Old World Serpents and New World Dragons: The Evolutionary Dynamics of Pythons and Liolaemid Lizards

Abstract

Since Darwin and Wallace, evolutionary biologists have been fascinated with the processes driving the evolution of biodiversity, species richness, ecology and morphology. Combining phylogenetic, geological, paleontological, ecological and climatic perspectives can shed light on evolutionary patterns at deep time scales to answer questions about the evolution of biodiversity. At the same time, looking at more shallow scales, where species are initially forming and diverging, can shed light on phylogeographic aspects of evolution. In this thesis I use two animal systems with comparable ages to understand evolutionary processes at both deep and shallow time-scales. Pythons (Pythonidae) are a diverse group of snakes distributed across the tropics and subtropics of the Old World. Their moderate species diversity (44 currently recognized species), is eclipsed by their remarkable morphological and ecological diversity. Liolaemid lizards (Liolaemidae), found in the southern portion of South America, also are diverse in their habitat use, reproductive mode and thermal niche, but are extremely species rich, with almost 300 currently recognized species. Both pythons and liolaemid lizards, based on different and common attributes, have been recognized as examples of an adaptive radiation. My research on pythons is summarized in the first three chapters. In Chapter 1, I use a phylogenomic dataset of over 300 loci and mitochondrial genomes, along with fossil calibrations, to reconstruct a time-calibrated species tree. Biogeographic reconstruction analyses reveal pythons likely originated in Asia, and only recently crossed Wallace’s line into Australo-Papua. Lineage and morphological diversification analyses show that once they crossed this line, pythons display a sudden burst of species, ecological and morphological diversity, consistent with ecological opportunity following colonization of a new environment with novel niches. In Chapter 2, I reconstruct ontogenetic allometric trajectories to illustrate changes in python head and body shape from juveniles to adults. I find that heterochrony (a dissociation of size, shape and age) is the initial developmental process responsible for pythons to diverge in morphology. Although this process involves shifting the rate and timing of development, it still requires that the direction of change along morphological space remain the same. However, I also find that at deeper evolutionary levels, the direction of shape change with size becomes evolvable, allowing pythons to explore more extreme phenotypes than previously available. Chapter 3 looks at phylogeographic patterns of one of the most iconic snakes in the world: the green tree pythons (Morelia viridis complex). Using a similar dataset as described above, but analysed in a species-delimitation and population structure framework, I identified deep phylogeographic structure within the group, and recognize two species, one of which comprises three subspecies. These taxa are highly cryptic, and the central cordillera of New Guinea likely played a key role in maintaining isolation between the lineages, allowing them to diverge. Chapter 4 switches the focus to the liolaemid lizards. I first sequenced mitochondrial loci for 40 species for which genetic data did not exist, and combined this with data from previous studies to build a time-calibrated tree with 258 terminal taxa, the most complete phylogenetic inference of the group to date. I tested multiple hypotheses on the diversification and evolutionary dynamics of liolaemids, all of which centered on the importance of the uplift of the Andes mountains in generating biodiversity. I find that most lineages originated in Andean regions before dispersing into surrounding biomes. Combined with findings that speciation rates are correlated with rising altitudes, I postulate the Andes acted as a species pump for the most diverse lizard group of South America. Moreover, using available data on distribution, thermal niche and reproductive mode (oviparity versus viviparity), I find that viviparity was the key factor in enabling liolamids to adapt to the high mountains of the Andes. Importantly, when lineages dispersed back into the lowlands, my results for Liolaemus, the most diverse genus in the family, suggest that several reversals back to oviparity must have occurred. This is one of the very few known cases where this reversal in reproductive mode has taken place. In Chapter 5 I focused on the Liolaemus (sensu stricto) subgenus, which is mostly distributed around the Andes between Chile and Argentina. Very rapid diversification events have rendered the phylogenetic relationships within this group extremely hard to resolve. I obtained tissues from almost 400 samples from the 87 described and 13 candidate species. I sequenced mitochondrial data using Sanger sequencing, and nuclear Single Nucleotide Polymorphisms (SNPs) using ddRADSeq, and used cutting edge phylogenetic reconstruction methods to resolve the evolutionary history of the group. I find very high levels of discordance between mitochondrial and nuclear data, suggesting extensive past introgression between lineages, highlighting single genetic markers and concatenation methods are of little use for resolving systematic relationships within the group. This is confirmed by the extensive reticulation events inferred by phylogenetic networks. Nevertheless, these results provide a huge step towards understanding the evolutionary patterns of Liolaemus in the context of the geological history of the Andes, and also help to improve a historically messy taxonomy, with the synonymy of many species and the likely species level status of many undescribed taxa. Finally, in Chapter 6, I explore the shallower evolutionary patterns of one of the most interesting groups of Liolaemus, the leopard lizards (Liolaemus leopardinus). Most species in this clade are found in the Andes mountains, but one is found in the isolated Costa cordillera. The mitochondrial genealogy suggests the taxon in the Costa mountains is sister to the rest, but the phylogenies estimated from thousands of independent SNPs instead suggest the pattern of all Andean species forming a clade is an artifact of past introgression. My molecular dating results suggest this group diverged during the Pleistocene, a period in time characterized by intermittent glacial and interglacial cycles. I suggest that past introgression occurred during glacial maxima, when taxa were forced down the mountains, enabling them to interbreed in the lowlands, followed by isolation in mountain tops during interglacial periods. I revise their taxonomy, sinking one species and describing another. My thesis highlights the importance of looking at different evolutionary timescales, geographic regions, and geohistorical processes to understand how biodiversity, is generated at all levels. Show more