Browsing by Author "Miller, Tom"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    Intraspecific variation in metacommunities: Linking regional heterogeneity to local population dynamics
    (2014-12-03) Dibble, Christopher Joseph; Rudolf, Volker H. W.; Miller, Tom; Kohn, Michael; Lane, David
    Variation within species drives differences in population dynamics, interactions between species, and the functioning of complex ecosystems. Fittingly, understanding the factors that govern this intraspecific variation remains a central goal of ecology and evolutionary biology. Local processes such as adaptation tend to increase divergence among distinct populations, while regional processes such as dispersal and gene flow tend to homogenize those differences. My research addresses how heterogeneity is maintained despite the movement of individuals around a landscape. Specifically, I use an experimental host-parasite system to propose and test mechanisms contributing to ecological differentiation. Initially, I found that variation in colonizer traits makes the order in which they arrive to a new habitat important. Intraspecific priority effects (IPEs) occur when early arrivers limit the growth of late arrivers, and drive context-dependent differences in growth among populations (Chapter 1). These effects of variation in individual traits and arrival order extended to alter interspecific competition and host/parasite interactions. My research indicates, then, that the process of community assembly depends not only on the traits of a dominant species in the environment, but also which individuals of that species get there first (Chapter 2). The relatively short-term importance of IPEs, however, may wane over time as individuals continue to disperse among populations. I tested the consequences of repeated dispersal, and found that they depend heavily on the expected fitness of migrants in their new habitat. That is, dispersal from a common source can increase ecological heterogeneity among populations if migrants have different effects in different environments (Chapter 3). The context-dependent effects of dispersal suggested an underlying trait-based mechanism. Specifically, I hypothesized that the effects of increasing trait variance in a population (e.g. via immigration) depend on the relationship between a population’s trait mean and the local environmental optimum. I found that increasing trait variance helps populations with suboptimal trait means, but harms populations already well-suited to the local environment, doubling their disease burden (Chapter 4). Overall, my research identifies novel ways in which intraspecific variation contributes to its own maintenance, limiting the ability of individual movement among populations to homogenize ecological and evolutionary differentiation.
  • Thumbnail Image
    Item
    Population dynamics in a changing world: the consequences of environmental variation for species with complex life histories
    (2019-04-18) Schultz, Emily; Miller, Tom
    Despite increasing awareness of the importance of environmental variation in ecological systems, there are still many open questions about how variation affects population dynamics. I studied several aspects of environmental variation and the roles it plays in population dynamics, particularly for species with complex life histories. My first chapter addresses spatial variation in the effects of native insect herbivores on the population dynamics of an invasive thistle, Cirsium vulgare. It shows that while the underlying population dynamics of the thistle were predicted to vary substantially across space in the absence of herbivores, native insects were able to drive thistle populations to negative growth rates across the study region. My first chapter treats populations as isolated entities, with no dispersal between different populations. However, there is often dispersal between disjoint populations, creating metapopulations, isolated population patches connected by dispersal. Therefore, my second chapter incorporates multiple populations and dispersal to test for the importance of within-patch heterogeneity for metapopulation persistence, using a whitebark pine metapopulation as a model. I found that for whitebark pine, within-patch heterogeneity had minimal effects on metapopulation, and the metapopulation was predicted to be stable with or without the observed within-patch heterogeneity. My final chapter also used the whitebark pine system to test for shifts in elevation distributions due to climate change. Whitebark pine populations were predicted to be buffered against climate change because they do not fill their climatically-determined fundamental niche, and they can therefore tolerate warmer and drier climates than they experience in their current range. Thus, environmental variation can play an important role in population dynamics, but different sources of variation might be important at different scales of organization.
  • Thumbnail Image
    Item
    The role of environmental variation and host gene flow on the vertical transmission and population prevalence of heritable symbionts
    (2017-11-17) Sneck, Michelle E.; Miller, Tom; Egan, Scott P.; Kohn, Michael H.; Bartel, Bonnie
    Heritable microbial symbionts, vertically transmitted from maternal host to offspring, have made an indelible contribution to the ecology and evolution of life on earth. For instance, the fixation of symbionts in hosts contributed to pivotal biological shifts, such as the evolution of vascular plants and eukaryotic cells. Vertically transmitted symbionts are often specialized to host genotypes and confer fitness benefits to hosts, including protection against abiotic and biotic stress. Despite their ubiquity and strong influence on hosts, our understanding of what drives the prevalence and persistence of heritable symbionts lags behind that of macro-organisms. Two factors are theorized to determine equilibrium frequencies of heritable symbionts: 1) symbiont vertical transmission rates, and 2) the relative fitness of symbiotic and non-symbiotic hosts. Therefore, characterizing when and how these factors vary in host populations are necessary first steps to predicting the population dynamics of heritable symbionts. Here, I used large-scale field surveys, greenhouse and common garden experiments, as well as demographic modeling approaches to test the hypothesis that outcrossing (i.e., gene flow) between genetically distant hosts disrupts symbiosis. Specifically, host outcrossing is hypothesized to create genetic incompatibilities between sexually reproducing hosts and their specialized clonal symbionts, which may reduce both vertical transmission rates and symbiont mediated mutualistic benefits. First, I found that symbiont prevalence in one host species negatively associated with drought, while symbiont genotype explained residual variation in vertical transmission rates. These results suggest that symbiont genotype, and to a lesser extent, climate variables play roles in shaping symbiont population dynamics, but substantial variability was unexplained. Second, I manipulated gene flow between hosts along a gradient of genetic distances and determined that symbiont vertical transmission was robust to host outcrossing, which remained high for several host generations. Lastly, I quantified the net effect of host outcrossing on symbiont population dynamics. Contrary to our hypothesis, host outcrossing did not disrupt mutualistic benefits of symbiosis, and instead, buffered hosts against deleterious effects of outbreeding depression. Together, my work provides strong evidence that host outcrossing does not disrupt symbiosis, and alternatively demonstrates that heritable symbionts are important players in the population dynamics of outcrossing hosts.
  • Thumbnail Image
    Item
    Time Based Bayesian Optimal Interval (TITE-BOIN) Design Algorithm Performance under Weibull Distribution on Simulated Phase I Clinical Trial Data
    (2015-12-03) Rogers, Donald; Yuan, Ying; Vannucci, Marina; Cox, Dennis; Miller, Tom
    In phase I clinical trials, our goal is to effectively treat the patient while minimizing the chance of exposing them to excessively toxic doses of a new drug. In order to choose the correct dose, we use an adaptive dose-finding design, the Bayesian optimal interval design (BOIN), to aid in this selection. Here, we propose evaluation of the Bayesian optimal interval design under the Weibull and uniform distribution, comparing it to a time-based algorithm of the BOIN (TITE-BOIN). Simulations show that under theWeibull distribution, standard BOIN surpasses the TITE-BOIN design in terms of recommendation of maximum tolerated dose and allocation of data. In addition, both designs under theWeibull perform better than the uniform distribution when selecting a dose. Further study of the effects of the Weibull parameters on the BOIN design and the duration of trial under the Weibull should be considered.