FISHERY BULLETIN: VOL. 86. NO. 4 



dances off California depend on nutrients from the 

 Gulf of Alaska, and Chelton et al. (1982) concluded 

 not only that phytoplankton off California depend 

 on nutrients from higher latitudes but also that 

 wind-forced coastal upwelling is relatively unim- 

 portant in supplying these nutrients. Despite this 

 uncertainty, growth in thaliaceans is so rapid that 

 it would seem at least many of those off the Men- 

 docino coast could result from local upwelling. 

 Regardless of what determines the growth of tha- 

 liaceans in this region, their appearance in large 

 numbers next to the beach is evidence that shore- 

 ward transport has developed. These animals are 

 readily detected by in situ observations, and also by 

 their occurrences in samples of both plankton and 

 gut contents (Table 3). 



Intermittent shoreward transport during the up- 

 welling season seems especially strong off northern 

 California. This is evident in the upwelling indices 

 of Bakun (1973), and also in that fewer thaliaceans 

 get inshore at this time of year, both to the north, 

 off Oregon (Hubbard and Pearcy 1971), and to the 

 south, off central and southern California (Black- 

 burn 1975). Thus, it is clear that the major foods of 

 5. mystinus along the Mendocino coast are most 

 available during downwelling episodes of the upwell- 

 ing season. 



Despite the increased productivity of the upwell- 

 ing season, S. mystinus experiences relatively poor 

 feeding conditions during that season's upwelling 

 episodes. Not only are fewer prey taken during 

 upwelling than during downwelling, a higher pro- 

 portion of the fish have empty stomachs (Fig. 13). 

 This is because the shoreward flow that transports 

 zooplankters from offshore during downwelling is 

 replaced by the seaward flow that is part of the 

 upwelling condition. 



There is, however, one relatively large gelatinous 

 zooplankter from offshore that is most available as 

 prey during upwelling. This is the chondrophore 

 Velella velella, a pelagic hydrozoan known as "by- 

 the- wind-sailor". It is, in fact, entirely because of 

 this animal that the food category "Pelagic Hydro- 

 zoa" ranked second as food during upwelling epi- 

 sodes (Table 2). Because V. velella floats on the 

 water's surface and is equipped with a sail-like struc- 

 ture (Fig. 8), its movements are determined more 

 by wind than by current. The species includes two 

 forms distinguished by whether their sails are ori- 

 ented to the left of or to the right of the main body 

 axis. This orientation determines their direction in 

 sailing before the wind— left-handed individuals sail 

 to the left of the wind direction and right-handed 



individuals sail to the right (Bieri 1959). Although 

 it has been reported that the right-handed form 

 predominates off California (Morris et al. 1980), all 

 those we examined from off Mendocino were left- 

 handed, and so would have been driven shoreward 

 by the northerly winds that generated upwelling. 



In the absence of favored open-water zooplank- 

 ters, 5. mystinus increased consumption of near- 

 shore hyperbenthic zooplankton, e.g., mysidaceans 

 and gammarideans (Table 2). But fewer of these 

 organisms were taken than might be expected, 

 based on their great abundance during much of the 

 upwelling season. Acanthomysis sculpta, the mysid 

 most often taken as prey, typically aggregates in 

 large swarms within 2 m of the seafloor. To prey 

 on them, adult 5. mystinus must leave the upper 

 levels of the water column in a departure from their 

 usual feeding mode that may reduce feeding effec- 

 tiveness. In addition, most hyperbenthic zooplank- 

 ters probably are too small to be ready prey of these 

 fish. Although 5-7 mm mysids (Table 2) should be 

 large enough, most other taxa are less than 2 mm. 

 Organisms as small as 1 mm occur in the diet, in- 

 cluding some thought to be strictly benthonic, e.g., 

 smaller of the gammaridean Jassa sp. (which also 

 occurred in plankton collections; Table 2). But such 

 forms may be ingested (and taken by plankton 

 nets) while attached to drifting plant fragments. 

 Although 5. mystinus has a smaller mouth than 

 most of its congeners, presumably as an adaptation 

 to planktivory (Hallacher and Roberts 1985), the 

 adults appear unable to consume the larvae of neritic 

 species, e.g., cirripedeans, that, with maximum 

 dimensions of 1 mm or less, often are the most 

 numerous of the zooplankton (Table 2). These lar- 

 vae are major prey of juvenile 5. mystinus (unpubl. 

 data; Gaines and Roughgarden 1987), which further 

 suggests it is their small size that precludes them 

 as prey of the adult. 



Foods most often consumed in the absence of pre- 

 ferred zooplankters, however, were plant materials. 

 In fact, during upwelling episodes more plants were 

 consumed than anything else (Table 2), and even 

 during downwelling plants were the second-ranked 

 food category (Table 3). Although these rankings are 

 inflated by undigested plant tissues, certain algal 

 materials appear to be important foods. The avail- 

 ability of plant foods to supplement prey shortages 

 was strongly seasonal, however. Thus, the sharply 

 reduced availability of plants during winter and 

 early spring undoubtedly contributed to the preva- 

 lence of empty stomachs among fish collected at 

 those times. 



740 



