Interpreting phenotypic variation in plant allelochemistry: problems with the use of concentrations.

Ecologists often use concentrations of defensive compounds as measures of plant allocation to defence and/or allelochemical production. I demonstrate that this practice may lead to erroneous conclusions because plants produce and allocate molecules (quantities) of compounds whereas concentrations reflect the distribution of these quantities in plant tissues and are, therefore, functions of plant biomass. As a tool for distinguishing between shifts in allelochemical production versus changes in plant biomass in determining allelochemical concentrations, I suggest using a technique known as graphical vector analysis (GVA) which has been developed for diagnosing nutrient limitations in forest stands, but has seldom been applied by researchers studying plant allelochemicals. I used data from several published studies to demonstrate how GVA can be applied to interpret ontogenetic and environmental effects on allelochemical levels and to compare the results obtained for different allelochemical types, plant species, treatments and experiments. These examples show that changes in plant biomass per se are an important source of variation in allelochemical concentrations and, therefore, concentration data can be easily misinterpreted if changes in absolute content and plant biomass are not considered simultaneously. Because studies reporting variation in allelochemical concentrations have been considered as tests for general theories of plant chemical defence, evidence in support of or against these theories should be re-examined using multivariate techniques such as analysis of covariance, allometric analysis and GVA.