Problems with Predicting Size-Fractionated Productivity from Chlorophyll Biomass in Marine Ecosystem Models

Estimates of primary productivity are critical components of marine ecosystem models. Models are often constructed with 2 or more size classes of phytoplankton as a way to partition flows of productivity through the microbial food web or directly to larger grazers. In the absence of measurements of size-fractionated primary productivity, this partitioning is done based on estimates of size-fractionated biomass, usually measured as chlorophyll-a. This requires, then, that modelers make the assumption that the contributions of a specific size-fraction to total biomass scales directly with its contributions to total primary productivity. When, if ever, is this assumption valid? I will present results of an analysis of approximately 30 data sets (to date) taken from a variety of marine ecosystems. These data sets include concurrent measurements of size-fractionated biomass and size-fractionated primary productivity, from which we can derive an estimate of the error associated with assuming direct proportionality. Preliminary results show that, when all data were considered together, contributions to primary productivity by the microphytoplankton (here defined as those greater than 20 microns in size) would be underestimated by an average of 4.2% if calculated based directly on relative contributions to chlorophyll-based biomass (n = 74; values ranged from a 49% underestimate to 16% overestimate). For the picophytoplankton (0.2 to 2 microns), contributions to primary productivity were overestimated by 4% (n = 113; values ranged from a 31% underestimate to a 51% overestimate). These errors do not include further variability that may be associated with converting from chlorophyll to carbon units, for example. Further analyses will look for trends in the data sets, including variations with depth, season, and biogeographic province.