Many pelagic animal species in the marine environment and in lakes migrate to deeper water layers before
sunrise and return around sunset. The amplitude of these diel vertical migrations (DVM) varies from several
hundreds of metres in the oceans to approx. 5–20 m in lakes. DVM can be studied from a proximate and
an ultimate point of view. A proximate analysis is intended to reveal the underlying behavioural mechanism
and the factors that cause the daily displacements. The ultimate analysis deals with the adaptive significance
of DVM and the driving forces that were responsible for the selection of the traits essential to the behavioural
mechanism. The freshwater cladoceran Daphnia is the best studied species and results can be used to model
migration behaviour in general. Phototaxis in Daphnia spp., which is defined as a light-oriented swimming
towards (positive phototaxis) or away (negative phototaxis) from a light source, is considered the most
important mechanism basic to DVM. A distinction has been made between primary phototaxis which occurs
when light intensity is constant, and secondary phototaxis which is caused by changes in light intensity. Both
types of reaction are superimposed on normal swimming. This swimming of Daphnia spp. consists of
alternating upwards and downwards displacements over small distances. An internal oscillator seems to be
at the base of these alternations. Primary phototaxis is the result of a dominance of either the upwards or
the downwards oscillator phase, and the direction depends on internal and external factors: for example,
fish-mediated chemicals or kairomones induce a downwards drift. Adverse environmental factors may
produce a persistent primary phototaxis. Rare clones of D. magna have been found that show also persistent
positive or negative primary phototaxis and interbreeding of the two types produces intermediate progeny:
thus a genetic component seems to be involved. Also secondary phototaxis is superimposed on normal
swimming: a continuous increase in light intensity amplifies the downwards oscillator phase and decreases
the upwards phase. A threshold must be succeeded which depends on the rate and the duration of the relative
change in light intensity. The relation between both is given by the stimulus strength versus stimulus duration
curve. An absolute threshold or rheobase exists, defined as the minimum rate of change causing a response
if continued for an infinitely long time. DVM in a lake takes place during a period of 1·5–2 h when light
changes are higher than the rheobase threshold. Accelerations in the rate of relative increase in light intensity
strongly enhance downwards swimming in Daphnia spp. and this enhancement increases with increasing fish
kairomone and food concentration. This phenomenon may represent a ‘decision-making mechanism’ to
realize the adaptive goal of DVM: at high fish predator densities, thus high kairomone concentrations, and
sufficiently high food concentrations, DVM is profitable but not so at low concentrations. Body axis
orientation in Daphnia spp. is controlled with regard to light–dark boundaries or contrasts. Under water,
contrasts are present at the boundaries of the illuminated circular window which results from the maximum
angle of refraction at 48·9° with the normal (Snell's window). Contrasts are fixed by the compound eye and
appropriate turning of the body axis orients the daphnid in an upwards or an obliquely downwards
direction. A predisposition for a positively or negatively phototactic orientation seems to be the result of a
disturbed balance of the two oscillators governing normal swimming.
Some investigators have tried to study DVM at a laboratory scale during a 24 h cycle. To imitate nature,
properties of a natural water column, such as a large temperature gradient, were compressed into a few cm.
With appropriate light intensity changes, vertical distributions looking like DVM were obtained. The results
can be explained by phototactic reactions and the artificial nature of the compressed environmental factors
but do not compare with DVM in the field.
A mechanistic model of DVM based on phototaxis is presented. Both, primary and secondary phototaxis
is considered an extension of normal swimming. Using the light intensity changes of dawn and the differential
enhancement of kairomones and food concentrations, amplitudes of DVM could be simulated comparable
to those in a lake.
The most important adaptive significance of DVM is avoidance of visual predators such as juvenile fish.
However, in the absence of fish kairomones, small-scale DVMs are often present, which were probably
evolved for UV-protection, and are realized by not enhanced phototaxis. In addition, the ‘decision-making
mechanism’ was probably evolved as based on the enhanced phototactic reaction to accelerations in the rate
of relative changes in light intensity and the presence of fish kairomones.