Methods:

Bacteria were enumerated essentially as described in our EqPac data
documentation and the article in the EQPAC I DSR volume. Multiple
images of all samples were acquired, processed and analyzed with Zeiss
software. We used a new algorithm which removed most, but possibly not
all of the prochlorophytes, so these data represent heterotrophic
bacterial counts only. We will be comparing samples with Rob Olson and
David Caron to constrain the bacteria counts. We also include estimates
of the mean volume of cells in each sample.

Bacterial abundance times cell volume yields total biovolume per liter,
an index of bacterial biomass, to which carbon per unit volume factors
can (and will) be applied. All raw images have been archived and can be
made available by separate arrangement.

Bacterial production was estimated by the incorporation of 3H-thymidine
and 3H-leucine into cold, 5% TCA-insoluble extracts, essentially as
described in our earlier JGOFS work, although using the new method of
Smith & Azam (see the Azam data documentation for Process-1).
Production estimates are forthcoming.

Analytical precision for bacterial abundance and cell volume.

We do not routinely perform replicate determinations of bacterial
counts.  However, reanalysis of 92 slides from the March series gave
the following coefficients of variation (sd/mean) for bacterial
abundance and cell volume:

property    	mean  	st.dev.	CV  (%)	    n
-  -  -  -  -  -  -  -  -  -  -  -  -  
abundance*	0.53  	0.061 	11.7  	    92
cell volume 	0.037 	0.006	16.4	    92
-  -  -  -  -  -  -  -  -  -  -  -  -
*units for abundance are x 10^9 cells/liter. volumes are um3/cell;
these are mean values for 92 replicate determinations of each
property.


Hugh Ducklow				Phone: 804-684-7180
School of Marine Science		Fax:   804-684-7293
The College of William and Mary		Email: duck@vims.edu
Box 1346				Courier: Rte. 1208, Greate Road, 
Gloucester point, VA 23062-1346	        http://www.vims.edu/bio/microbial/