Sampling strategies and associated errors in estimating epilimnetic chlorophyll in eutrophic lakes

The spatial distributions of chlorophyll a were studied in Lake Mendota and Delavan during thermal stratification in 1971 and 1972. Both lakes are calcareous and support large epilimnetic chlorophyll standing crops (CSC) (milligrams per square meters) of planktonic blue‐green algae. The root‐mean‐square (rms) coefficient of variation (CV) between chlorophyll samples l m apart in the epilimnetic profile was 23% during sunny windless interludes following cold fronts. The CV for samples within the profile increased as the two thirds power of their separation distance. During conditions of elevated wind stress the interstrata CV for chlorophyll converged to the pure analytical error (2–3%). For Mendota, the rms CV (CSC) varied as 7 ds or, alternatively, 9.5 ds2/3, where ds represents the interstation separation distance (kilometers). The rate of increase was greater for Delavan (∼13 ds), probably because of increased nutrients, hence the increased incidence of large self‐shaded and buoyancy‐prone chlorophyll standing crops that are readily advected by the wind. Profile sampling near the center of Mendota or Delavan provides an unbiased estimate of pelagic mean CSC with a coefficient of variation of 14%. Lake stations at the margins of the pelagic zone (NE, NW, SW, and SE) are biased for pelagic mean CSC and also have large temporal CV. The station biases reflect the prevailing westerly winds; the large CV reflect the cyclical wind shifts related to cyclonic storms.