Non-destructive estimation of leaf area of Citrus varieties of the Kotra Germplasm Bank

Recently, mathematical modeling and computer expertise are advancing hastily. Their progression has been smooth sailing. The advancements have expedited and speeded up our scientific analyses. Hence, it is fruitful and essential to take advantage of the opportunities. Leaf area is among the most important plant properties which are directly related to ecological and physiological variables of a plant including leaf area index, light interception, evapotranspiration, photosynthesis, and growth. Thus, its calculation is extremely important. In this study, leaf area of species typica tress in Citrus and Subtropical Fruits Research Institute of Iran named Kotra Germplasm Bank include Orange (Citrus sinensis), Mandarin (Citrus reticulata), Lime (Citrus aurantifolia), and Lemon (Citrus lemon) were estimated using a non-destructive method Artificial neural network (NN) and by measuring quantitative leaf variables including width, length and a combination of width and length. For this purpose, four genera from each species were chosen and 200 leaves from different parts of their crown were collected. The width and length of the leaves were measured in the lab using a ruler, and their area was measured by a leaf area meter. This disquisition answered if GMDH-type NN was able to be applied to assess the area of the leaf as deferent according to particular variables consisting of a leaf with and leaf length. The average width, length, and area of leaves values significantly differed among the studied species as per the results.GMDH type NN provides a thriving tool for efficient detection of the model in data and precisely anticipating a proceeds indicator based on search input data and it’s able to be used to predict leaf area according to width and length.