UC Berkeley

What is the effect of isolation in the formation of new lineages? Volcanic islands provide an ideal scenario to explore this question due to their various degrees of isolation, environmental variability and generally understood geology. On the other hand, the dispersal abilities, diverse ecology, and known natural history of spiders make them a suitable group of study to address this big question.

Because, the formation of new clades on islands start with a colonization event, I first studied long distance dispersal across different Pacific archipelagoes for spiders of the genus Tetragnatha. Then, I focused on the early stages of diversification for this group in the Hawaiian archipelago using population genetics studies combining Sanger sequencing of mitochondrial genes and Exon Capture for nuclear markers. Finally, I studied the effect of isolation at the morphological level. In particular, by looking the convergent evolution of color polymorphisms among species present on islands and continental areas. For this I examine the case of the family Theridiidae.

Colonization of remote archipelagoes corresponds to a very rare event. The Long- jawed spiders, genus Tetragnatha, are widely distributed around the world including oceanic islands. In Hawai'i there are 37 species of one or two colonization events from North America. In the Marquesas there are 5 species and phylogenetic evidence suggests an origin somewhere in the Asia. For the species present in the Society Islands, Rapa Nui and D. raptor (Hawai'i), their biogeographic origin is still unsolved. In order to examine colonization patterns across the Pacific Rim, I used freshly collected and museum specimens for phylogenetic analysis. The DNA from historic collections was sequenced using newly design primer pairs to amplify overlapping short fragments of COI. The phylogenetic reconstruction shows two of the species from the Society Islands (T. rava and T. tuamoaa) and D. raptor in the same clade as other species from Asia. T. moua (Society Islands) is sister to T. nitens, which has a circumtropical distribution making strong biogeographic statements difficult. The species from the Marquesas appear linked to a basal polytomy, while the one from Rapa Nui is in the same clade as the Hawaiian species. Other new species from Hawai'i were added into the analysis, but their phylogenetic position within the Web building clade is not conclusive. The first chapter provides preliminary information about origin of Tetragnatha spiders in Pacific islands. The implementation of Next Generation Sequencing technologies and inclusion of more species could improve this understanding.

Once the founder event occurs lineages could follow different diversification trajectories. Adaptive radiation is one of them in which many species evolve on a short period of time occupying different ecological niches. The study of their temporal dynamic it is not a simple problem to address in natural conditions. However, the chronosequence of the Hawaiian Islands provides "snapshots" of different stages of the diversification process. Here, I propose a population-based mechanism, which associated with the changes on the landscape attempts to explain the patterns of species diversity. By comparing species present on the young (Big Island: T. anuenue and T. brevignatha) and middle age islands (Maui: T. waikamoi and T. brevignatha), I tested the following predictions: (1) higher genetic structure in the young island populations and (2) lower genetic structure in populations from the middle age island. To test these hypotheses I sequenced three mitochondrial genes (COI, ND1 and Cytb). For my first prediction I found that the mitochondrial haplotype distribution is not related to the specific volcano configuration, but rather with breaks in forest types associated with rain regimens. The second prediction was not supported. Instead, I found strong genetic isolation between populations on Maui. Finally, a time calibrated coalescent reconstruction of T. brevignatha provided evidence for a population expansion during Pleistocene glacial periods. It also suggests that the Windward/Leeward population break is related to more recent events than the original colonization of the island. Species evolving in Hawai'i seem to be affected not only by the geologic development of the islands, but also by more recent climatic events.

On this context, I studied the genetic structure of newly formed species. In the third chapter I examined the genetic signatures of young speciation events by applying the transcriptome-based Exon Capture approach to the study of three closely related species (T. brevignatha, T. waikamoi and T. macracantha) from Maui Nui and Big Island. The combined data showed the existence of 5 clades: T. brevignatha Big Island, T. brevignatha Maui, T. macracantha, T. waikamoi and a WaikBrevMac clade. The last one includes specimens from all three nominal species. For T. brevignatha, the molecular divergence between the populations of Maui and Big Island are similar to "species level" comparisons. The Big Island population shows a strong break between the Leeward and Windward populations. For T. waikamoi nuclear data suggests that the population break between Haleakalā and West Maui is less pronounced than the break based on mitochondrial data. In the case of T. macracantha, we found a separation between the Maui and Lana'i populations. Moreover, there are two very divergent lineages co- existing in Kīpahulu Valley. Lastly, the finding of the WaikBrevMac group adds a new layer of complexity where recent speciation events suggest divergent lineages with convergent (or are the ancestral) morphologies and no genetic admixture. The addition of species with different eco-morphologies will complement these findings and also inform about the genetic signatures of ecological speciation.

Finally, to examine the effects of isolation in morphology I addressed the phenomena of convergent evolution in a color polymorphism on island and continental species. This kind of convergence has been widely studied, as it corresponds to a unique situation where the convergence leads to diversity. In the spider family Theridiidae, the independent evolution of the abdominal color polymorphism has been well described in at least four species (Enoplognatha ovata, E. latimana, Theridion grallator and T. californicum). Among the shared characteristics on these cases of color polymorphism are: (1) the presence of "Yellow" as the double recessive and more common variant, (2) single loci Mendelian inheritance (except in Big Island population of T. grallator), (3) constant frequencies of variants among poorly connected populations and (4) evidence of the action of natural selection. The genus Selkirlkiella from the temperate rainforest of southern South America has several species with some degree of color polymorphism. Here, I documented the color polymorphisms of S. alboguttata (Robinson Crusoe Island) and S. luisi (Valdivia, Chile). Selkirlkiella alboguttata displays 6 morphs and S. luisi two morphs. The "Yellow" morph was the more common in both species. Based on a molecular phylogeny, we confirm that this genus is closely related to Enoplognatha. The presence of color polymorphism in this genus appears to be an event of convergent evolution at the family level, while between species it is likely due to common ancestry. There is also evidence of other cases of color polymorphism in the family. Finally, this phenomenon seems to be associated with the "under leaf community".

Spiders are a very diverse group of organisms. Due to their exceptional dispersal abilities, they have colonized even the most remote islands on the planet. The implementation of new genetic methods as well as the collection of specimens from remote areas will reveal even more insights about the biology of these organisms. This knowledge will extremely valuable to expand the understanding of the effects of isolation in the generation of biodiversity.