Living organisms are under selection not only for one, but also for several inheritable characters at the same time. Well-sampled and well-supported phylogenies are necessary for the studies of character evolution and their history. The poison frogs (Dendrobatidae) are a well-known example of aposematism in anurans. They include ~270 species of Neotropical frogs with aposematic (toxic and conspicuous) and non-defended (palatable and cryptic) species. The origin of aposematism in poison frogs is puzzling, because of its predicted low probability of establishment due to the prey's increased conspicuousness. Previous studies suggested a single origin of toxicity and warning coloration. By expanding taxon sampling of the group, I reexamined the phylogenetic correlation between the origins of toxicity and warning coloration. I found four or five independent origins of aposematism; by using simulations, I rejected hypotheses of one, two, or three origins of aposematism (P < 0.002). I also found that diet specialization is linked with the evolution of aposematism and has evolved independently at least two times. Poison frogs are endemic to the Neotropic, which is one of the Earth's largest reservoir of biodiversity. I reconstructed the biogeography of the poison frog clade and rejected an Amazonian center-of-origin in favor of a model expanding over the Neotropics. I inferred 14 dispersals into and 18 out of Amazonia to adjacent regions; the Andes were the major source of dispersals into Amazonia. Significant percentage of dendrobatid diversity in Amazonia and Chocó resulted from repeated immigrations, with radiations at <10.0 million years ago. In contrast, the Andes, Venezuelan Highlands, and Guiana Shield have undergone extended in situ diversification at near constant rate since the Oligocene. Poison frogs have significant variation on their physiological characteristics. I measured resting and active metabolic rates of 54 species. I traced metabolic measurements along aposematism, diet specialization, molecular rates, and body mass. I found a synergistic and co-adapted functionality of active metabolic rates with all previous traits that is perhaps the consequence of the increase in complexity in most biological systems. My thesis has expanded the knowledge of the biology, phylogenetic history, and biogeography of the poison frogs. / text

Aposematic signals advertise chemical defense or other forms of unprofitability to predators and may be parasitized by dishonest signallers through Batesian mimicry. Warning signals do not provide perfect avoidance, however, meaning many aposematic phenotypes evolve to balance between signal saliency and mitigating detection. For Batesian mimics the cost of a predator encounter should be greater, which begs the question of whether imperfect mimicry leads Batesian mimics to exhibit more muted signals than their models. I sought to test this hypothesis and other related hypotheses using the Ameerega-Allobates poison frog mimicry complex native to the Ecuadorian Amazonian rainforest. In Chapter 2 I found that, while many elements of the mimic’s signal were less salient than in the model, the mimic had higher detectability than the model. I also found that saliency discrepancies across colour patches produced variation in detectability across different body postures and viewing angles in both species. In Chapter 3 I turned my focus to how the balance between cryptic and salient signal elements changes across ontogeny, and how shifts in that balance may affect detectability. I found that both species undergo ontogenetic colour change, with the mimic improving in resemblance to the model as it developed and the model refining its aposematic signal as it grew. These colour changes impacted the crypsis efficacy of the mimic in that different colour stages had differential detectability. In Chapter 4 I tested for behavioural associations with the ontogenetic increase in mimetic fidelity in Al. zaparo. I found that colouration alone did not explain variation in behaviour, and that body size and environmental conditions impacted boldness and activity. / Dissertation / Doctor of Philosophy (PhD) / Antipredator colouration can take many different forms and can function according to different underlying strategies. Crypsis, also known as camouflage, operates by evading detection – effective crypsis means that predators don’t see the prey in the first place, increasing the prey’s survival. Using crypsis as an antipredator strategy means that a species should only occupy space where its colours blend in, however, and when hiding from predators it’s difficult to attend to other ecological needs like foraging, conquering and/or defending territory, or pursuing reproduction. These opportunity costs are thought to be reduced under aposematism, or ‘warning colouration’, where an animal evolves bright colouration that advertises some sort of secondary defense like toxic stings or secretions. Aposematic protection stems from standing out in the environment and being distinguishable from other prey, providing reduced predator attention across contexts such as foraging, reproduction, thermoregulation, or simple movement throughout the environment. These benefits can then be parasitized by a dishonest ‘mimic’ species which has no chemical defense but copies the warning signal of a ‘model’. Warning signal protection isn’t perfect, however, and most aposematic species still experience some form of predation threat. This means that there may be a limit to how conspicuous a signal an aposematic species is favoured to evolve. Accordingly, evidence shows that aposematic species incorporate both cryptic and aposematic elements into their colouration. The aim of this thesis is to examine how this balance is achieved by a poison frog and a non-toxic mimic, how this balance may change across development, and whether there are behavioural implications of changes in colour strategy. I found that both species have evolved a salient colour patch that they can flexibly expose or hide to modulate their detectability. I also found that the balance between crypsis and aposematism in the mimic and model shifts over time through developmental colour change. Lastly, I found that body size and environmental conditions affect behaviour in developing individuals of the mimic species.