numerous runs of the program. Whatever uncertainty that leaves about the reality of
the actual mechanisms, the model produces some fascinating results. The basically
flat total phytoplankton stocks of the region are again reproduced, but the small and
large phytoplankton have different seasonal cycles (Plate 4.1).
Fig. 4.10 State variables and transfers in the subarctic Pacific model of Denman et al.
(2006). Schematic representation of the nitrogen version of the ecosystem model. PS
represents nano- and picophytoplankton, PL diatoms, Z 1 microzooplankton, Ni nitrate,
Na ammonium, D detritus and bacteria, Z 2 (t) the imposed annual cycle of
mesozooplankton, f 1 formation of sinking diatom aggregates, f 2 loss to detritus of
unassimilated food of mesozooplankton that does not end up as ammonium, and f 3
grazing on diatoms by microzooplankton.
(After Denman et al. 2006.)
Small cells pulse at the surface in spring; diatoms become abundant during summer
in the deeper layer between the seasonal thermocline and the permanent halocline,
and then their stocks expand upward to the surface in fall. October peaks in
chlorophyll were noted in time-series sampling done from the weather ships several
decades ago. The model also produced blooms of first small and then large cells in
response to reducing the iron-limitation parameters, increases on roughly the same
time scale as the phytoplankton stock changes that occurred in the SERIES iron-
addition experiment in July–August of 2002 (Boyd et al. 2005). The vertical
separation of different classes of primary producers constitutes a hypothesis that
should be examined in the field.