Diversity dynamics in Nymphalidae butterflies: Effect of phylogenetic uncertainty on diversification rate shift estimates

TL;DR

This study demonstrates that phylogenetic uncertainty significantly influences diversification rate estimates in Nymphalidae butterflies, highlighting the importance of accounting for such uncertainty to avoid misleading conclusions.

Contribution

We extended the MEDUSA method to incorporate phylogenetic uncertainty, revealing its impact on diversification rate shift estimates in butterfly lineages.

Findings

Phylogenetic uncertainty causes large variation in diversification rate estimates.

Only 3 of 13 significant shifts are consistent across most trees.

Hostplant shifts are linked to diversification rate changes.

Abstract

The family Nymphalidae is the largest family within the true butterflies and has been used to develop hypotheses explaining evolutionary interactions between plants and insects. Theories of insect and hostplant dynamics predict accelerated diversification in some scenarios. We investigated whether phylogenetic uncertainty affects a commonly used method (MEDUSA, modelling evolutionary diversity using stepwise AIC) for estimating shifts in diversification rates in lineages of the family Nymphalidae, by extending the method to run across a random sample of phylogenetic trees from the posterior distribution of a Bayesian run. We found that phylogenetic uncertainty greatly affects diversification rate estimates. Different trees from the posterior distribution can give diversification rates ranging from high values to almost zero for the same clade, and for some clades both significant rate increase and decrease were estimated. Only three out of 13 significant shifts found on the maximum credibility tree were consistent across more than 95% of the trees from the posterior: (i) accelerated diversification for Solanaceae feeders in the tribe Ithomiini; (ii) accelerated diversification in the genus Charaxes, and (iii) deceleration in the Danaina. By using the binary speciation and extinction model (BISSE), we found that a hostplant shift to Solanaceae or a codistributed character is responsible for the increase in diversification rate in Ithomiini, and the result is congruent with the diffuse cospeciation hypothesis. A shift to Apocynaceae is not responsible for the slowdown of diversification in Danaina. Our results show that taking phylogenetic uncertainty into account when estimating diversification rate shifts is of great importance, and relying on the maximum credibility tree alone potentially can give erroneous results.