Conclusions While the nutrient budget gives a general idea of the major sources of nitrate in the basin, it is still limited in its ability to determine the amounts of pollution that actually enter the groundwater and end up in the springs. Further analysis in this area could include various types of modeling to predict actual nitrate flows to groundwater from all sources in the springshed. Using spatial and hydrological data to predict areas in the springshed that are most vulnerable to groundwater contamination would also be useful to managing nutrients in the springshed. Further study of the karst systems and the movement of groundwater in the springshed could also provide insight into how nutrients are transported from the landscape to the springs. The results of the CVM survey imply that Columbia County residents place a substantial value on water quality in the Ichetucknee River and are willing to pay for and participate in a program to improve it. Using CVM is one of the few ways to place a value on a good that is otherwise completely unknown. This information is important for policymakers to have in order to make informed decisions that affect the water resources that citizen's depend on for drinking water, tourism, and recreation (Loomis, 2000b). The results of this study help inform the debate over the allocation of funds for groundwater protection, and should provide useful information for policymakers considering stormwater planning, land-use changes in the county, and approval for increased septic tanks permits. The results also imply that if a referendum program of this nature were placed on a ballot in Columbia County, it could face a fair chance of passing. If a flat tax increase for the entire county is unfeasible, one possible action might be to create a special taxing district along hydrologic boundaries, to better ensure that the people who are most drastically impacting the resource are the ones who pay more to improve it. While this study provides an estimate of