lines in Box 4.4 (i.e. lines starting “%”, with instructions for making the changes). If
you run the model with those changes (and set nyear = 3), the output will be much the
same, but with somewhat realistic variability in the summer values of N, P, and H,
both during summer and between years. Since the random variation in mixing is about
the same magnitude as the onset of mixed-layer deepening in fall, the summer
oscillations usually replace the fall bloom.
(^) Having established a model that produced recurring seasonal cycles including a
spring bloom, Evans and Parslow set out to discover what would eliminate blooms.
Many oceanic ecosystems, mostly HNLC systems (see Chapter 11), consistently do
not exhibit seasonal blooms. They tried eliminating variation in mixed-layer depth.
That damps the cycle amplitude in our model, too, but cycles remain. Simplify the
program appropriately and try that. If the model is realistic in its fundamental
relationships, then it appears that blooms are affected by, but do not solely depend
upon, mixed-layer variation. The principal drivers must then be the cycle of insolation
and the response of grazers to it. Next, they asked whether the parameters describing
grazing could be responsible for the absence of strong phytoplankton cycling seen in
many oceanic areas, and found them strongly effective. In our version, increasing the
maximum per capita grazing rate, c, from 0.35 to 0.6 flattens the cycles of
phytoplankton stocks, eliminating the spring bloom (Fig. 4.6c), despite the ongoing
cycling of illumination and mixing rate. Nutrients, while cycling, remain high in the
mixed layer all year. That, in fact, remains the theory explaining the absence of
blooms in the subarctic Pacific and subantarctic. As will be seen (Chapter 11), it is
thought (and tested) that trace-metal limitation sets up phytoplankton–grazer relations
appropriate for grazing to consistently balance phytoplankton growth, but with
oscillations somewhat like those of summer in our model.
(^) When you try different parameters in this program, you will likely find some sets
with strong day-to-day oscillations and some producing negative values of state
variables. That is because of the whole-day time-steps. Changes of actual oceanic
phytoplankton biomass do come in daily increments (with increase in the light,
decrease in the dark), so there is a touch of realism in this choice. We have followed
Evans and Parslow in sticking to quite a shallow (80 m) winter mixed layer, much
shallower than usual North Atlantic winter mixing. A useful exercise to try is
reprogramming the model to increase winter mixing, say, to 300 m, i.e. below the
usually calculated critical depths (defined in Chapter 11). To limit bloom duration you
may need stronger grazing. You will not get a classic North Atlantic bloom, at least
not with the initial parameters.