serve as a very large reservoir for storage of anthropogenic 

 CO,. The physical, chemical, and biological mechanisms still 

 are not well understood. It has been suggested that the capacity 

 of the oceans to store CO, may have been overestimated 

 {Brewer etal.. 1989). However, much research still needs to be 

 conducted before an accurate understanding of the role of the 

 oceans emerges. TTie Bering Seaecosystem may be an important 

 area for storing carbon, both by burial in the sediments and via 

 transport to areas of deep-water formation. 



Materials and Methods 



Primary productivity measurements were carried out at 30 

 of the 1 13 stations occupied (Fig. 1). Standard '^C methods 

 using liquid scintillation counting were employed on samples 

 from two depths. One sample was collected near the surface, 

 the other from the subthermocline layer. The latter from the 

 light absorption maxima if one was identified by a SeaTek //; 

 situ beam transmissometer. Samples were collected with 8-1 

 Niskin samplers and samples from each layer were homogenized 

 in a 20-1 carboy prior to subsampling. Triplicate subsamples 

 were incubated with 2.5 ^ Ci NaH'^'CO, at each of eight light 

 intensities from 0.25 to 200 |a Ein m - s '. Fluorescent lamps 

 were used as the light source. Additional samples were kept in 

 total darkness and at ambient, sea-surface, natural light. 

 Incubations took place in water-cooled chambers at near- 

 ambient sea-surface temperature. After 1-h incubations, samples 

 were filtered through 0.4 |i m pore Gelman metricel filters. 

 Filters were acidified in vials with 0.5 ml 1 .0 N HCl and 

 counted in Ecolume scintillation cocktail (ICN Biomedicals, 

 Inc.). 



Primary production rates were calculated and normalized 

 to chlorophyll a concentrations (Strickland & Parsons, 1972). 

 Alkalinity and total CO, concentrations were estimated by 

 titration (Strickland & Parsons, 1972). Chlorophyll was 

 determined fluorometrically by G. Holmes and W. Robie at the 

 time ofsampling (Strickland & Parsons, 1972). Photosynthesis 

 versus light intensity (P-I) parameters were estimated with a 

 hyperbolic tangent function (Jassby & Piatt, 1976) and with a 



function incorporating photoinhibition (Platte/ a/., 1980). I'he 

 best fit model was determined from the sum of squares obtained 

 by using nonlinear least-squares regression (Systat, Inc.). 



Incident sea-surface light intensity was monitored 

 continuously with a photosynthetically active radiation (PAR) 

 sensor (Li-Cor LI-I92S) mounted on a post removed from 

 most light interferences and connected to a data logger (Li-Cor 

 LI- 1 000). Subsurface light was measured and the extinction 

 coefficients calculated at each station with a LI- 1 85 quantum 

 meter equipped with a LI-192S sensor (Li-Cor). 



The incident light, extinction coefficients, chlorophyll 

 versus depth distribution, and P-I parameters were used in a 

 numerical model to calculate integral production through the 

 water column throughout the day. The program was written in 

 Fortran 77 (Lahey Computer Systems) and could also be used 

 to calculate seasonal and annual estimates of production. 



Results 



Figure 2 shows incident surface light intensity (as photon 

 flux density). The maximum values ranged from about 350 to 

 1,900 |i Einm-s'. Storms were common near the Aleutians 

 and the Polar Front, which accounted for some of the PAR 

 variability. 



The areal primary production throughout the Bering and 

 Chukchi Seas was relatively high, with a mean of 

 1.8 g C m-d'. The maximum value estimated was 

 15.3 g C m-d ', while the low value was 0.174 g C m- d '. 



The distribution of primary production had several peaks 

 throughout the region, especially in shallow waters on the 

 shelf. Not all the shelfarea was productive, however. Some of 

 the lowest values recorded during this cruise occurred south of 



2000 n 



1500- 



Cfi 



M 



I 





1000 



500- 



ovm 



Day 



Fig. 1. Locations of stations where primary productivity was measured. pjg 2. Photon Hux density (15 min means) measurements taken during the 



cruise. July-August, 1988. 



219 



