Abstract
Phytoplankton functional types are groupings of many species into a smaller number of types according to their ecological or biogeochemical role. Models describe phytoplankton functional types by a set of traits that determine their growth rates or fitness. Traits for functional types are often determined from observations on a small number of species under laboratory conditions. Functional types can be composed of a large number of species with very different trait values, so the representation of a type by an average trait value may not be appropriate. A potential solution is to estimate trait values from observations of the aggregate biomass of phytoplankton functional types in natural populations. We report on some recent efforts to extract trait values from time-series data using Bayesian statistical models and discuss some challenges of this approach.
