White-tailed deer (Odocoileus virginianus) harvests from wildlife management areas in northeastern Georgia were compared with habitat variables, population variables from the previous fall harvest and hunters/km2 • Three years of harvest data from 8 Blue Ridge Mountain areas and 12 years of data from an Upper Piedmont area were analyzed. A significant stepwise multiple regression model (R2 = 0.57) related deer harvest numbers to the following habitat variables for the mountain areas: number and hectares of agricultural food plots, kilometers of roads, hectares of oak timber (all positive), and hectares of clearcuts (negative). On the Piedmont area, similar analyses yielded a model (R2 = 0.60) with number of bucks harvested related to number of hunters, number of food plots, food plot hectares (all positive), pine timber hectares (negative), and percent population harvested the previous year (negative). Despite a food plot component of only 0.13% of the total land area, much of the variability in the mountain deer harvest was explained by food plots and associated road access. These components were possibly responsible for 4 additional deer harvestedlha of food plot and approximately 40% of the current deer harvest. The Piedmont harvest was influenced by similar habitat factors plus hunting pressure and deer population characteristics from the previous year. Overall, analyses indicated that an increase in food plot hectares and hunter access in both study areas increased deer harvest.
