A comparison of in situ and simulated in situ methods for estimating oceanic primary production

Primary production data measured by in situ (IS) and ‘simulated’in situ (SIS) incubations were compared. To minimize differencesbetween the two types of incubations, SIS experiments were conductedin temperature-controlled incubators in which the spectral distributionand irradiance were adjusted to approximate IS conditions. ISavailable irradiance (I_Is) was computed from vertical attenuationof integrated surface irradiance. Vertical attenuation was estimatedusing a spectral irradiance model, validated by measured profilesof the vertical attenuation coefficient. IS incubations werecarried out using two methods. The first involved deploymentof bottles on a drifting array for whole-day (dawn to dusk)incubations. The second method employed an autonomous submersibleincubation device that performed short term (<1 h) incubationsat multiple depths. Differences between whole-day IS and SISincubation estimates were attributed partially to differencesbetween I_IS and SIS-available irradiance (I_SIS). Photosynthesis-irradiance(P-I) properties of IS and SIS populations from the whole-dayincubations were not significantly different. P-I propertiesof the short-term IS and SIS populations were significantlydifferent, although estimates of P^B (mg C mg Chl^–1 h^–1)from contemporaneous IS and SIS incubations did not differ by>40%. Integrated water-column primary production (IPP) estimatedusing P-I models derived from SIS data were within 15% of ISestimates of IPP.