84
Even if one knew the appropriate weighting factors, the comparison of growth rates could not
be made by averaging the component growth rates. The reason is that growth rate and specific
activity are related in a logarithmic manner. The average value of a logarithm is not equal
to the logarithm of the average value of the argument of the logarithm. One can, however,
show that the carbon-specific growth rate of the entire population is equal to a weighted
average of the carbon-specific growth rates of the population components. The reasoning is
similar to that used in the foregoing example of chI a specific activities. Hence one can say
that the carbon-specific growth rate of the entire population should lie somewhere between the
extreme carbon-specific growth rates of the population components, being close to the
minimum if phytoplankton carbon is dominated by the slowest-growing component and close
to the maximum if phytoplankton carbon is dominated by the fastest- growing component.
Calculating the abundance of different algal taxa using pigment concentrations
Multiple regression analysis
Recently, Gieskes et al (1988) have used a mathematical analysis of algal pigment fingerprint
series described earlier (Gieskes and Kraay, 1983b) by which it is possible to estimate the
contribution of the major taxonomic groups to the total chi a contents of natural phytoplankton
populations. This method may suffer from the instability of the carotenoid-to-ChI a ratio,
which is in several classes determined by the physiological condition of cells (e.g., Klein,
1988) and the level of light adaptation (see previous section). Bidigare et al. (1989a)
summarized existing literature on light adaptation by stating that in the ocean's phytoplankton
the ratio of non- photosynthetically active pigments-to-Chl a decreases with depth while the
ratio of photosynthetically-active pigments-to-Chl a increases with depth. Gieskes et al. (1988)
avoided the ratio problem by first arranging similar populations by cluster analysis, then doing
multiple regression analysis on each group separately.
Another approach, and application to picoplankton populations from oligotrophic oceans
Another method developed with G. Kraay and M. Veldhuis, which calculates biomass of
individual taxonomic groups from pigment data is more straightforward. Errors made with this
approach are minimal when cell size accounts for variations in pigment concentration per unit
biomass. A few examples will be presented to illustrate how effective this method is. I use
the chlorophyll and zeaxanthin concentrations per cell of picoplankton Prochlorophytes given
in Chisholm et al. (1988). Concentrations recently published by Veldhuis and Kraay (1990)
Even if one knew the appropriate weighting factors, the comparison of growth rates could not
be made by averaging the component growth rates. The reason is that growth rate and specific
activity are related in a logarithmic manner. The average value of a logarithm is not equal
to the logarithm of the average value of the argument of the logarithm. One can, however,
show that the carbon-specific growth rate of the entire population is equal to a weighted
average of the carbon-specific growth rates of the population components. The reasoning is
similar to that used in the foregoing example of chI a specific activities. Hence one can say
that the carbon-specific growth rate of the entire population should lie somewhere between the
extreme carbon-specific growth rates of the population components, being close to the
minimum if phytoplankton carbon is dominated by the slowest-growing component and close
to the maximum if phytoplankton carbon is dominated by the fastest- growing component.
Calculating the abundance of different algal taxa using pigment concentrations
Multiple regression analysis
Recently, Gieskes et al (1988) have used a mathematical analysis of algal pigment fingerprint
series described earlier (Gieskes and Kraay, 1983b) by which it is possible to estimate the
contribution of the major taxonomic groups to the total chi a contents of natural phytoplankton
populations. This method may suffer from the instability of the carotenoid-to-ChI a ratio,
which is in several classes determined by the physiological condition of cells (e.g., Klein,
1988) and the level of light adaptation (see previous section). Bidigare et al. (1989a)
summarized existing literature on light adaptation by stating that in the ocean's phytoplankton
the ratio of non- photosynthetically active pigments-to-Chl a decreases with depth while the
ratio of photosynthetically-active pigments-to-Chl a increases with depth. Gieskes et al. (1988)
avoided the ratio problem by first arranging similar populations by cluster analysis, then doing
multiple regression analysis on each group separately.
Another approach, and application to picoplankton populations from oligotrophic oceans
Another method developed with G. Kraay and M. Veldhuis, which calculates biomass of
individual taxonomic groups from pigment data is more straightforward. Errors made with this
approach are minimal when cell size accounts for variations in pigment concentration per unit
biomass. A few examples will be presented to illustrate how effective this method is. I use
the chlorophyll and zeaxanthin concentrations per cell of picoplankton Prochlorophytes given
in Chisholm et al. (1988). Concentrations recently published by Veldhuis and Kraay (1990)