Research Topics
Population genetic
|
Members of my lab often pursue research projects that establish expectations for levels of population structure, polymorphism, or divergence. These projects are focused on determining our ability to detect or distinguish processes using the observed patterns in empirical genetic data. Such studies shed light on whether certain hypotheses can be tested effectively with a given type or amount of genetic data given sampling variance and the variance among replicate outcomes of random processes. Examples of such work is population structure (FST) for nuclear genetic markers and organelle genetic markers for different rates of pollen and seed gene flow (Hamilton and Miller 2002), comparing population differentiation for neutral molecular loci estimated by FST and for the additive genetic component of quantitative traits estimated by QST (Miller et al. 2008), and estimators of genetic effective population size based on average two-locus gametic disequilibrium in unphased genotype data (Hamilton et al. 2018).
|
Ecological genetics |
While ecological genetics has a long history, it is only relatively recently that biologists have started to explore how the full range of population genetic processes (drift, selection, gene flow) interacts with and also causes ecological variation. This new change of emphasis is called “eco-evolutionary dynamics,” especially when there is a feedback between ecological and population genetic variation.
Current work is focused on the foundation salt marsh plant Spartina patens. First, the lab is using computer simulations to validate measures to estimate the frequency of clonal reproduction in populations that are partly clonal and partly sexual. The frequency of clonal reproduction is a fundamental characteristic of reproduction (similar to rates of selfing and outcrossing) that has often been overlooked in studies of natural populations. Using a set of novel microsatellite genetic markers, the lab is also working to identify the frequencies of genets (unique multilocus genotypes) and ramets (stems with identical multilocus genotypes) in natural Spartina patens populations. These data, coupled with the results of simulations, will be used to estimate clonal reproductive rates within and among populations as well as ecological correlates of the rate of clonal reproduction. Understanding the mechanisms that cause of spatial patterns of genetic variation in S. patens can also be used to define and test hypotheses for the mechanisms that influence interactions with Spartina alterniflora. Gina Wimp and I are working to develop an empirical system to study eco-evolutionary dynamics in salt marsh plants and insects. We have developed microsatellite genetic markers for two planthopper insect species. With these genetic tools in place, we can study the dynamics of genetic polymorphism in the plant and the insects. We will test if genetic variation within plant patches influences the genetic effective population size and rates of gene flow in the resident insect populations. We also plan to alter ecological conditions such as predator density and nutrient availability, to determine how those variables impact the genetic effective population size in the resident insect populations. |
Molecular clock rate variation
|
The original molecular clock hypothesis predicts that the rate of substitution of neutral mutations should be both constant per year and independent of the effective population size. However, heterogeneity in substitution rates is commonly observed. Such rate heterogeneity may be explained by the action of natural selection or by modifications of the neutral molecular clock. One possible neutral explanation for rate heterogeneity is that substitution rates are not constant per year but constant per generation. In that case, species with different generation times could experience different divergence rates per year. Another possibility is that locus- and lineage-specific natural selection causes substitution rate variation.
