Sensing tree for yield forecasting under different irrigation.

Abstract

Tree itself is assumed to be a better indicator of water stress. Sensing of plant behavior in relation to leaf physiology, plant water status and canopy reflectance are the major factors indicating the water need of the trees. In this study, different response factors (leaf physiological parameters, leaf nutrients, leaf water content and canopy reflectance) of citrus tree have been observed under differential water stress condition by supplying deficit irrigation and fruit yield has been forecasted based on these factors. For the first year a yield response model has been formulated employing principal component regression (PCR) methodology and the model has been validated for second year data. Among different factors, leaf-N, leaf-K, stem water potential stress index, stomatal conductance and water band index have been found as the best predictors for yield and resulted higher accuracy ( ) in yield prediction of citrus tree. Overall, the study reveals that sensing tree is one of the better options to quantify water stress for efficient irrigation scheduling and to get target yield from orchards.