Search

Supplementary materials, methods and results

Supplementary Figures, Tables and Methods Descriptions

The importance of the environment in shaping phenotypic evolution lies at the core of evolutionary biology. Chipmunks of the genus Tamias (subgenus Neotamias) are part of a very recent radiation, occupying a wide range of environments with marked niche partitioning among species. One open question is if and how those differences in environments affected phenotypic evolution in this lineage. Herein we examine the relative importance of genetic drift versus natural selection in the origin of cranial diversity exhibited by clade members. We also explore the degree to which variation in potential selective agents (environmental variables) are correlated with the patterns of morphological variation presented. We found that genetic drift cannot explain morphological diversification in the group, thus supporting the potential role of natural selection as the predominant evolutionary force during Neotamias cranial diversification, although the strength of selection varied greatly among species. This morphological diversification, in turn, was correlated with environmental conditions, suggesting a possible causal relationship. These results underscore that extant Neotamias represent a radiation in which aspects of the environment might have acted as the selective force driving species' divergence.

3.2. Diversification times in the genus Pteronotus The rising of Mormoopidae is estimated in the late Oligocene, ca. 32 Ma (95% HPD = 31–36.7 Myr). Divergence time of the Most Recent Common Ancestor (MRCA) of genus Pteronotus dates to the middle Miocene, around 16.1 Ma, with a confidence interval (95% HPD) ranging from 12.3 to 20.2 Myr. Diversification events giving rise to the main clades of Pteronotus are evenly distributed along the genus phylogeny. The first was the basal split between the ancestor of Clade 4 (subgenus Phyllodia) and the branch giving rise to the other Pteronotus clades. Splitting between Clade 3 and the ancestor of Clades 1 and 2 also took place early in the genus evolutionary history (~ 13.9 Ma), followed by the separation of Clades 1 and 2 at 11.3 Ma. Emergence of closely related species started at the late Miocene, around 7.5 Ma, with differentiation between P. macleayi and P. quadridens within Clade 2, and extended through all Pliocene less than 3 Ma with cladogenesis within Clades 4, 3 and 1. Origin and diversification of most extant Pteronotus lineages is recent, occurring during the Pleistocene (Fig. 1 and Table 2). Table 1 Evolutionary rate statistics for each loci investigated in the present study. Values were estimated under the uncorrelated lognormal relaxed molecular clock. HPD = High Posterior Density. Table 2 Estimates for the Most Recent Common Ancestor (MRCA) of Pteronotus lineages sampled for two or more specimens in the multilocus dating analysis. 3.3. Ancestral geographic ranges Results of the six biogeographic models implemented in BioGeoBEARS are presented in Table 3, which describes Log-likelihood values found for each of them. Comparative statistics show that DIVAj (LnL = ¯ 42.27) and DIVA (LnL = ¯ 44.07) models best fitted the data, indicating assumptions on geographic range evolution made by dispersal-vicariant method mostly satisfy patterns of geographic distribution of Pteronotus lineages. Implementation of DIVA model includes two free anagenetic parameters, d (dispersion rate/range expansion) and e (extinction rate/range retraction), and fixed cladogenetic parameters allowing equal probability of vicariance for all descendant size ranges, but disallowing subset sympatric speciation (Matzke, 2013). The assignment of the free parameter J in the model, specifying weights for jump-dispersal events in the cladogenesis matrix, produced significant improvement of the biogeographic model (AICwt = 0.69 for DIVAj versus 0.31 for DIVA; Table 3). As previously shown for other islands groups (Matzke, 2014), inclusion of the J parameter in the model was important to explain some biogeographic events in the genus Pteronotus, such as colonization of Caribbean islands. Table 3 Log-likelihood values of the six biogeographic models tested in the present study. Models performance was compared through Akaike Information Criterion (AIC) and Akaike weight (AICwt), the last being estimated just between the two better AIC values. The most probable states for the ancestral node ranges for the genus Pteronotus were estimated using the DIVAj model (Fig. 2). Geographic range estimates of early nodes in the phylogeny had relatively low probabilities in all biogeographic models (data not shown), demonstrating higher uncertainty of biogeographic inferences. Area combinations displayed with black dashed lines in phylogeny nodes had frequencies smaller than 10% in the estimates, but that still are the most probable states amongst the 638 possible area combinations. The most probable alternative states of ancestral ranges according to DIVAj model are listed in Table S3. According to the DIVAj model, the ancestor of Pteronotus inhabited a wide geographic area, ranging from Central America and Jamaica to the Amazon Region. An initial vicariant event split this population in the ancestral lineage of Clade 4 in South America and the ancestral lineage of Clades 1 + 2 + 3 in Central America (both ancestors could not have raised in CA because dispersal-vicariant model does not allow subset sympatric speciation). The ancestral population of Clade 4 later expanded its range into Central America, while ancestor of Clades 1, 2 and 3 diversified and dispersed to South America. In the subsequent cladogenesis event, between Clades 1 and 2, the first dispersal jump to Greater Antilles in the genus evolutionary history took place. The dotted-lined arrows in Fig. 2 highlight the shift in ancestral geographic ranges from Caribbean Coast and Central America to Jamaica and Cuba at Clade 2 ancestor. Diversification events within Clades 3 and 4 also required dispersal jumps, as displayed by the same arrows in the figure. Within Clade 4, both Amazon occupation by the ancestor of Clade 4C and the second colonization event in the Greater Antilles by the ancestor of Clade 4A were presumed to involve shifts in ancestral distributions. A similar scenario is estimated by the DIVA model (Fig. S 2), except that occupation of new areas occurred via anagenetic dispersal of ancestral populations instead of through jumps at cladogenesis events. 