Department of Statistics, University of Nebraska, Lincoln, NE, USA.
Department of Statistics, Iowa State University, Ames, IA, USA.
Methods Mol Biol. 2022;2539:57-68. doi: 10.1007/978-1-0716-2537-8_7.
It is essential that the scientific community develop and deploy accurate and high-throughput techniques to capture factors that influence plant phenotypes if we are to meet the projected demands for food and energy. In recognition of this fact, multiple research institutions have invested in automated high-throughput plant phenotyping (HTPP) systems designed for use in controlled environments. These systems can generate large amounts of data in relatively short periods of time, potentially allowing researchers to gain insights about phenotypic responses to environmental, biological, and management factors. Reliable inferences about these factors depends on the use of proper experimental design when planning phenotypic studies in order to avoid issues such as lack of power and confounding. In this chapter, the topic of experimental design will be discussed, from basic principles to examples specific to controlled environment plant phenotyping. Examples will be provided based on the package agricolae in the R statistical language.
如果我们要满足对食物和能源的预计需求，科学界就必须开发和部署准确、高通量的技术来捕捉影响植物表型的因素。鉴于这一事实，多个研究机构投资于自动化高通量植物表型（HTPP）系统，这些系统专为在受控环境中使用而设计。这些系统可以在相对较短的时间内生成大量数据，这可能使研究人员能够深入了解对环境、生物和管理因素的表型反应。要对这些因素进行可靠的推断，就需要在计划表型研究时使用适当的实验设计，以避免出现缺乏效力和混杂等问题。在本章中，将讨论实验设计的主题，从基本原则到具体的受控环境植物表型设计示例。将基于 R 统计语言中的 agricolae 包提供示例。
