Pheromones as mixtures: geometric designs and response surface modeling for optimization of mating disruption of the citrus leafminer Phyllocnistis citrella

Authors: Lapointe, Stephen; STELINSKI, LUKASZ - University Of Florida

Submitted to: Meeting Proceedings
Publication Type: Proceedings
Publication Acceptance Date: 7/15/2009
Publication Date: 10/27/2009
Citation: Lapointe, S.L., Stelinski, L. 2009. Pheromones as mixtures: geometric designs and response surface modeling for optimization of mating disruption of the citrus leafminer Phyllocnistis citrella. Meeting Proceedings. Fifth Asia Pacific Conference on Chemical Ecology, October 27-30, 2009, Honolulu, Hawaii.

Interpretive Summary: Insect sex pheromones can be used to disrupt the mating behavior of pest species and thereby reduce their populations and the damage they cause without recourse to pesticides. Insect sex pheromones are often blends of one or more compounds that are produced in certain ratios by the insect to attract mates. Because blends are mixtures, studies aimed at identifying optimal proportions of blends to achieve mating disruption must use special experiment designs. In mixtures, the ingredients must total to a constant value. Proportion designs are constrained because the percentage of one ingredient is defined by the percentage of the other ingredient. Experiment designs developed to address mixtures are ideally suited for many areas of experimental biology including pheromone blend studies because they address the confounding of proportionality and concentration intrinsic to other commonly-used designs. If the response variable is a function of the proportionality of components, then mixture designs allow for complete exploration of the design space as well as analysis of individual component and nonlinear blending (interaction) effects. A proportion/amount design coupled with response surface modeling (RSM) allowed us to identify optimal blends of a two-component pheromone system for attraction and orientation disruption of the leafminer moth, Phyllocnistis citrella, a major pest in citrus growing areas around the world. It has been suggested that natural pheromone blends produced by the insect will be the most effective blend for purposes of mating disruption. However, given the experimental approaches employed in this field to date, it is not possible to conclude that optimal blends for disruption have been identified in most cases. RSM and mixture designs represent a highly efficient method that allow chemical ecologists to completely explore the experiment design space and thereby identify optimal blend ratios and dosages for mating disruption. If this approach were extended to more species, we believe a more complete and possibly different understanding might evolve concerning how mating disruption works.

Technical Abstract: Pheromone blends composed of one or more components are mixtures. Mixtures present a unique problem for experiment design when response (e.g., insect attraction to a blend) is a function of proportionality of the blend's components. In mixtures, the ingredients must total to a constant value. Proportion designs are constrained because the percentage of one ingredient is defined by the percentage of the other ingredient. Scheffé polynomials also differ because they do not have an intercept (the concept has no application in mixtures). Experiment designs developed to address mixtures are ideally suited for many areas of experimental biology including pheromone blend studies because they address the confounding of proportionality and concentration intrinsic to factorial and one-factor-at-a-time designs. If the response variable is a function of the proportionality of components, then mixture designs, but not factorials, allow for complete exploration of the design space as well as analysis of individual component and nonlinear blending (interaction) effects. A proportion/amount design coupled with response surface modeling (RSM) allowed us to identify optimal blends of a two-component pheromone system for attraction and orientation disruption of the leafminer moth, Phyllocnistis citrella, a major pest in citrus growing areas around the world. It has been suggested that natural pheromone blends produced by the insect will be the most effective blend for purposes of mating disruption. However, given the experimental approaches employed in this field to date, it is not possible to conclude that optimal blends for disruption have been identified in most cases. RSM and mixture designs represent a highly efficient method that allow chemical ecologists to completely explore the experiment design space and thereby identify optimal blend ratios and dosages for mating disruption. If this approach were extended to more species, we believe a more complete and possibly different understanding might evolve concerning how mating disruption works.