population mode to the variable cycle length shows that, for this animal, staying close
to a preferred light intensity is a proximate cause for vertical movement. The
scattering layer experiment, particularly the response to an eclipse, shows that at least
some marine plankton or micronekton also shift vertical position in response
primarily to changes in irradiance. Elaborate extensions of such experimental work
have been carried out. They show many variants, including species which respond to
the rate of change of light intensity, not to intensity per se.
(^) The obvious ultimate value of moving down into dark layers during daytime is to
avoid visual predators, and that is certainly the principal selective advantage that has
driven the recurring evolution of DVM in many distinctive animal groups, including
every phylum represented in the plankton. As stated above, the darkness at depth is
one of the few forms of cover available in pelagic habitats. This adaptive value of
DVM is simpler to propose than to prove, but clever observational designs and some
luck have provided proof that predator avoidance is primary.
(^) The strongest indication that predation avoidance is the adaptive value of DVM
comes from species in which it seems to be optional. Bollens and Frost (1989a)
compared the night and day vertical distributions of Calanus pacificus females in a
fjord between the spring periods of two years. In 1985 the day and night distributions
were identical, while in 1986 there was a strong shift between deeper day and
shallower night distributions. In both years, strong migrations set in eventually and
were marked in August (Fig. 8.19). This difference was accompanied by absence from
net samples of sand lance larvae early in 1985 (resulting from failed spawning), while
trawl tows in May 1986 showed ∼13 juvenile sand lance (and a few other
planktivorous fish) per 10,000 m^3 in the upper layer of the fjord. By summer of both
years, other planktivorous fish had moved into the fjord, producing threats from visual
predation equivalent to that from sand lance in spring 1986. Presence of predators
seemed to correspond to adoption of DVM by the copepod, and more importantly, the
copepod did not migrate vertically unless the predators were present. At least in this
instance, the cost of migrating is clear, the migrating Calanus had to forgo surface
feeding during daylight. There is no phytoplankton food in this fjord at the depths
occupied in daytime when migrating, and copepod guts are empty at depth from dawn
to dusk (Dagg et al. 1989).
Fig. 8.19 Comparisons based on stratified net tows of day (open bars) and night (dark
bars) vertical distributions of Calanus pacificus females in Dabob Bay, Washington
State. Diel vertical migration was absent in April, obvious in June and August.
(^) (After Bollens & Frost 1989a.)