Biology Reference
In-Depth Information
(3) there are high levels of fluctuation in recruitment (possible evidence
for weak regulation, except in the earliest life-cycle stages).
The authors point out that the time until effective explosion or extinction
may be long (possibly 100 years for thousandfold changes in abundance)
under weak regulation, but that there are few records implying greater
persistence of particular fish stocks than this. Furthermore, fluctuations in
population size are so great that, in spite of the circumstantial evidence for
regulation, analysis of stock recruitment diagrams rarely provides evidence
for or against regulation. Because of the difficulties in demonstrating
density-dependent regulation, the authors performed a simulation study
to test whether ''stochastic regulation'' is capable of maintaining fish stocks
over a wide range of fishing mortality rates. The following assumptions are
the basis for the presumed stochastic regulation:
(1) recruitment of fish stocks is highly variable but non-negative, leading
to a skewed probability distribution better described, for example, by
a log-normal than a symmetric (such as a normal) distribution;
(2) variability of recruitment increases with declining stock sizes; and
(3) median recruitment is deterministically determined.
However, assumptions (2) and (3) have not been proven. Nevertheless,
the authors conclude that it seems indeed plausible or even likely that the
only regulatory mechanism for fish stocks is stochastic, i.e., an increased
and non-normal variability in low stock sizes. The effect would be
''strong regulation in the mean,'' resulting from the increasing excess of
the mean over the median when population sizes are small. These results
may be suggestive, but require confirmation by an analysis of real long-
term data sets. Is stochastic ''regulation,'' even if it occurs, really strong
enough to produce the effects shown in the simulation study? Also, is the
term regulation appropriate, considering that ''regulation'' in the simula-
tions is based on purely stochastic events, and that density-dependent
processes are not involved? Finally, as pointed out by the authors them-
selves, an approach at least as valuable would be to examine empirically
the causes of variability themselves.
Crawley (
1990
) reviewed long-term studies of plant populations and
found evidence for regular patterns in relatively few of these; density-
dependent processes are important in certain populations for which
several generations have been monitored, and, in contrast to animal
populations, where the significance of competition is ''a matter of debate,''
it is ''quite clearly of over-riding importance'' in most plant assemblages.
