The function of coloration and inter-island diversification of pachyrhynchid weevils

• Main Authors: TSENG, HUI-YUN, 曾惠芸
• Other Authors: LIN, CHUNG-PING
• Format: Others
• Language: en_US
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/67873775773914519197

博士 === 東海大學 === 生命科學系 === 105 === Pachyrhynchid weevil (Coleoptera: Curculionidae) is a group of insects with astonishing and diverse coloration, but their biology and diversification are poorly known. This dissertation studied the function of these diverse colors and how these weevils and their colors have evolved on the oceanic islands where they inhabited. Chapter 1 aims to test whether these conspicuous colorations function as aposematic signals for the arboreal predator, Swinhoe’s tree lizard (Japalura swinhonis). Pachyrhynchus tobafolius and Eupyrgops waltonianus were selected as potential prey, and the predatory responses of allopatric versus sympatric lizard populations were compared by presenting weevils with intact or artificially-removed markings. Sympatric lizards were found to attack masked weevils more frequently than intact ones, whereas allopatric lizards attacked intact weevils with a significantly higher rate due to the novelty of the preys. The results supported the hypothesis that the conspicuous markings of pachyrhynchid weeils function as aposematic signals for lizards. Chapter 2 tested the stepping-stone hypothesis of colonization of the Pachyrhynchus orbifer complex in the Taiwan-Luzon volcanic belt using phylogenies reconstructed from a multi-locus data-set. Divergence time estimates of all extant island lineages (<1 Mys of Middle Pleistocene) are much more recent than the geological ages of the islands (22.4-1.7 Mys). Complex and stochastic colonization of P. orbifer was inferred to have involved both northward and southward dispersals, which are strongly inconsistent with the strict stepping-stone hypothesis. Speciation of the P. orbifer complex may have occurred only between islands, suggesting that peripatric speciation through founders of stochastic dispersals was the major evolutionary driver. Chapter 3 evaluated the effect of filtering parameters in the process of assembling RAD-sequences. The remarkable advances in next-generation sequencing technology provide a powerful tool for phylogenetics and population genetics, especially in recently divergent populations and shallow phylogenetic relationships, which were problematic using a limited number of loci. However, potential errors from different sources of RAD-seq may cause incorrect inference in downstream analyses. RAD sequences of P. orbifer species complex with different filtering parameters combinations in Populations of Stacks program were used to evaluate the stability of inferred phylogenies by comparing the number of retained loci, number of SNPs, heterozygosity, percentage of polymorphic sites, bootstrap supports and tree topologies. These comparisons showed that even processing the loci in the population level caused different levels of variation in the phylogenetic inferences, and setting higher values of r, p and m might not be always suitable. We recommend choosing filtering parameters by testing different combination of parameters.