Abstract
The length of the midrib, petiole and maximum width of the lamina of 300 sour orange leaves were used to develop leaf area prediction models. Four models exhibited good predictive ability as indicated by R2- which ranged 0.89-0.93. However, two of these models were identified as being efficient for field purposes. These were A= -14.95 + 9.96(W) and A= 0.79 + l.45 (W2) where A is the estimated leaf area, W is the maximum lamina width (cm). These two regression models afford an accurate, unsophisticated, non-destructive means of estimating total leaf area of container-grown sour orange rootstocks. The four models were less accurate in estimating the leaf areas of ten-year-old Citrus aurantium trees.