4. Discussion 4.1. Origin and diversification patterns in the genus Pteronotus Our time-scale estimates suggest that early cladogenesis events within the genus Pteronotus occurred gradually, encompassing a large time range during the evolutionary history of this group. The age of the Most Recent Common Ancestor (MRCA) of genus Pteronotus was estimated to be between 12.3 and 20.2 Ma, which is old when compared to divergence of other bat genera (Lim, 2008; Rojas et al., 2016; Velazco and Patterson, 2013). Despite this, the genus Pteronotus harbors a relatively small number of species, which are morphologically homogeneous in their skull and external characters (Simmons and Conway, 2001). The lack of phenotypic variation confounded early species designations based solely on morphology, while this genus is characterized by deep genetic divergences among lineages (Pavan and Marroig, 2016). Interestingly, the MRCA of the genus Pteronotus lived around 16 Ma, which represents just half of age of the MRCA of Family Mormoopidae, estimated to be at least 32 Myr old. This large interval of ca. 16 Myr between family and genus origins may imply the ancestor of Pteronotus underwent a long period of anagenetic evolution or, alternatively, that basal lineages filling this temporal gap became extinct later. Biogeographic reconstruction of DIVAj model provides a scenario with both vicariant events and dispersal jumps to explain distributional patterns of extant Pteronotus lineages. This model proposes a widely distributed ancestor for the genus Pteronotus, which does not favor any of the current biogeographic hypotheses on the group origin (Morgan and Czaplewski, 2012; Smith, 1972). It is important to highlight, however, that ancestral range estimates by DIVA model seeks to minimize dispersal-extinction costs (Ronquist, 1997). Thus, it tends to reconstruct wide ancestral ranges at early nodes (which are directly related to the maximum number of areas settled) to explain distributional patterns of phylogeny tips (Lamm and Redelings, 2009; Ree et al., 2005). In spite of this, it seems that founder event speciation was an important mode of lineage splitting in the genus Pteronotus, particularly within Clade 4., Published as part of Pavan, Ana C & Marroig, Gabriel, 2017, Timing and patterns of diversification in the Neotropical bat genus Pteronotus (Mormoopidae)., pp. 61-69 in Molecular phylogenetics and evolution 108 on pages 63-66, DOI: 10.1016/j.ympev.2017.01.017, http://zenodo.org/record/7845815, {"references":["Matzke, N., 2013. BioGeoBEARS: Biogeography with Bayesian (and Likelihood) Evolutionary Analysis in R Scripts. University of California, Berkeley, Berkeley, CA.","Matzke, N. J., 2014. Model selection in historical biogeography reveals that founder-event speciation is a crucial process in island clades. Syst. Biol.","Lim, B. K., 2008. Historical biogeography of New World emballonurid bats (tribe Diclidurini): taxon pulse diversification. J. Biogeogr. 35, 1385 - 1401.","Rojas, D., Warsi, O. M., Davalos, L. M., 2016. Bats (Chiroptera: Noctilionoidea) challenge a recent origin of extant neotropical diversity. Syst. Biol.","Velazco, P. M., Patterson, B. D., 2013. Diversification of the Yellow-shouldered bats, Genus Sturnira (Chiroptera, Phyllostomidae), in the New World tropics. Mol. Phylogenet. Evol. 68, 683 - 698.","Simmons, N. B., Conway, T. M., 2001. Phylogenetic Relationships of Mormoopid Bats (Chiroptera: Mormoopidae) Based on Morphological Data. Bulletin of the American Museum of Natural History 258, 97 pp.","Pavan, A. C., Marroig, G., 2016. Integrating multiple evidences in taxonomy: species diversity and phylogeny of mustached bats (Mormoopidae: Pteronotus). Mol. Phylogenet. Evol. 103, 184 - 198.","Morgan, G. S., Czaplewski, N. J., 2012. Evolutionary history of the Neotropical Chiroptera: the fossil record. In: Gunnell, G. F., Simmons, N. B. (Eds.), Evolutionary History of Bats: Fossils, Molecules and Morphology. Cambridge University Press, New York, USA, pp. 105 - 161.","Smith, J. D., 1972. Systematics of the chiropteran family Mormoopidae. Miscellaneous Publication, Museum of Natural History, University of Kansas, vol. 56, pp. 1 - 132.","Ronquist, F., 1997. Dispersal-vicariance analysis: a new approach to the quantification of historical biogeography. Syst. Biol. 46, 195 - 203.","Lamm, K. S., Redelings, B. D., 2009. Reconstructing ancestral ranges in historical biogeography: properties and prospects. J. Systemat. Evol. 47, 369 - 382.","Ree, R. H., Moore, B. R., Webb, C. O., Donoghue, M. J., 2005. A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution 59, 2299 - 2311."]}

Figures S1 to S3 and tables S1 to S7

Most biological systems are formed by component parts that to some degree are inter-related. Groups of parts that are more associated among themselves and are relatively autonomous from others are called modules. One of the consequences of modularity is that biological systems usually present an unequal distribution of the genetic variation among variables. Estimating the covariance matrix that describes these systems is a difficult problem due to a number of factors such as poor sample sizes and measurement errors. We show that this problem will be exacerbated whenever matrix inversion is required, as in directional selection reconstruction analysis. We explore the consequences of varying degrees of modularity and signal-to-noise ratio on selection reconstruction. We then present and test the efficiency of available methods for controlling noise in matrix estimates. In our simulations, controlling matrices for noise vastly improves the reconstruction of selection gradients. We also perform an analysis of selection gradients reconstruction over a New World Monkeys skull database in order to illustrate the impact of noise on such analyses. Noise- controlled estimates render far more plausible interpretations that are in full agreement with previous results.