Biological Oceanography

(^) One such model was developed by Simon Wood and applied to a Pseudocalanus
newmani population in Dabob Bay, Washington, by Ohman and Wood (1996). The
method needs data on the progression of stage abundances. In order to avoid obtaining
both stage duration and mortality from the field data, the user must apply laboratory
estimates of development rate under the food, temperature, and other conditions
pertaining in the field. Unfortunately, there is no guarantee that development rate in
the field is the same as that measured in experiments. The progression of stage
abundances is then fitted by a smoothing function with several assumptions. One is
that the population is closed, which is, of course, the basic assumption of any
horizontal life-table method. For zooplankton, this restricts the method to a few, rather
odd sites with very little net advection to carry animals either in or out. Ages at
advance to each stage are assumed to be fixed: ages in successive stages do not
overlap. Unfortunately age-within-stage at molt can vary by a factor greater than two
for copepods raised in conditions as nearly identical as can be arranged or imagined to
occur in the field (Carlotti & Nival 1991). Even from the same clutch of eggs, some
fourth copepodites will be younger than some second copepodites! Wood’s method
assumes that can’t happen and may or may not be robust against it. A main feature of
the method is fitting of a surface represented by cubic splines to the abundance data.
That provides mathematical functions that can be integrated explicitly, as well as
getting rid of some of the troubling chaos of real data. The splines are then modeled
(the method is very complex) by equations for each stage reproducing the function in
terms of throughput (stages advance at rates specified from rearing experiments) and
stage mortality (the final output).
(^) The results (Fig. 8.11) are exponential stage-specific mortality rates (d−1). For P.
newmani in Dabob Bay, the variations among stages were reasonably consistent
between two years, suggesting that mortality varies strongly from stage to stage. As in
Johnson’s study, the nauplii were not distinguishable from those of sympatric
congeners, so their mortality was approximated from likely input estimated from