lakes (for example, to a lake district) and because manipulative whole lake experiments often use adjacent lakes as reference lakes to evaluate treatment effects. The low coherence in this set of lakes in general and of individual variables such as chlorophyll concentrations and yellow perch year-class strength are discussed. 9. We found no studies that explicitly addressed interlake coherence; one long-term study of forest production made it clear that coherence among sites in a landscape will be a function of scale.