For requirement (iii), most mesopelagic squid, many fish, and most euphausiids
have a photophore in close association with one or both eyes, usually actually built
into the eye. It is probable that these produce a neural signal in the visual system
proportional to the overall luminescent intensity. The photophore in the eye must be
regulated by the same process regulating those producing ventral illumination, thus
providing a feedback loop for adjustment to a match. In addition to the demonstration
of such a feedback by Warner et al., experiments with squid by Young et al. (1979)
show that matching is disrupted when light-sensitive vesicles on the dorsal side of the
body are covered. Both the principal eyes and dorsal light-sensitive vesicles are
involved in determination of downwelling intensity, since matching deteriorates when
any of those organs is covered.
(^) Bioluminescent countershading is in some instances even more complex. Young and
Mencher (1980) made spectral scans of photophore output at different temperatures
(Fig. 12.5) from the ventral skin of Abraliopsis pacificus, a small, muscular squid
abundant near Hawaii. At the colder (8°C) temperatures of its daytime, mid-water
habitat, it produces a narrow spectrum centered at 472 nm, very close to the
downwelling spectrum. At the warmer temperature (23°C) of the surface layers, it
migrates up into at night, it produces a much broader spectrum, one remarkably close
to that of moonlight viewed through 20 m of water. Thus, it not only matches the
downwelling intensity, it matches the spectral composition through a dramatic change.
To do this, it has three types of ventral photophores, two with blue filters, one with
red filters (Plate 12.1). The selection of a matching combination is not made by