Experimental Design for Plant Improvement

Abstract

AbstractSound experimental design underpins successful plant improvement research. Robust experimental designs respect fundamental principles including replication, randomization and blocking, and avoid bias and pseudo-replication. Classical experimental designs seek to mitigate the effects of spatial variability with resolvable block plot structures. Recent developments in experimental design theory and software enable optimal model-based designs tailored to the experimental purpose. Optimal model-based designs anticipate the analytical model and incorporate information previously used only in the analysis. New technologies, such as genomics, rapid cycle breeding and high-throughput phenotyping, require flexible designs solutions which optimize resources whilst upholding fundamental design principles. This chapter describes experimental design principles in the context of classical designs and introduces the burgeoning field of model-based design in the context of plant improvement science.