™ 5 4000 



S 9 3000 



Fig. 4. Average zooplanklon abundance (mean nuniber/waler lype plus/ 

 minus one standard error) in each of three water types in the northwest 

 Bering Sea. Abbreviations are as indicated in Fig. 3. 



Auklet distributions (i.e., all auklet species considered 

 collectively) did not differ significantly between ASW and 

 BSW when the distributions of those water types were 

 considered at the surface, and at 5, 10, and 15-m depths, 

 respectively (see Fig. 5). Mean abundances of auklets in both 

 these water types were significantly greater than the mean 

 abundance of auklets in CCW, however. This pattern changed 

 at the 20-m depth; auklets were significantly more abundant in 

 ASW than BSW (Table 3). Chukchi Coastal water was not 

 observed anywhere in the cruise track at the 20-m depth. 



Discussion 



hnplicit in all studies of this type are the assumptions that 

 physical features of the pelagic habitat in which birds forage 

 may enhance prey availability and that birds can exploit these 

 features in some way to promote foraging efficiency. In this 

 study, seabirds are unevenly distributed among the water types 

 detected at the surface in the northwest Bering Sea. At these 



large spatial scales, it appears that seabirds are foraging 

 selectively in water types in which their preferred prey type 

 may be most abundant. Planktivorous birds preferred ASW, 

 which contains the greatest abundance of large oceanic 

 zooplankters. Piscivores, in turn, preferred CCW, which may 

 have the greatest abundances of fish of the three water types. 

 This last conclusion is limited by a lack of information about 

 prey-sized fish distribution in this area at this time, but what 

 data are available suggest that oceanographic conditions in 

 ACW (low salinities, shallow depths, high stability) provide 

 better habitat for small/juvenile fish than BSW or ASW. In that 

 CCW also has these properties, it may also be assumed to be 

 good prey fish habitat. 



How the different water types affect foraging success via 

 specific foraging strategies requires further investigation. Where 

 AS W is observed at the surface, it may be upwelling zooplankters 

 that would otherwise be unavailable to surface-foragers. Birds 

 employing this strategy are known to congregate at mesoscale 

 features such as convergent fronts that concentrate prey (Brown, 

 1980). Further, they are among the most mobile and far- 

 ranging foragers of the seabirds included here, and are thereby 

 least limited to a single broad domain. 



Deep-feeding birds might be expected to be found in water 

 types in which vertical structure concentrates prey and thereby 

 reduces energetic costs of diving (Hunt et ai, 1990). Deep- 

 feeding birds in this study were found in highly-stratified areas 

 where a lens of CCW overlay BSW or ASW. Because deep- 

 feeding birds are exploiting a foraging niche unavailable to 

 surface- or shallow-feeders, their distributions might be related 

 to distributions of preferred subsurface waters. Planktonic 

 prey abundance and distribution can be better assessed than 

 those of fishes with the sampling methods used here, so the 

 relationships between planktivorous, deep-feeding auklets and 

 zooplankton was used to explore this idea. However, when 



TABLE 3 



Auklet abundance ( x /km- ) in each water 



type in the northwest Bering Sea in July 



and August 1988. Water mass positions 



are estimated at five depths: 0, 5, 10, 



15 and 20 m. 



' Water type positions did not differ appreciably at depths 



above 15 m. 

 * p < 0.05, determined by Kruskall-Wallis test. 



393 



