Every plant science experiment starts with a design that will be adapted to answer a specific biological question and involves evaluation of phenotypic traits. Plant phenotyping has advanced from manual measurements of physiologically relevant parameters to high‐throughput phenotyping platforms that use robotics and imaging sensors. Yet, this game‐changing technology has its own challenges, namely data analysis and interpretation. The improved quality of the sensors used in the phenotying experiment provides increased understanding, however the insight provided on the research question is limited by the experimental design. Aspects such as replication or spatial variability are important to consider when designing the experiment conducted in highly controlled environment as well as under field conditions. With wider availability of cameras and other sensors, we are able to record increasing number of plant traits. This results in the phenotypic bottleneck moving from data acquisition to data analysis. Throughout this article, we present practical considerations and potential shortcomings of phenotyping systems and suggest some solutions to the challenges of plant phenotyping through streamlined and reproducible data analysis pipelines.
Key Concepts
Plant phenotypes are complex, resulting from the interaction between genotype and environment.
The phenotype can be divided into traits, for example, biomass can be dissected into leaf area, branches/tillers, fruits.
The relationship between traits depends on the environment, genotype and treatment.
Each phenotyping method is optimised to answer a specific research question.
Exploring the relationships between phenotypes and their changes across genotypes/treatments increases our understanding of the underlying physiological processes.
Experimental design should include an optimal number of replicates and sample randomisation to ensure a successful interpretation of phenotypic results.
Phenotyping results require detailed statistical analysis to be adequately interpreted.