Experimental Designs for Next Generation Phenotyping

Abstract

The increase in popularity of high-throughput genotyping in breeding programs is associated with recent advances in DNA sequencing technology and large decreases in genotyping costs. However, the limits of using genotyping for making predictions and, therefore, identifying potential candidate materials for selection thus reside in the quality of the phenotyping. High-throughput phenotyping technologies have been developed and implemented prior to planting and during cultivation. Much of this phenotyping has occurred in relatively small and restricted environments where many influential factors in the quality of phenotype can be adequately controlled. In many situations, however, it is necessary to perform phenotyping under field conditions. In this case, depending on the characteristic of interest to be collected, the influence of factors difficult to be controlled in such adverse conditions can cause the need for use of alternatives that can ensure a sufficiently accurate and precise phenotyping. In this sense, the science of Statistics contributes with an important role, either in the use of traditional basic concepts, in the planning of controlled experiments, or in modeling and developing appropriate analyzes. This chapter will discuss several experimental designs that can potentially be used for phenotyping under variable conditions, describing their various characteristics. Also it will address on topics related to the problem of obtaining accurate and precise phenotypic information, and the role of statistics in the success of this venture so fashionable today.