The evolutionary dynamics of complex ecological traits - including multistate representations of diet, habitat, and behavior - remain poorly understood. Reconstructing the tempo, mode, and historical sequence of transitions involving such traits poses many challenges for comparative biologists, owing to their multidimensional nature and intraspecific variability. Continuous-time Markov chains (CTMC) are commonly used to model ecological niche evolution on phylogenetic trees but are limited by

more »

... e assumption that taxa are monomorphic and that states are univariate categorical variables. Thus, a necessary first step when using standard CTMC models is to categorize species into a pre-determined number of ecological states. This approach potentially confounds interpretation of state assignments with effects of sampling variation because it does not directly incorporate empirical observations of resource use into the statistical inference model. The neglect of sampling variation, along with univariate representations of true multivariate phenotypes, potentially leads to the distortion and loss of information, with substantial implications for downstream macroevolutionary analyses. In this study, we develop a hidden Markov model using a Dirichlet-multinomial framework to model resource use evolution on phylogenetic trees. Unlike existing CTMC implementations, states are unobserved probability distributions from which observed data are sampled. Our approach is expressly designed to model ecological traits that are intra-specifically variable and to account for uncertainty in state assignments of terminal taxa arising from effects of sampling variation. The method uses multivariate count data for individual species to simultaneously infer the number of ecological states, the proportional utilization of different resources by different states, and the phylogenetic distribution of ecological states among living species and their ancestors. The method is general and may be applied to any data expressible as a set of observational counts from different categories.