Species-specific phytoplankton growth rates via diel DNA synthesis cycles. IV Evaluation of the magnitude of error with computer-simulated cell populations

Abstract

computer model was constructed to evaluate the magnitude of error embedded in the estimation of phytoplankton growth rate via cell cycle analysis. Attempts have been made to assess errors from different parts of computational procedures, errors caused by variations in curve fitting, and errors arising due to situations that do not match the assumptions used In the original denvations. Results indicated that errors involved In measuring the fraction of a population in the combined S, G2, and M phase have minimal influence on the accuracy of estimated growth rate When errors in the estimation of the duration of cell cycle phase are included, an averaged deviation of 11 % was observed. If the additional errors in the measurement of single-cell DNA content are also considered, the averaged deviation in estimated growth rates increases to 22 %. Changing the sampling interval from 1 h to 2 h created only a minor increase in relative error. Among different methods for defining maximum values of phase fractions, the periodic curve fitting technique performed the best. For special situations that do not match original assumptions, a nght-skewed S phase fraction curve caused the largest deviation, but the deviation was less then 41 %.