Abstract
Primary production rates derived from model photosynthesis–light (P–l) curves and daily solar radiation data were compared with direct measurements using "simulated" in situ incubations in arctic and temperate marine waters. On the average, model estimates were slightly higher in surface waters and significantly lower at the bottom of the euphotic zone (1% light level) than were the measured values. This could be partially explained by spectral differences in incubation light sources. However, areal production rates were statistically indistinguishable. Use of P-l models without a photoinhibition term and incorporating P-l pararmeters from mixed-layer populations gave the best overall agreement with directly measured production rates.