HOBSON ET AL.t CREPUSCULAR AND NOCTURNAL ACTIVITIES OF CALIFORNIA FISHES 



ators can be led by luminescent plankton to prey 

 over some distance (Nichol 1962), but we suggest 

 the approach must be made with great stealth to 

 avoid turbulence and resulting luminescence. 

 Surely such limitations preclude many kinds of 

 predatory activities after dark, especially when 

 the prey are as agile as most small fishes. In fact, 

 the diminished threat from predators that smaller 

 reef fishes enjoy at night (Hobson 1973, 1975, 1979) 

 may stem largely from the difficulties predators 

 have at this time moving undetected to within 

 striking range. 



If fishes use luminescent plankton to detect 

 predators and prey, undoubtedly they have experi- 

 enced strong selection pressures to enhance this 

 detection. An obvious adaptive response to such 

 pressures would be a match of the scotopic visual 

 pigments to the emission spectra of the lumines- 

 cent plankton. Certain vertebrates living at great 

 depths, or in the open ocean, reportedly have 

 scotopic pigments that match the luminescent 

 emissions of organisms with which they interact, 

 socially or as predator or prey (Clarke 1936; Munz 

 1958a; Lythgoe and Dartnall 1970; McFarland 

 1971; Locket 1977). These animals, however, inter- 

 act directly with the luminescent organisms, 

 whereas we stress indirect interactions. Of course, 

 the principle is the same either way — detection is 

 enhanced by matching visual pigment absorption 

 to a luminescent emission. 



The emission spectra of most luminescent 

 plankton, as exemplified by Noctiluca miliaris and 

 Gonyaulax polyedra (Figure 5), indicate that 

 fishes would sense the emitted photons best with 

 blue-sensitive visual pigments that have Xmax val- 

 ues near 490 nm. But this holds only before the 

 light has passed through water. As noted above, 

 the spectrum of light changes as it travels through 

 water, with the degree of change sharply affected 

 by the water's clarity. Clearly any such change 

 will favor a different k^^ value in visual pig- 

 ments. 



The relative effectiveness of visual pigments 

 with differing X^ax values in waters of differing 

 clarities can be estimated with some simple calcu- 

 lations. Given the relative attenuation of light at 

 each wavelength, which is a fiinction of water 

 type, we can compare the relative photoabsorption 

 of different visual pigments at increasing dis- 

 tances from the luminescent source. Let us con- 

 sider, for example, how a series of pigments with 

 ^max values at 10 nm intervals between 450 and 

 550 nm, each at 0.4 absorbance units, would ab- 



sorb the light emitted by Noctiluca as this light 

 passes through differing clarities of seawater, as 

 defined by Jerlov (1968). The relative photoabsorp- 

 tion of each pigment at a given distance from the 

 source can be estimated by multiplying that pig- 

 ment's percentage absorption at each wavelength 

 by the relative amount of light available at that 

 wavelength and then integrating the products 

 over all wavelengths (Figure 14). Thus, to see 





c 



CO 



z 

 o 



I- 

 o 



I 



Q. 



450 500 



WAVELENGTH >v 



550 



max 



(nmJ 



FIGURE 14. — Expected relative effectiveness of visual pigments 

 with differing A^iax values (positioned at wavelengths indicated 

 by circles) in sensing luminescent emissions ofNoctiluca miliaris 

 under varying circumstances. Top curve, for reference, repre- 

 sents effectiveness at zero range (no alteration of emission spec- 

 trum by intervening water); dashed curve (with open circles), for 

 comparison, represents effectiveness at distance of 3 m in clear 

 tropical water (water type lA of Jerlov 1968); three lower curves 

 represent relative effectiveness in waters of decreasing clarity 

 (water types 1, 5, and 7 of Jerlov 1968). Vertical line crosses each 

 curve at the optimal >^max position. Method of calculation in text. 



luminations from an organism like A/", miliaris at a 

 distance of 3 m in coastal waters (as at Santa 

 Catalina), fishes would best have visual pigments 

 with Amax values between 500 and 510 nm. And as 

 range increases, and water clarity decreases, 

 photodetection of this luminant source would be 

 improved by shifting the \^^ position slightly, but 

 continuously, toward the greener wavelengths 

 (Figure 15). 



In reality, of course, the reduced visibility in 

 turbid water sharply limits the practical extent of 

 such a shift. During heavy ph5rtoplankton blooms, 

 for example, even large objects in full daylight are 



25 



