Three factors motivate the search for
mathematical formulations of the relationship
between photosynthesis and light for
phytoplankton (the light-saturation curve).
First, the simulation and prediction of
changes in phytoplankton biomass by
means of mechanistic models require a
quantitative expression for photosynthesis.
Second, if in situ productivity could be estimated
with sufficient precision from more
easily measured variables (irradiance,
chlorophyll a, nutrients, temperature, etc. )
we could collect productivity data with
greater spatial and temporal resolution for
the same amount of effort. Third, a mathematical
formulation enables us to detcrmine
the physiological parameters of photosynthesis
operationally by choosing those
parameter values which produce the best
fit between the equation and the experimental
data.
Of the equations that have been proposed
to describe the utilization of light by
phytoplankton, a few are based on grossly
oversimplified models of photosynthesis;
the rest are completely empirical. This being
the case, the only valid criterion for
judging the relative merits of the different
formulations is their ability to describe
data with the minimum number of well
chosen parameters. As discussed at some
length in Platt et al. (1975), by “well
’ Present address : Lawrence Berkeley Laboratory,
Energy and Environment Division, University
of California, Berkeley 94720.
chosen” we mean parameters which are, as
far as possible, mututally independent, and
which have an informative physical interpretation,
that is a significance transcending
their role as operators in a particular
mathematical expression.
Here we compare the merits of eight different
representations of the photosynthesis-light
relation judged according to this
criterion of ability to describe experimental
data. We treat only that part of the curve
up to and including light saturation. We
shall not consider the range of light levels
above the threshold of photoinhibition,
where the introduction of at least one additional,
independent parameter would be required
for a complete description (Platt et
al. 1975).
The test data comprise measurements of
14C uptake and irradiance from 188 experiments
made at three coastal locations in
Nova Scotia over a period of 2 years. We
thank B. Irwin, P. Dickie, and D. Rudderham
for technical help and K. L. Denman
for constructive discussion. T. T. Bannister
and J. Steele made helpful comments on
the manuscript. We are indebted to S. Taguchi
for permission to refer to his unpublished
data.
Basis for compa