incorptiration rates ranging from 1 .4 pinoles 1 ' h ' in coastal 

 waters with temperatures of 6.5°C to rates of 4.7pmoles 1 ' h ' 

 at the shelf break with temperatures of 7.5-1 0°C. In the Celtic 

 Sea where water-column temperatures varied from 8 to 15°C. 

 thymidine rates ranged from 0.24 to 0.81 pmoles (values 

 converted from data given in Table 3 in Joint & Pomroy. 1983). 

 In coastal shelf waters off NW Spain with temperatures of 

 I0-I8°C, rates of 0.1 to 10.1 pmoles 1' h' were reported 

 (Hanson ei al., 1986a; Hanson et al.. accepted). In other 

 temperate waters, rates ranged from 0. 1 to 20 pmoles I ' h ' for 

 California Coastal waters (calculated from data in Table 1 in 

 Fuhrman ct al.. 1980) and for southeastern US shelf waters 

 (Hanson t-i al., 1988). Therefore, results from the Chirikov 

 basin and surface waters of the south Bering Sea show that 

 bacterioplankton during this late summer period appeared as 

 productive as bacterioplankton on many continental shelves 

 and oceanic ecosystems in northern temperate and southern 

 polar regions. 



Bacterioplankton are the most abundant group of marine 

 organisms in pelagic communities, yet the least understood in 

 regard to population structure, function, and interaction with 

 other pelagic communities in marine food webs. Total 

 bacterioplankton counts varied little with water type on the 

 north Bering Sea Shelf (overall 4.2 x 10"! 0.2 [S.E.J cells 1 ' ). 

 The highest density of bacterioplankton occurred in the surface 

 waters ofthe south Bering Sea (about 1.3x 10''cells I ')■ These 

 densities are quite similar to values reported for other polar or 

 subpolar regions (Fuhrman & Azam, 1980, 1982; 

 Hanson et al., 1983; Garrison et al.. 1986; Pomeroy & 

 Deibel, 1986; Douglas et al.. 1987; Kottnieier & Sullivan, 

 1987). 



Estimate of Bacterioplankton Productivity and Growth Rate.s 

 Because of the uncertainty in the proportion of 

 bacterioplankton that use thymidine for DN A synthesis relative 

 to total metabolically active cells ( Douglas c/t//., 1987), we can 

 only estimate the productivity of the bacterioplankton in the 

 Bering Sea ecosystem. Our estimates are based on a theoretical 

 conversion factor of 2 X 10"* cells produced (mole of thymidine 

 incorporated) ' (Fuhrman & Azam, 1982), the accuracy of 

 which depends on a number of assumptions that have been 

 discussed previously (Fuhrman & Azam, 1982; Ducklow & 

 Hill. 1985; Douglas e/rt/., 1987). 



Empirically derivedCF'sgenerally range 1 to5x lO'^cells 

 produced (mole of thymidine incorporated) ' (Kirchman «'?«/., 

 1982; Riemann et al., 1984, 1987; Ducklow & Hill, 1985). 

 Acknowledging the relative accuracy ofthe theoretical CF. we 

 applied the theoretical CF and report the productivity of the 

 bacterioplankton in the Chirikov on the order of 1 -5 x I O*" cells 

 produced 1 ' h ', average 3 x 10" cells 1 ' h ' (Table 3). These 

 rates of cell productivity in the Chirikov basin are on the same 

 order as rates measured in other high and low latitude 

 ecosystems. In McMurdo Sound and the Ross Sea. Antarctica, 

 Fuhrman and Azam ( 1980) estimated cell productivity rangmg 

 from<(). 1 to 21 X 10" cells 1' h' (rates adjusted 1.54 times; a 

 theoretical CF of 1.3 x 10"* cells (mole of thymidine 

 incorporated] ' was originally applied to thymidine 

 incorporation for cell productivity estimates). 



TABLK 3 



Bacterioplankton production (mg carbon m - d '), bioinass 



(g carbon m -), growth rate (u, d '). and doubling time 



(Ln 2/u,days) in the Chiriko\ basin and south Bering Sea. 



August 1 988. Bacterioplankton production based on estimates 



from thymidine (Thy) incorporation and frequency of 



dividing cells (FDC). 



Dividing-cell productivity by the total number of 

 bacterioplankton, an estimate of the specific growth rate of 

 bacterioplankton population can be calculated. Specific growth 

 rates in the three water types in the Chirikov basin are given in 

 Table 3. Growth rates were not significantly different across 

 the basin. Rates averaged 0. 1 8 day ' ( or a population doubling 

 time of roughly 5 days). The doubling time of 5 days is similar 

 to the doubling times reported for temperate coastal and shelf 

 waters ( 1 to 4 days, Fuhrman & Azam, 1982; 4 days. 

 Joint & Pomroy. 1983; 0.8 to 10 days, Hanson et al., 1986b, 

 1988). 



Assuming a thymidine-active subpopulation of 50% ofthe 

 total number of bacterioplankton in the Chirikov basin, the 

 doubling time of this subpopulation is 2.5 days. The doubling 

 time of the thymidine-active bacterioplankton in Canadian 

 Shelf waters off Nova Scotia ranged from 0.5 to 1.2 days 

 (Douglas et al., 1987). Thus, mean growth rate for 

 bacterioplankton of high latitude ecosystems are in general 

 comparable to rates calculated for bacterioplankton in low 

 latitude environments. 



Hagstrom et al. ( 1979) proposed a frequency of di\'iding 

 cells (FDC) method to estimate bacterioplankton growth rates 

 without incubation and radioactive organic substrates. 

 Theoretical consideration and empirical evidence have shown 

 that the frequency of cells in the dividing state is proportional 

 to the growth rate ofthe population ( Newell & Christian, 1981; 

 Larsson & Hagstrom, 1982; Hanson ('/«/., 1983). Tocalculate 

 growth rates by the FDC technique, a basic assumption is that 

 all cells are metabolically active. But because of inactive cells 

 in the population. FDC values underestimate the growth state 

 of the active population. Thus, estimates of bacterioplankton 

 growth rates using FDC error conservatively. 



73 



