Jointly Modeling the Evolution of Discrete and Continuous Traits

Abstract

Whether modeling the evolution of a discrete or continuous character, the focal trait of interest does not evolve in isolation and require comparative methods that model multivariate evolution. Progress along these lines has involved modeling multivariate evolution of the same class of character and there are fewer options when jointly modeling traits when one character is discrete and the other is continuous. Here we develop such a framework to explicitly estimate the joint likelihood for discrete and continuous characters. Specifically, our model combines the probability of observing the continuous character under a generalized OU process with the probability of the discrete character under a hidden Markov model, linked by a shared underlying regime. We use simulation studies to demonstrate that this approach, hOUwie, can accurately evaluate parameter values across a broad set of models. We then apply our model to test whether fleshy and dry fruits of Ericaceae lineages are correlated with their climatic niche evolution as represented by the aridity index. Consistent with expectations, we find that dry fruits have higher rates of climatic niche evolution, that the climatic niche of fleshy fruits is more conserved, and dry fruits have a more humid climatic optimum.

A trait‐based approach to determining principles of plant biogeography

“…Phylogenetic regression methods that allow the error term to be modeled according to different assumptions of how continuous traits evolve (Ho and Ané, 2014) and shift detection methods that allow the parameters of continuous trait evolution models to vary across the phylogeny (Uyeda and Harmon, 2014; Khabbazian et al, 2016) also allowed for more biologically realistic pictures of continuous trait evolution. Models that jointly estimate discrete and continuous trait evolution (Tribble et al, 2021; Boyko et al, 2022) allow for traits and environmental variables to influence one another throughout evolution in cases where they are correlated. Similarly, trait‐dependent biogeographical models that jointly estimate trait and range evolution (Sukumaran and Knowles, 2018; Quintero and Landis, 2020) allow dispersal parameter estimates to be conditioned on the presence of a certain trait.…”

The evolutionary responses of life‐history strategies to climatic variability in flowering plants

“…Here, we assess the relationship between climatic factors and the evolution of life‐history strategies in flowering plants. To that end, we apply recent developments in trait evolution models (Boyko et al ., 2023) to explicitly incorporate climatic niche variation's impact on life‐history strategy evolution. We account for the heterogeneity of evolutionary histories in flowering plants and the habitats associated with them by analyzing a broad sample of clades with global distribution and where multiple transitions between annual and perennial strategies are observed.…”

Long-term responses of life-history strategies to climatic variability in flowering plants

“…First, we wanted to accurately model correlations between climatic niche evolution and life history characters within each of our 32 clades. This was done by fitting a set of 15 hOUwie models with 100 stochastic mappings per iteration and adaptive sampling enabled (Boyko et al 2022). hOUwie is a recently developed framework that explicitly models the joint evolution of discrete and continuous characters.…”

“…The parameters we allowed to vary in our model are rates of transition between annual and perennial ( q ), the phenotypic optima of the climatic niche ( θ ), and the rate of climatic niche evolution ( σ 2 ). This means we analyzed BMV, OUV, OUM, and OUMV type models, as well as BM1 and OU1 (Boyko et al 2022). We conducted model-averaging and compared parameter estimates within hOUwie to test for: (1) a relationship between climatic optima and life history strategy, and (2) whether evolutionary rates of annuals are greater than those of perennials across all climatic variables.…”

“…Here, we assess the relationship between climatic factors and the evolution of life history strategies in flowering plants. To that end, we apply recent developments in trait evolution models (Boyko et al 2022) to explicitly incorporate climatic niche variation’s impact on life history strategy evolution. We account for the heterogeneity of evolutionary histories in flowering plants and the habitats associated with them by analyzing a broad sample of clades with global distribution and where multiple transitions between annual and perennial strategies are observed.…”