• Guldin, James M.
  Department of Forest Resources, University of Arkansas at Monticello
  
• Baker, James B.
  USDA Forest Service, Southern Forest Experiment Station, Monticello, AR
  
• Lockhart, Brian R.
  University of Arkansas at Monticello
  

Four soil-site evaluation methods that estimate the site index of loblolly pine (Pinus taeda L.) were tested for accuracy across a range of sites in southern Arkansas. Site index was measured on each plot using tree height-age methodology. Estimates of site index for each of 51 plots were made using (1) Zahner's field procedure, (2) Zahner's regression procedure, (3) Baker-Broadfoot's field procedure for sweetgum (Liquidambar styraciflua L.), and (4) a modification of the Baker-Broadfoot sweetgum field procedure for loblolly pine. The order of highest to lowest correlation between each method and measured site index was the unmodified Baker-Broadfoot method> modified Baker-Broadfoot method > Zahner field method > Zahner regression method. The most accurate method was the modified Baker-Broadfoot method, which estimated site index within +-10 feet of measured site index 9 times in 10.