Daily values of evapotranspiration from a watershed in Volusia County, Florida, were estimated for a 2-year period (January 1998 through December 1999) by using an energy-budget variant of the eddy correlation method and a Priestley-Taylor model. The watershed consisted primarily of pine flatwood uplands interspersed within cypress wetlands. A drought-induced fire in spring 1998 burned about 40 percent of the watershed, most of which was subsequently logged. The model reproduced the 449 measured values of evapotranspiration reasonably well (r²=0.90) over a wide range of seasonal and surface-cover conditions. Annual evapotranspiration from the watershed was estimated to be 916 millimeters (36 inches) for 1998 and 1,070 millimeters (42 inches) for 1999. Evapotranspiration declined from near potential rates in the wet conditions of January 1998 to less than 50 percent of potential evapotranspiration after the fire and at the peak of the drought in June 1998. After the drought ended in early July 1998 and water levels returned to near land-surface, evapotranspiration increased sharply; however, the evapotranspiration rate was only about 60 percent of the potential rate in the burned areas, compared to about 90 percent of the potential rate in the unburned areas. This discrepancy can be explained as a result of fire damage to vegetation. Beginning in spring 1999, evapotranspiration from burned areas increased sharply relative to unburned areas, sometimes exceeding unburned evapotranspiration by almost 100 percent. Possible explanations for the dramatic increase in evapotranspiration from burned areas could include phenological changes associated with maturation or seasonality of plants that emerged after the fire or successional changes in composition of plant community within burned areas.

Variations in daily evapotranspiration are primarily the result of variations in surface cover, net radiation, photosynthetically active radiation, air temperature, and water-table depth. A water budget for the watershed supports the validity of the daily measurements and estimates of evapotranspiration. A water budget constructed using independent estimates of average rates of rainfall, runoff, and deep leakage, as well as evapotranspiration, was consistent within 3.8 percent. An alternative water budget constructed using evapotrans-piration estimated by the standard eddy correlation method was consistent only within 9.1 percent. This result indicates that the standard eddy correlation method is not as accurate as the energy-budget variant.