Abstract
Mean irradiance in the surface layer was calculated as a percent of incident irradiance using Secchi depths and thermocline depths for two sets of lakes: (1) 20 nearly circular lakes in northwest Minnesota, USA, sampled at a similar season, and (2) 1140 lakes of heterogeneous shapes in Ontario, Canada, sampled throughout the ice-free season. For both, as lake area increased, the thickness of the surface layer increased. Consequently, lakes larger in area had lower mean irradiance in their surface layer. Lake depth had less influence, but in the Ontario lakes, as depth increased, attenuation decreased, and mean irradiance increased. Lake area (log transformed) explained ca. one-half of the variation in mean irradiance in the homogeneous sampling of Minnesota lakes, while area and depth together explained less of the variation in the heterogeneous sampling of Ontario lakes. Shape, i.e. shoreline development, was unimportant compared with the gross measures of size, area and depth. Mean irradiance is complexly related to lake size; when depth increases slowly with area, larger lakes may have lower mean irradiance in their surface layers. But, when depth increases quickly with area, mean irradiance can be unrelated to area.