chamber from a valve at one end; a second sample is withdrawn at the end of 

 the experiment. An ambient bottom water sample is also taken, and a dark 

 bottle is filled with water at the start of the experiment and sampled at the 

 end. The samples are analyzed for metals and nutrients after filtration. Benthic 

 fluxes are calculated from the change in the concentration of a constituent in a 

 chamber, the mean height of the chamber (chamber volume/enclosed sediment 

 surface area), and the length of the experiment. Dark bottle "fluxes" are 

 calculated in the same maimer as benthic fluxes, taking the dark bottle 

 concentration at the end of the experiment as the final concentration, and the 

 ambient bottom water concentration as the initial concentration. The dark 

 bottle results are important for metals; because if they are equal to zero within 

 analytical uncertainties, they indicate that metal analyses are not seriously 

 affected by sample contamination during collection. 



Metal concentrations were measured with the techniques used for pore 

 waters. The ambient water column metal concentrations at Jamestown North 

 are about 10 ppb (0.2 nM) for Mn and Fe, 3 ppb (0.06 fiM) for Ni, 2 ppb (0.04 

 IdM) for Cu, and 0.1 ppb (.001 fM) for Cd. Precisions in individual metal 

 analyses are ±10%; hence uncertainties in fluxes are 40 juM m"^ day'^ for Mn 

 and Fe, 8 /uM m"2 day"^ for Cu and Ni, and 2 luM m'- day'* for Cd. Histograms 

 of dark bottle results at Jamestown (Figure 2-3) are roughly as predicted from 

 these errors. 



Results of experimentally determined fluxes of Mn""" and nutrients at 

 Jamestov^ North are given in Table 2-3, along with model fluxes for these 

 chemicals presented earlier. The model fluxes agree quite well with the 

 measured values, indicating that the system is well characterized. 



A histogram of Jamestown North metal fluxes is shown in Figure 2-4 , and 

 averages are tabulated in Table 2-4. Manganese fluxes are similar to the model 

 values. Pore water iron concentrations are similar to those of manganese, and 

 comparable fluxes are predicted. Observed iron fluxes are about an order of 

 magnitude lower than manganese (and predicted) fluxes; this is ascribed to 

 rapid oxidation of iron in the supernate following diffusion out of sediments. 

 Nickel, copper, and cadmium fluxes are predicted to be negligible (see 

 discussion of pore water values, above) and in fact measured fluxes are 

 generally equal to zero within the analytical uncertainty. 



From the concentration of constituents in the Bay, the average height of the 

 water column (taken as 10 m) and the fluxes, we can calculate doubling times 

 of metals in the Bay with respect to benthic fluxes. These values are given in 

 Table 2-4. Upper limits on Cu, Ni and Cd doubling times were calculated taking 

 the flux as less than or equal to the sum of the mean flux and one standard 

 deviation. Doubling times are to be compared with residence times of water in 

 Narragansett Bay of about one month. The results show that, in the 



18 



