IZEM, R*; MARRON, JS; KINGSOLVER, JG; UNC-Chapel Hill; UNC-Chapel Hill; UNC-Chapel Hill: Analyzing Variation in Thermal Performance Curves

A thermal performance curve (TPC) represents the performance of an individual or genotype as a continuous function of temperature. Patterns of genetic variation in thermal performance curves and other continuous reaction norms are important for determining evolutionary responses to selection; but identifying and interpreting such patterns of variation are often difficult with standard linear methods such as principal components or factor analysis. We describe a new statistical approach that decomposes variation in TPCs into three modes of variation of biological interest: vertical shift (Faster-Slower) mode, horizontal shift (Hotter-Colder) mode and generalist-specialist mode. The vertical shift mode of variation can be modeled with standard linear analysis, but effective analysis of the other modes of variation has motivated the development of a new non-linear methodology. This new method visualizes and quantifies the contribution of each mode of variation to the total genetic or phenotypic variation in the population. We illustrate the method using data on variation in TPCs for growth rate in Pieris rapae caterpillars. The analyses indicate that the generalist-specialist mode contributes substantially to broad-sense genetic variation in this population, but reveal little genetic variation in overall growth rate (vertical shift). The general approach should be applicable to other types of continuous reaction norms.