During summer, the Chukchi Sea is supplied with relatively 

 warm water through the Bering Strait which results in ice-free 

 conditions over 50-100% of its area by late summer (Stringer 

 etal., 1980). Three distinct water types that conserve salinity, 

 but not temperature, flow northward through Bering Strait into 

 the Chukchi Sea (Coachman et ai. 1975; Aagaard, 1984; 

 NOAA, 1988). Anadyr water on the west is the most saline 

 (>33 ppt), followed by Bering Shelf water in the center, and 

 finally Alaska Coastal water (<31.5 ppt) on the east. Anadyr 

 water mixes extensively with Bering Shelf water and loses its 

 identity soon after entry into the Chukchi Sea (e.g., Grebmeier 

 etal.. 1989). During its northward flow, Alaska Coastal water 

 is augmented by additional freshwater inputs from Kotezebue 

 Sound. Siberian Coastal water intrudes from the East Siberian 

 Sea along the Siberian coast and is characterized by salinities 

 <32 ppt (Coachman et ai. 1975). The general flow from the 

 strait eventually veers eastward past Pt. Hope, steered in part by 

 bathymetry. Tidal amplitudes are small and do not 

 contribute significantly to water mass movements (NOAA, 

 1988). 



Seabird Censuses 



Seabirds were counted aboard the research vessel (RA') 

 AA"«£/e/?»A' A'o/o/ev between 8 and 15 August 1988. Shipboard 

 counts were conducted along transect lines running along the 

 Chukchi Sea Continental Shelf, generally in east-west directions 

 (Frontispiece). Sixty-two 10-min transect counts were 

 conducted in the southern Chukchi Sea portion of the joint 

 expedition. A total of 129.5 km- was censused at an average 

 ship speed of 15 knots (Table 1 ). 



Transects were taken while the ship was in motion and 

 used a 300 x 300 m zone (90° sector) extending forward and 

 abeam of the vessel on the side with the best viewing conditions 

 (Tasker et ai. 1984). Ship position (latitude/longitude) was 

 recorded at the beginning and end of each transect count. All 

 birds sitting or fiying within the transect were counted, but 

 sitting and flying birds were recorded separately. Counts were 

 conducted between oceanographic stations by a single observer 

 from the ship's flying bridge 12 m above the ocean surface. 

 Binoculars were frequently used to detect birds missed by the 

 unaided eye. Thick-billed (llria lomvia) and common 

 (U. aalge) murres, jaegers, and small dark alcids could not 

 always be identified to species. All data were logged on 

 standard USFWS project forms. 



In addition to the transect counts, 18 station counts were 

 obtained while the ship was stopped for oceanographic sampling. 

 Station counts recorded all birds seen within 600 m of the ship 

 during a 15-min period. 



Ecological research on the RA' Akademik Korolev was 

 undertaken in five ecosystems in the Bering and Chukchi Seas 

 ( Bering Sea Shelf, Bering Sea Slope, Gulf of Anadyr, Chirikov 

 basin, and southern Chukchi Sea). The oceanographic data 

 used in this analysis are from the southern Chukchi Sea, the 

 only region from which we had sufficient seabird transects and 

 station counts. Along the transect lines, the vessel was stopped 

 to deploy instruments for measuring physical and biological 

 attributes of the water at 32 stations (Stations 44-75, 

 Frontispiece ). A CTD hydrocast unit recorded depth, pressure, 

 water temperature, conductivity, salinity, and density (o, ). 

 Vertical cross-section profiles along transect lines were obtained 

 from programs developed at the Institute of Marine Science, 

 University of Alaska, Fairbanks, and Woods Hole 

 Oceanographic Institution. 



Analysis 



Relationships of seabirds to water masses were examined 

 by first categorizing all 10- and 15-min counts by the water 

 type in which they occurred. Ship position, sea surface 

 temperature, and salinity were used to place each transect 

 within the proper water mass. Seabird abundances were 

 expressed both as the number km - and hour counted. Numbers 

 of birds per 10-min transect were used as sample units in 

 multiple comparisons (Kruskal-Wallis rank sums analysis) 

 across all water mass types. When there were significant 

 overall differences, individual comparisons between water 

 mass types were made using Mann- Whitney {/-tests. If values 

 of U were outside limits of regular probability tables due to 

 large sample size, the approximate normal deviate Z was used 

 as the test statistic (Snedecor& Cochran, 1980). Zvalueswere 

 corrected for tied groups. 



Results 



Water Mass Distrihiitions 



Three distinct water masses were recorded within the 

 southern Chukchi Sea during August 1988 (Figs. 2-6). We 

 follow previous convention in designating these water mass 

 types (e.g., Coachinan e! ai. 1975; Coachman, 1987). 



TABLE 1 



Allocations of census effort by water mass type: 



SCW=Siberian Coastal water; BSAW=mixed Bering 



Shelf-Anadyr water; ACW=Alaska Coastal water. 



10-min transect counts 



15-min Station Counts 



sew 23 

 BSAW 24 

 ACW 1 5 



230 

 240 

 150 



15.0 

 15.0 

 15.0 



13-16 48.07 

 15-16 50.16 

 15 31.35 



2.09 

 2.09 

 2.09 



7 105 



8 120 

 3 45 



38: 



