This role of density stratification is illustrated by the 
data from the Bay in 1977 shown in Slide 1. As the water-column 
density gradient, expressed as change in sigma-t over change in 
depth, increases from February to June to oxygen concentration 
in the lower layer decreases from near saturation to anoxia. 
Even in February, when temperatures are still rather low, if 
stratification increases there are enough "natural agencies" 
consuming oxygen in the deep water that oxygen concentration 
declines. Slide 2 shows the relationships among the annual 
cycles of deep water tmperature, oxygen concentration and 
salinity, and change in density over change in depth during 1964 
to 1966 and 1969 to 1971 in the middle portion of Chesapeake 
Bay. Again, we see that through the annual cycle, the oxygen 
concentration in the deep water is generally low or zero in 
summer and early fall and increases at otheer times of the 
year. The oxygen plot in Slide 2a shows reoxygenation events in 
August of both 1964 and 1965, after which the deep layer again 
lost oxygen before the surface water temperature decreased to 
produced seasonal reoxygenation. Short-term reoxygenation 
during summer was not observed in the data set for 1969 to 1971. 
The emerging picture is that anoxia has been a recurrent 
feature with varying intensity. There are other observations, 
particularly about fisheries and submerged aquatic vegetation, 
in the Chesapeake Bay which lead us to believe that the system 
is under stress. However, we are faced with a difficult 
interpretational problem because the data are spotty in time and 
space. We must make some judgements about the quality of the 
data and how to use them. Also, the mechanism is more complex 
than presented by Sales and Skinner. Slide 3 shows two graphs 
of change in dissolved oxygen concentration vs. change in 
salinity over the same depth from the upper to the lower layer. 
Each datum point in Slide 3a represents a vertical profile in 
the mid-bay during the month of July in ten different years 
between 1949 and 1980. Linear regression analysis yields an 
equation for the line through the points which has a regression 
coefficient of 0.87. If we assume that the vertical structure 
of the water column and its processes can reach a steady-state 
in the summer, then the line in Slide 3a represents the steady- 
state relationship between the salinity gradient and the 
dissolved oxygen gradient. This further implies steady-state 
between oxygen utilization in the deeper layer and its resupply 
via the surface. If we apply this model to other seasons, we 
might be able to test for a steady-state in the utilization and 
resupply processes. 
90 
