UC Berkeley

The processes that generate and maintain biological diversity in the natural world have long captivated evolutionary biologists. In aposematic organisms, where conspicuous coloration warns predators of toxicity and/or unpalatability, phenotypic variation is particularly compelling because many factors are believed to contribute to color and pattern diversity within and among species. My dissertation focuses on a group of aposematic frogs endemic to Madagascar, commonly called Malagasy poison frogs (genus Mantella). The bright coloration observed in many Malagasy poison frogs coincides with the presence of toxic skin compounds derived from arthropod prey. Although several species within this group are endangered, little is known about their natural history and the diversity of color and pattern that occurs is poorly understood. Conservation efforts are further complicated by unresolved genetic relationships among phenotypically divergent species and populations. By characterizing patterns of genetic and phenotypic variation, my dissertation research aims to clarify the evolutionary processes contributing to color diversity and to inform management efforts for this group.

To better understand the factors contributing to phenotypic diversity, I first quantified genetic structure and color and pattern variation across populations of three closely-related species of Malagasy poison frogs, Mantella aurantiaca, Mantella crocea, and Mantella milotympanum. Although my genomic analyses identified three distinct genetic clusters within this complex, they did not correspond to current species designations. In some instances, populations presumed to be “morphs” of one species were actually identified as genetically distinct units. By demonstrating the complexity of distinguishing between species, intraspecific phenotypic variants, and populations, my study highlights the need to re-evaluate how species and morphs are classified, especially in aposematic organisms. In addition to the conservation implications of my study, I also found evidence that a variety of mechanisms, including selection, drift, and hybridization, are contributing to the color diversity observed in these populations.

In my second chapter, I further examined patterns of diversity among M. aurantiaca, M. crocea, and M. milotympanum populations by assessing the relationship between color and pattern variation and alkaloid-based chemical defenses. Unexpectedly, I detected very limited correlations between alkaloid composition and frog color, pattern, or conspicuousness. Additionally, differences in frog conspicuousness were not associated with differences in frog alkaloid diversity or abundance. Collectively, these results suggest that aposematic signals in this group may not be quantitatively honest, where increasing conspicuousness corresponds to increasing toxicity. Geographic distance, however, was highly correlated with alkaloid variation, implicating differences in habitat and associated characteristics as important determinants of variation in chemical defenses. Overall, my results indicate that geography, rather than phenotypic diversity, has a strong role in structuring chemical variation among populations.

In my final chapter, I aimed to clarify the status of the endangered Mantella cowani and its relationship to the common Mantella baroni by quantifying genetic and phenotypic variation across putative M. cowani – M. baroni hybrid and parental populations. Although previous mitochondrial studies have suggested hybridization between M. cowani and M. baroni in the wild, my genomic analyses did not recover admixture between the two species at this location. In fact, my population genetic analyses did not reveal any substantial genetic structure, even between M. cowani and M. baroni populations. Despite the lack of genomic differentiation, M. cowani was phenotypically distinct from both M. baroni and putative M. cowani – M. baroni hybrids. These discordant patterns of genomic and phenotypic diversity could possibly be explained by longstanding hybridization, ongoing migration, or selection acting on a small portion of the genome. Despite the lack of genetic variation among populations, I found M. cowani to be a morphologically distinct population. Further, the putative M. cowani – M. baroni hybrid population demonstrated novel phenotypes not observed in either species. My findings raise important questions as to how different aspects of biological diversity should be prioritized in conservation, especially in aposematic species where phenotypes have clear ecological relevance and may be important for survival.

My dissertation provides integrative studies of phenotypic and genetic diversity for several species of Malagasy poison frogs and has important implications for both evolutionary biology and conservation. Greater understanding of adaptation and diversity in this group will be essential to future conservation efforts and to the preservation of Madagascar’s unique biodiversity.