FISHERY BULLETIN: VOL. 79, NO. 1 



their activities cover observations over 15 yr in 

 southern California — from San Diego north to 

 Point Conception — but our more detailed obser- 

 vations, along with the light measurements and 

 analysis of visual pigments, refer to Santa 

 Catalina Island (lat. 33°28 ' N, long. 118°29 ' W), 35 

 km from the mainland (Figure 1). Here the water 

 is consistently warmer and more transparent than 

 on the adjacent mainland; during our study sur- 

 face temperatures ranged between about 11 ° and 

 20° C, and underwater visibility generally ex- 

 ceeded 10 m. Thus, when related to comparable 

 data collected earlier in the tropics (Hobson 1968a, 

 1972, 1974; Munz and McFarland 1973; McFarland 

 and Munz 1975a), these results offer a conserva- 

 tive measure of differences between warm- 

 temperate and tropical habitats. 



METHODS 



Determining the Spectral Composition of 

 Submarine Sunlight 



The spectral distribution of submarine light was 

 measured with a Gtunma 3000R spectroradiome- 

 ter^ mounted in an underwater housing (Munz 

 and McFarland 1973). The instrument, fitted with 



■•Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



a cosine receptor head and calibrated in photons 

 per square centimeter per nanometer per second, 

 draws a quantal irradiance spectrum. Radiance of 

 the backlighting along a particular line of sight 

 was measured by restricting the angle of view of 

 the receptor head to a narrow cone (ca. 0.008 

 steradians). Usually radiance was determined 

 along the zenith, horizontal, and nadir lines of 

 sight. 



Because we were interested in comparing the 

 spectral distribution of submarine light for differ- 

 ent water conditions and along different lines of 

 sight, results have been normalized and are pre- 

 sented in terms of relative number of photons. The 

 light levels that occur at twilight were beyond the 

 spectroradiometer's sensitivity for measurement 

 of spectral radiance. At twilight, therefore, spec- 

 tral irradiance and not spectral radiance was 

 measured. Irradiance data are reported in terms of 

 absolute numbers of photons. 



To facilitate comparisons, several of the spectral 

 curves were indexed by calculating their A.P50 val- 

 ues (Munz and McFarland 1973; McFarland and 

 Munz 1975a). The AP50 value represents the 

 wavelength within the visible spectrum (400-700 

 nm) that halves the total number of photons under 

 a spectral curve. Because underwater light is usu- 

 ally homochromatic and fairly symmetrical in dis- 

 tribution, XP50 provides a useful single index to a 

 spectrum. 



34' 



33° 



Figure 1.— The study area in southern 

 California. Santa Catalina Island was 

 the site of detailed observations, includ- 

 ing light measurements and analysis of 

 visual pigments in fishes. 



32° 



I I i^i^^j I J )■ T-IJll"! J It ri 1 tl { 11 t T f I I I ( I I 1 T't* T! r 15 T f J 

 ^.SANTA BARBARA 

 PT CONCEPTION 



LOS ANGELES 



50 100 



Ul I— I I— I U-l 1— I J 



KILOMETERS 



I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 



120° 



119° 



118° 



