Adaptive identification of environmental stress for the management of plant growth

Abstract

A model of a plant as a player in a game against an uncertain environment is presented using J.L. Harper's conception of a plant as a set of populations of different types of modular units. Following D.F. Rhoades' conception of a plant's production of secondary chemical compounds (tannins and lignins etc.) for defence against variable environmental contingencies including herbivory, physical and weather stress, we postulate n types of modular units each producing a secondary substance which will serve as n control variables for the plant (or plant manager) in this semi-game. It is shown that an auxiliary model called a predictor can be designed which, using read-out data on amounts of produced chemicals, tracks or identifies, in adaptive fashion, the environmental variations. Such identification is advantageous from the management point of view because production of these chemicals is metabolically expensive and consequently identification may help reduce energy expenditures. The mathematical method uses Antonelli's theory of Volterra-Hamilton systems, control theory of deterministic systems and, in particular, the nonlinear identification theory of Skowronski.