Decoupled evolution of floral traits and climatic preferences in a clade of Neotropical Gesneriaceae

Background Major factors influencing the phenotypic diversity of a lineage can be recognized by characterizing the extent and mode of trait evolution between related species. Here, we compared the evolutionary dynamics of traits associated with floral morphology and climatic preferences in a clade composed of the genera Codonanthopsis, Codonanthe and Nematanthus (Gesneriaceae). To test the mode and specific components that lead to phenotypic diversity in this group, we performed a Bayesian phylogenetic analysis of combined nuclear and plastid DNA sequences and modeled the evolution of quantitative traits related to flower shape and size and to climatic preferences. We propose an alternative approach to display graphically the complex dynamics of trait evolution along a phylogenetic tree using a wide range of evolutionary scenarios. Results Our results demonstrated heterogeneous trait evolution. Floral shapes displaced into separate regimes selected by the different pollinator types (hummingbirds versus insects), while floral size underwent a clade-specific evolution. Rates of evolution were higher for the clade that is hummingbird pollinated and experienced flower resupination, compared with species pollinated by bees, suggesting a relevant role of plant-pollinator interactions in lowland rainforest. The evolution of temperature preferences is best explained by a model with distinct selective regimes between the Brazilian Atlantic Forest and the other biomes, whereas differentiation along the precipitation axis was characterized by higher rates, compared with temperature, and no regime or clade-specific patterns. Conclusions Our study shows different selective regimes and clade-specific patterns in the evolution of morphological and climatic components during the diversification of Neotropical species. Our new graphical visualization tool allows the representation of trait trajectories under parameter-rich models, thus contributing to a better understanding of complex evolutionary dynamics. Electronic supplementary material The online version of this article (doi:10.1186/s12862-015-0527-6) contains supplementary material, which is available to authorized users.

Appendix S1. Phylogenetic analysis and relative divergence time estimation The Codonanthopsis, Codonanthe and Nematanthus clade has ca. 52 species and is distributed throughout the Neotropical rainforest with a center of diversity in the Brazilian Atlantic forest (BAF). Recent phylogenetic and biogeographic analyses of the Neotropical Gesneriaceae have shown that the CCN group is divided in two sister clades . One is composed of Codonanthe sensu stricto and Nematanthus (39 species) that are endemic to the BAF (Chautems 2009), while the other includes Codonanthopsis (13 species) that occurs throughout the tropical rainforest of Central America, northern South America and Amazonia . We sampled 27 Nematanthus, 8 Codonanthe and 11 Codonanthopsis species, as well as 13 outgroup species representing different Episceae genera and other related Neotropical tribes such as Gesnerieae (Gesneria humilis), Gloxineae (Kohleria spicata) and Sinningieae (Sinningia schiffneri). This taxonomic sample includes all known species in the three genera CCN except C. anisophylla, C. chiricana, N. mirabilis, N. striatus, and the probable hybrid species N. kuhlmannii and N. mattosianus for which plant material was not available. All sampled species were sequenced for two nuclear (ITS, ncpGS) and four plastid regions (atpB-rbcL spacer, rpl16 intron, rps16 intron, trnL-trnF spacer) using the procedure described in Perret et al. (2013).
Voucher information and Genbank numbers for trnL-trnF and rps16 sequences are published in Perret et al. (2013). Genbank numbers for other markers are provided in Table S1. We used MAFFT (Katoh and Kuma 2002) and Guidance (Penn et al. 2010) to align sequences and subsequently remove all sites with a confidence score below 0.8; the final matrix contained 4,484 bp. We identified the GTR+G model of substitution as the best model for each DNA region, using AIC method as implemented in the phymltest function in R (ape package; Paradis et al. 2004). The only exception was the ncpGS gene, which was modeled with HKY85+G.
We used these models in Bayesian inference (BI: MrBayes 3.2; Ronquist et al. 2012) and all model parameters were unlinked across gene partitions. We performed two runs of BI for the complete alignment matrix, where each run consisted of four chains of 10 7 generations, sampling every 10 3 steps. We determined chain convergence and burn-in length (20% of the sampled generations) by examining trace plots of each parameter in Tracer v.1.4 (Rambaut and Drummond 2007). A consensus tree was calculated by removing the burn-in period and combining the two runs. Phylogenetic relationships and relative divergence times were estimated with a relaxed clock model applying uncorrelated log-normal prior distribution for the rates of substitution and a Yule prior on the rates of speciation using BEAST (Drummond et al. 2012). Each DNA region was treated as a separate data partition, allowing parameters of each region to be unlinked. Three independent analyses were performed by including either all 46 species sampled, the 38 species with morphological data or the 43 species with climatic data (see below). All three analyses included two independent runs of 2*10 7 generations each. The CCN clade was constrained to be monophyletic based on previous studies (Clark et al. 2012;Perret et al. 2013).
Due to the absence of closely related fossils in the CCN clade, difficulties to align sequences from fast evolving genes in higher taxa, and the multiple issues of secondary calibrations (Sauquet et al. 2012), an accurate absolute divergence-time estimation for this group is problematic. However, our analyses did not require absolute times of divergence because our goal is to compare the evolution of different traits within the CCN clade. We therefore used BEAST to produce ultrametric trees by setting the prior for the time of the most recent common ancestor of CCN clade to an almost fixed root of 1 (normal distribution with mean=1, sd=1e-06). We also carried out a MrBayes (v3.2; Ronquist et al. 2012) analysis for comparison (See Figure S6). Finally, we examined the evolution of three binary traits: geographic distribution, pollination syndromes, and floral orientation (see Table   S1). The biogeographic data was obtained from Perret et al. (2013), while information on pollination syndromes and resupination was obtained from available field observations of pollinators (see Table S2) and observations on living plants. Ancestral state reconstructions were performed using the Maximum Clade Credibility (MCC) tree and the function rayDISC (root=maddfitz) in the corHMM R package (Beaulieu et al. 2013).

Appendix S2. Visualization of continuous trait evolution
Trait simulations for each trait used the MCC tree of the BEAST analyses used the functions sim.rates (phytools R package, Revell 2012), for the multiple BM model, and OUwie.sim (Ouwie R package, Beaulieu et al. 2012) for the OU models. The sim.rates function allows for multiple BM rates of evolution, here estimated by rjMCMC (geiger R package, see above) and mapped on each branch of the phylogenetic tree. The OUwie.sim function requires OU parameters and tree painted with the defined selective regimes. All parameters were set according to the model selected for each trait (Table 1) and the root states were based on BM maximum likelihood estimation, even for OU models, because of the impossibility to estimate root state under these models. Table S1. List of species sampled, voucher information,and floral orientation (FO, 0= non-resupinated, 1= resupinated). Genbank accession numbers in italics correspond to sequences submitted with this manuscript, non-italics correspond to sequences reported in Perret et al. 2013. Binary trait information for geographic distribution (GD, 0= other biomes, 1= BAF), pollination syndrome (PS, 0= bee-pollinated and 1= hummingbirdpollinated).