406
x-ray emission (PIXE) has been used to examine surface and intracellular concentrations of
Ag, Cu, Hg, Zn, Cd, Mg, P, S, Cl, Ar, K, Ca, Fe, Sr, and Pb in Chara corallina (Jahnke
et al., 1981) Synechococcus. Oscillatoria. Scenedesmus, and Ankistodemus (MacArthur et al.,
1985). Accelerator mass spectrometry (AMS) may be useful for determining long-lived
radioisotopic compositions of particles containing as little as a few ng of material (Elmore and
Phillips, 1987). These, and other analyses, may prove valuable in helping to elucidate the
pathways and fate of sorbed or colloidal materials in the ocean.
In so far as the virtues of individual cell or particle analysis are concerned, the question
emerges as to how to abstract useful information from potentially very large data sets. New
numerical tools are being developed to specifically deal with flow cytometry data (e.g. Li,
1990; Demers et al., ms). On the other hand, there are attempts, in the present volume, to
deal with the problem of scaling from cells or particles up to satellite images (e.g. papers in
this volume, Legendre and Le Fevre; Lewis and Cullen; Holligan and Balch; Morel). This
problem is not unique to oceanography, and it plagues terrestrial plant ecologists as well
(Osmond, 1989) at such a point that it sometimes completely impedes progress. Satellite
images are powerful because they integrate information over large areas. However, attempts
to describe the optical properties of the upper ocean as inferred from satellites, using the
properties of individual cells or particles, is a bit like resolving whether light is best described
as a particle or an electromagnetic wave. Both approaches are valid, but the approach of
inferring bulk optical properties from the number and types of individual particles is daunting,
perhaps a bit like taking the census of sheep by counting their legs and dividing by four.
Instead, it seems that a more useful application of optical or other individual cell properties
would be in inferring the effects of the physical or chemical environment from apparent cell
responses. Such an approach is exemplified by Dusenbury (pers. comm.), who examined
phytoplankton responses to the effects of diurnal heating of the upper mixed layer. This is
attractive in that the information is not lost in volumes of data, and also because it would be
difficult to obtain the information as cleanly by any other approach. Our suggestion is that
analyses of individual cells should be preferably applied to those areas where information
about individual cells is important, and not used as sophisticated overkill for bulk property
measurements.
x-ray emission (PIXE) has been used to examine surface and intracellular concentrations of
Ag, Cu, Hg, Zn, Cd, Mg, P, S, Cl, Ar, K, Ca, Fe, Sr, and Pb in Chara corallina (Jahnke
et al., 1981) Synechococcus. Oscillatoria. Scenedesmus, and Ankistodemus (MacArthur et al.,
1985). Accelerator mass spectrometry (AMS) may be useful for determining long-lived
radioisotopic compositions of particles containing as little as a few ng of material (Elmore and
Phillips, 1987). These, and other analyses, may prove valuable in helping to elucidate the
pathways and fate of sorbed or colloidal materials in the ocean.
In so far as the virtues of individual cell or particle analysis are concerned, the question
emerges as to how to abstract useful information from potentially very large data sets. New
numerical tools are being developed to specifically deal with flow cytometry data (e.g. Li,
1990; Demers et al., ms). On the other hand, there are attempts, in the present volume, to
deal with the problem of scaling from cells or particles up to satellite images (e.g. papers in
this volume, Legendre and Le Fevre; Lewis and Cullen; Holligan and Balch; Morel). This
problem is not unique to oceanography, and it plagues terrestrial plant ecologists as well
(Osmond, 1989) at such a point that it sometimes completely impedes progress. Satellite
images are powerful because they integrate information over large areas. However, attempts
to describe the optical properties of the upper ocean as inferred from satellites, using the
properties of individual cells or particles, is a bit like resolving whether light is best described
as a particle or an electromagnetic wave. Both approaches are valid, but the approach of
inferring bulk optical properties from the number and types of individual particles is daunting,
perhaps a bit like taking the census of sheep by counting their legs and dividing by four.
Instead, it seems that a more useful application of optical or other individual cell properties
would be in inferring the effects of the physical or chemical environment from apparent cell
responses. Such an approach is exemplified by Dusenbury (pers. comm.), who examined
phytoplankton responses to the effects of diurnal heating of the upper mixed layer. This is
attractive in that the information is not lost in volumes of data, and also because it would be
difficult to obtain the information as cleanly by any other approach. Our suggestion is that
analyses of individual cells should be preferably applied to those areas where information
about individual cells is important, and not used as sophisticated overkill for bulk property
measurements.