Applications of Principal Component Analysis to Horticultural Research

Horticultural researchers often must measure complex traits, such as vigor, reproductive performance, morphology, or adaptability, and develop relationships with treatments or associated variables. Complex traits, however, are a composite of individual traits that often vary together in response to imposed treatments or evolutionary pressures. Identifying a single variable representative of the complex trait may not be possible, so the researcher is faced with the possibility of separately examining many related variables. If the researcher uses univariate statistics to quantify differences or relationships, then the number of separate analyses required will equal the number of individual variables measured. With many types of biological data, however, correlation among variables is common (multicollinearity), and information provided by separate univariate analyses will be redundant.