Behrenfeld and Falkowski 1997a). P
b is a function of incident photosynthetically
available radiation (PAR) and thus varies with time of day, depth in the water
column, season, and cloudiness. Fully resolved NPP models attempt to characterize this variability in the dynamic underwater light field. However, many
simpler NPP algorithms employ time- and depth-integrated parameters and calculate productivity as a function of incident daily PAR and a maximum daily
assimilation efficiency for the water column P
b
opt
. Behrenfeld and Falkowski
(1997a) summarize and compare the various classes of models, distinguishing
between approaches by time, depth, and wavelength resolution.
A key derived property for many ecological applications is daily water-column–integrated NPP (
P
PP). In addition to subsurface light availability described
above, assessment of
P
PP requires assumptions regarding other depth-dependent
properties. In particular, surface mixed layer depth and vertically-varying nutrient
loads, grazing pressures, and light conditions give rise to variations in biomass and
phytoplankton physiological state (photoacclimation and growth rate). Numerous
approaches have been developed to account for these effects. Depth-integrated
models generally assume the water column is composed of two layers, one light
saturated and the other light limited. An empirical function then relates the fraction
of the water column that is light saturated to the incident irradiance. Behrenfeld
and Falkowski (1997a) demonstrate that this approach is sufficient to capture
[80 % of the variance in observed
P
PP when evaluated over a wide range of
trophic conditions. Depth-resolved NPP models may take many forms. In some
cases, vertical structure is prescribed using empirical relationships with surface
properties (e.g., characterizing the profile of chlorophyll from surface chlorophyll
concentration). More complex approaches incorporate information on mixing
depths to assign an upper layer of uniform biomass and physiology, and then
below this depth iteratively adjust chlorophyll stocks and physiological state based
on models of photoacclimation, attenuation, and a prescribed shift from nutrient
limitation to light limitation at depth (e.g., Westberry et al. 2008). As our
knowledge of vertical variability improves, it will be these depth-resolved models
that will provide the appropriate model scaffolding to incorporate this information
to achieve improved NPP assessments.
All chlorophyll-based models of NPP, from simple depth-integrated algorithms
to fully resolved time- and depth-dependent models, require a characterization of
assimilation efficiencies (P
b ; P
b
opt , etc.). In most cases, this aspect of NPP models is
least mature. The most common approach is to relate assimilation efficiency to sea
surface temperature (SST), with a somewhat bewildering array of SST-dependent
models proposed (see Fig. 8.4 in Behrenfeld and Falkowski 1997a). One of the
most commonly employed functions expresses assimilation efficiency as an
increasing exponential function of temperature. The Q 10 for this exponent was
based on a compilation of laboratory phytoplankton growth rates, where an
exponential relationship was fit to the maximum observed growth rates over a wide
range of temperatures (after Eppley 1972). There is no a priori reason to assume
that this growth-rate-based function has any direct physiological relevance to the
208
T. K. Westberry and M. J. Behrenfeld
b is a function of incident photosynthetically
available radiation (PAR) and thus varies with time of day, depth in the water
column, season, and cloudiness. Fully resolved NPP models attempt to characterize this variability in the dynamic underwater light field. However, many
simpler NPP algorithms employ time- and depth-integrated parameters and calculate productivity as a function of incident daily PAR and a maximum daily
assimilation efficiency for the water column P
b
opt
. Behrenfeld and Falkowski
(1997a) summarize and compare the various classes of models, distinguishing
between approaches by time, depth, and wavelength resolution.
A key derived property for many ecological applications is daily water-column–integrated NPP (
P
PP). In addition to subsurface light availability described
above, assessment of
P
PP requires assumptions regarding other depth-dependent
properties. In particular, surface mixed layer depth and vertically-varying nutrient
loads, grazing pressures, and light conditions give rise to variations in biomass and
phytoplankton physiological state (photoacclimation and growth rate). Numerous
approaches have been developed to account for these effects. Depth-integrated
models generally assume the water column is composed of two layers, one light
saturated and the other light limited. An empirical function then relates the fraction
of the water column that is light saturated to the incident irradiance. Behrenfeld
and Falkowski (1997a) demonstrate that this approach is sufficient to capture
[80 % of the variance in observed
P
PP when evaluated over a wide range of
trophic conditions. Depth-resolved NPP models may take many forms. In some
cases, vertical structure is prescribed using empirical relationships with surface
properties (e.g., characterizing the profile of chlorophyll from surface chlorophyll
concentration). More complex approaches incorporate information on mixing
depths to assign an upper layer of uniform biomass and physiology, and then
below this depth iteratively adjust chlorophyll stocks and physiological state based
on models of photoacclimation, attenuation, and a prescribed shift from nutrient
limitation to light limitation at depth (e.g., Westberry et al. 2008). As our
knowledge of vertical variability improves, it will be these depth-resolved models
that will provide the appropriate model scaffolding to incorporate this information
to achieve improved NPP assessments.
All chlorophyll-based models of NPP, from simple depth-integrated algorithms
to fully resolved time- and depth-dependent models, require a characterization of
assimilation efficiencies (P
b ; P
b
opt , etc.). In most cases, this aspect of NPP models is
least mature. The most common approach is to relate assimilation efficiency to sea
surface temperature (SST), with a somewhat bewildering array of SST-dependent
models proposed (see Fig. 8.4 in Behrenfeld and Falkowski 1997a). One of the
most commonly employed functions expresses assimilation efficiency as an
increasing exponential function of temperature. The Q 10 for this exponent was
based on a compilation of laboratory phytoplankton growth rates, where an
exponential relationship was fit to the maximum observed growth rates over a wide
range of temperatures (after Eppley 1972). There is no a priori reason to assume
that this growth-rate-based function has any direct physiological relevance to the
208
T. K. Westberry and M. J. Behrenfeld