Yield forecasting in cardamom (Elettaria cardamomum Maton) plantations under intensive management

Abstract

A study was undertaken to develop a model for forecasting the yield in cardamom plantations under intensive management. Thirteen biometrical characters namely tillers per clump, tiller height leaves per tiller, vegetative buds per clump, bearing tillers per clump, panicles per clump, panicle length, racemes per panicle, capsules per raceme, seeds per capsule, leaf length, leaf breadth and recovery percentage were chosen as explana tory variables and they exhibited a precision of about 82%. Step down regression resulted in the retention of only four characters namely, panicles per clump, racemes per panicle, capsules per raceme and leaf breadth with which yield can be estimated with around 77% precision.