•  
  •  
 

Abstract

An understanding of general forces affecting annual harvest is essential to the management of white-tailed deer (Odocoileus virginianus). A predictive model based on such factors would be valuable to managers. The relationship between 27 different variables and annual, legal deer harvest in Arkansas was evaluated for 1957-1986. Variables most affecting harvest were soybean acreage, hay acreage, number of days in the deer season, rain during the deer season, and total state timber production, total state pulpwood production, and deer harvest 2 years prior. Because significant autocorrelation and heteroscedasticity were present in the variables, log-linear, first differencing and non-linear quasi-Newton regression methods were used in addition to ordinary least squares. First differencing removed autocorrelation and heteroscedasticity, but fit was not acceptable (R2 = 0.710). Non-linear estimation of first differenced log transformed variables provided an acceptably high R2 (0.896) with high significance of the individual parameter estimators. Factors associated with habitat quality 2 years prior most affected present-year deer harvest.

Share

COinS