James Hagy
UM-CEES Chesapeake Biological Lab
P.O. Box 38
Solomons, MD 20688
hagy@cbl.cees.edu
Our work is directed toward developing an integrated empirically-based profile of the physical, chemical, and biological properties of the Patuxent River ecosystem, including the responses of these properties to external forcing.
The substantial base of information available suggests that this approach may be particularly effective for Patuxent River. The most substantial data source is the greater than 10-year record of spatially and termporally detailed water quality and other information gathered by the Chesapeake Bay Monitoring Program. Many other research programs and small short-term studies have provided unique opportunities for comparison as well as information needed to develop accurate conceptual models of the ecosystem.
The nutrient budget for Patuxent River described by Boynton et al. (Estuaries 18:1B, 1995) is one example of an integrated empirical ecosystem analysis. The mass balance approach central to nutrient budgets is useful in that reasonable bounds can be placed on even uncertain estimates of individual components of the balance since the inputs and outputs must sum to zero. Thus, as more information is learned about one component of the ecosystem, additional information is also implied about the other components. Of course, this analytical 'check' is also available to other modeling approaches.
Box modeling is a logical extension of the mass balance, although box models preceded nutrient budgets. Box models use the balance of salt and water transport to estimate rates of advective and non-advective exchange, and subsequently use these estimates to hind-cast, by mass balance, net biogeochemical fluxes with some degree of spatial and temporal resolution. The underlying concept is that as more relationships are drawn among the observed ecosystem parameters, the amount of information gained indirectly about the ecosystem grows, and the uncertainty surrounding any individual estimate is constrained by the others. Our box modeling effort draws heavily upon the monitoring program data and is one example of a productive synthesis that is possible only after a persistent, aggressive, and high-quality monitoring effort.
Thus far, the box models have generated estimates of net production (or consumption) of dissolved inorganic nitrogen, dissolved silica, bottom-layer dissolved oxygen, and organic carbon (net ecosystem metabolism). They have also provided some information on the issue of new versus regenerated nitrogen sources, the relative role of Patuxent River and Chesapeake Bay as nitrogen sources, the role of denitrification, and controls on hypoxia. Physical transport estimates based on the box models have quantified residence times as a function of space, time, and Patuxent River discharge rate. Many of the estimates were made for successive years, making it possible to evaluate in retrospect the effect of changes in nutrient loading.
In future work, we plan to draw more information into the analytical process. Much of this information is already available. For example, 14C-based plankton production estimates from the monitoring program, sediment nutrient and oxygen fluxes from the monitoring program, and an open water metabolism study at Benedict (B. Sweeney) each provide independent estimates of ecosystem processes. An upcoming deployment of the Chesapeake Bay Observing System Rover buoy at Broomes Island and a current sediment mapping effort will probably also contribute substantially.