Meeting Abstract
Adaptive landscapes have inspired evolutionary research for nearly a century. Until recently such landscapes have mostly served as metaphors rather than quantitative frameworks for research. Current methods which utilize landscape frameworks for research primarily employ evolutionary modelling, usually fitting data to Ornstein-Uhlenbeck models to make inferences about adaptive peaks. Recently alternative methods have been developed which utilize combinations of performance surfaces – multivariate visualizations of relationships between phenotype and functional performance – to explore features of adaptive landscapes and explain the distribution of species in phenotypic space. I illustrate these new methods using data on turtle shell shape and information on performance for three shell functions – strength, hydrodynamic efficiency, and self-righting ability. The performance surfaces for these functions are given varying weights and then combined to obtain a set of predicted performance optima in shell shape space. The distribution of actual turtle shells in shape space significantly overlaps these optima, particularly when the distribution of both optimal and “near optimal” (performance values within 99% of the maximum) locations are considered. The performance surface methods outperform modelling-based approaches in locating reasonable adaptive peaks and explaining the shape of the phenotypic distributions of turtle shells. In addition, these methods provide information about the relative importance of the individual functions in guiding turtle shell evolution and potentially in determining fitness. Performance surface-based methods show great promise for allowing researchers to more directly connect functional performance with macroevolutionary patterns of diversification, and to explain the distribution of species across phenotypic space.
