Rates of morphological evolution in Captorhinidae: an adaptive radiation of Permian herbivores

Abstract

The evolution of herbivory in early tetrapods was crucial in the establishment of terrestrial ecosystems, although it is so far unclear what effect this innovation had on the macro-evolutionary patterns observed within this clade. The clades that entered this under-filled region of ecospace might be expected to have experienced an "adaptive radiation": an increase in rates of morphological evolution and speciation driven by the evolution of a key innovation. However such inferences are often circumstantial, being based on the coincidence of a rate shift with the origin of an evolutionary novelty. The conclusion of an adaptive radiation may be made more robust by examining the pattern of the evolutionary shift; if the evolutionary innovation coincides not only with a shift in rates of morphological evolution, but specifically in the morphological characteristics relevant to the ecological shift of interest, then one may more plausibly infer a causal relationship between the two. Here I examine the impact of diet evolution on rates of morphological change in one of the earliest tetrapod clades to evolve high-fibre herbivory: Captorhinidae. Using a method of calculating heterogeneity in rates of discrete character change across a phylogeny, it is shown that a significant increase in rates of evolution coincides with the transition to herbivory in captorhinids. The herbivorous captorhinids also exhibit greater morphological disparity than their faunivorous relatives, indicating more rapid exploration of new regions of morphospace. As well as an increase in rates of evolution, there is a shift in the regions of the skeleton undergoing the most change; the character changes in the herbivorous lineages are concentrated in the mandible and dentition. The fact that the increase in rates of evolution coincides with increased change in characters relating to food acquisition provides stronger evidence for a causal relationship between the herbivorous diet and the radiation event.

Keywords: Adaptive radiation; Captorhinidae; Herbviore; Paleozoic; Tetrapod.

Figure 1. An illustration of the methods used to calculate disparity in this study.

(A) A hypothetical phylogeny illustrating as solid dots the data points that would be included under the method of Brusatte et al. (2011): the tip taxa A, B and C, and the Nodes 1 and 2; (B) The phylogeny plotted against a hypothetical trait, illustrating how the morphology of the lineage leading to Taxon C in time bin 1 and the morphology of the lineage leading to taxon A in time bin 2 may be inferred assuming a gradual model of evolution with no rate variation along a branch; (C) An illustration of how the same morphologies are inferred assuming a punctuated model of evolution, where the morphological change occurs at the speciation event.

Figure 2. The fit of models of diet evolution to the phylogeny of Captorhinidae.

Boxplots illustrating the distribution of 100 Akaike weights values calculated for each of the models of the evolution of diet as a discrete character, fit to the 100 time calibrated phylogenies of captorhinids. ER, Equal Rates; SYM, Symmetrical; ARD, All Rates Different.

Figure 3. The phylogeny of Captorhinidae, illustrating the evolution of diet.

Two of the 100 time calibrated phylogenies used in the analysis. The thick branches represent the observed ranges of each taxon. The colours of the tip labels represent the diet inferred for that taxon: Red, Carnivore; Blue, Omnivore; Green, Herbivore. The pie charts at each node represent the probability of each dietary regime inferred for that node, deduced by maximum likelihood ancestral state reconstruction. (A) MPT 1: Opisthodontosaurus is the sister to the clade containing Rhiodenticulatus and all captorhinids more derived. (B) MPT 2: Opisthodontosaurus is the sister to Concordia.

Figure 4. The phylogeny of Captorhinidae, illustrating the location of significant changes in rates of evolution.

Two of the 100 time calibrated phylogenies used in the analysis. The thick branches represent the observed ranges of each taxon. The colours of the tip labels represent the diet inferred for that taxon: Red, Carnivore; Blue, Omnivore; Green, Herbivore. The pie charts on each branch represent the proportion of the 100 time calibrated phylogenies which show significantly high or low rates of evolution along that branch: Red, significantly high rates; Blue, significantly low rates; White, no significant rate variation. (A) MPT 1: Opisthodontosaurus is the sister to the clade containing Rhiodenticulatus and all captorhinids more derived. (B) MPT 2: Opisthodontosaurus is the sister to Concordia.

Figure 5. A comparison of the mean rates of evolution within each dietary regime.

(A) Histogram illustrating the mean rate of discrete character evolution calculated for each dietary regime in each of the 10,000 stochastic maps of dietary evolution; (B) Histogram illustrating the mean rate of discrete character evolution calculated for the herbivorous lineages compared to a random selection of branches with an equal sample size in each of the 10,000 stochastic maps of dietary evolution.

Figure 6. A comparison of the morphological distances between taxa within each dietary regime.

Histogram illustrating the mean MORD distance between each taxon in each each dietary regime in each of the 10,000 stochastic maps of dietary evolution.

Figure 7. A comparison of disparity through time of the captorhinids in each dietary regime.

The disparity (sum of variances) calculated for all taxa within each dietary regime in each time bin. Values shown in the graph are the means of the values calculated in 10,000 stochastic maps of dietary evolution. The dashed line represents the mass extinction event dubbed Olson’s Extinction. (A) Morphology along each branch calculated assuming a gradualist model of evolution; (B) Morphology along each branch calculated assuming a punctuated model of evolution.

Figure 8. Lineage densities of captorhinids in each dietary regime.

Boxplots indicating the distribution lineage densities of the captorhinids in each dietary regime, calculated in each of the 10,000 stochastic maps of dietary evolution.

Figure 9. The proportion of characters within each skeletal region changing within each dietary regime.

Boxplots illustrating the distribution of the proportions of character changes in each skeletal region occurring in each dieatary regime, calculated in each of the 10,000 stochastic maps of dietary evolution. (A) Carnivores; (B) Omnivores; (C) Herbivores.

Figure 10. Phylomorphospace of captorhinids.

The phylogeny of captorhinids plotted over principal coordinates 1 and 2. Colours of lineages represent the die found to have the highest probability by the likelihood ancestral state reconstruction. Taxon labels—1, Thuringothyris; 2, Concordia; 3, Opisthodontosaurus; 4, Rhiodenticulatus; 5, Reiszorhinus; 6, Romeria prima; 7, Romeria texana; 8, Protocaptorhinus; 9, Saurorictus; 10, Captorhinus laticeps; 11, Captorhinus aguti; 12, Captorhinus magnus; 13, Captorhinikos chozensis; 14, Labidosaurus; 15, Captorhinikos valensis; 16, Labidosaurikos; 17, Moradisaurus; 18, Gansurhinus; 19, Rothianiscus; 20, MBCN 15730.