Soria-Hernanz, Barverman & Hamilton (2008) observed parallel substitution rates in chloroplast and mitochondrial regions within Brazil nut taxa (Lecythidaceae), with rate variation among taxa, suggesting that rate heterogeneity was caused by lineage differences rather than locus-specific natural selection. The generation time hypothesis in plants has focused on comparing substitution rates in annual and perennial species. In the pre-genomics era, tests of the generation time hypothesis in plants often used the nuclear ribosomal ITS sequences available for many plant taxa. Soria-Hernanz et al. (2008) showed that ITS regions lack power to distinguish relatively small substitution rate differences (< 3-fold), and that the annual Arabidopsis thaliana had a faster substitution rate than the close perennial relative A. lyrata at five nuclear loci providing more power to make the comparison. Braverman et al. (2016) used 256 genes to compare substitution rates in annual-biennial-outgroup triplets (annual Arabidopsis thaliana and biennial Arabidopsis lyrata with outgroups Brassica rapa, Capsella grandiflora, and Neslia paniculata; annual Brassica rapa and biennial Brassica oleracea with outgroup N. paniculata). The Arabidopsis comparisons consistently showed faster rates for the annual lineage. ANOVA showed about 6% of rate variation between the Arabidopsis and Brassica species trios, about 26–32% of rate variation among loci. In one ANOVA model using substitution rates from genes partitioned into odd and even codons, about 13% of substitution rate variation was due to lineage-by-locus interaction consistent with natural selection but interaction was not a significant without partitioned genes. Further work on substitution rate heterogeneity will utilize genome-scale sequence data in order to gather many loci that have the power to estimate potentially small divergence rate differences. Current work is ongoing with plants as well as insects. |
Population genetics of striped bass |
Striped bass (Mornone saxatalis) are a migratory fish species found along the east coast of the U.S. between North Carolina and Maine. They are anadromous (live as adults in salt water but reproduce in fresh water) and the largest spawning population is in the Chesapeake Bay. This predatory fish plays a crucial role in estuary and near-shore marine ecosystems and also forms the basis of recreational and commercial fisheries that generate millions of dollars of economic activity each year. There are several ongoing projects in the lab that have the goal of understanding genetic variation and population demographic history in this species.
Evolutionary responses to global change are expected to occur over decades, especially in long-lived organisms. Species are expected to evolve in response to changing environments, yet relatively few species have genetic data sets available to test hypotheses about the evolutionary mechanisms that will determine responses that span the range of successful adaptation to persistence to extinction. This project will address those knowledge gaps through both retrospective and prospective sampling and next-generation DNA sequencing (NGS) to directly observe the temporal and spatial evolutionary responses to long-term environmental change. The genetic data will be used to test for loci experiencing local adaptation, to estimate patterns of population differentiation and gene flow over time, and to estimate contemporary and historic genetic effective population sizes to quantify genetic drift. This project will collect a five-decade longitudinal genetic data set for the anadromous fish striped bass (Morone saxatilis) to test for past changes in local adaptation, and the organization of genetic variation within and among populations. Striped bass are long lived, migrate seasonally along the Northeast mid-Atlantic U.S. coast, and spawn in estuaries in Albemarle Sound (NC), Chesapeake Bay (MD and VA), the Delaware River (DE, PA & NJ), and the Hudson River (NY) that are predicted to experience above-average sea level rise in the coming decades. Based on numerous life history and environmental exposure variables, striped bass were ranked “very high” for climate vulnerability in a recent study of species of the Northeast U.S. continental shelf marine ecosystem. Further, striped are actively managed since they support major recreational and commercial fisheries, and the mid-Atlantic population is considered overfished. The availability of archival genetic samples spread evenly over time make this a unique project to examine genetic responses to environmental change over five decades. Both archived fish scale samples and young-of-year fish samples collected during annual recruitment surveys are available as a source of genomic DNA from 1970 to 2024 age cohorts at intervals of about 6-8 years. Single nucleotide polymorphism (SNP) data will be obtained from these samples using reduced-representation high-throughput sequencing (RAD-seq). Simulations on the draft striped bass genome predict The project will also expand an existing genetic data of 17 microsatellite genetic marker loci to include about 900 total individuals from seven Chesapeake Bay age cohorts 1970-2006, as well as the 2012 age cohort in Chesapeake Bay, Delaware and Hudson. |
Tropical forest genetics and forest fragmentation (older research) |
A long-term project in my lab revolves around the genetic impacts of tropical deforestation. This study utilizes experimentally created forest fragments in the Central Amazon region of Brazil at the Biological Dynamics of Forest Fragments Project (BDFFP). Within these fragments I am working to estimate levels of population structure and gene flow for the canopy tree Corythophora alta, a member of the Brazil nut family (Lecythidaceae).
Primary growth tropical forest is currently undergoing rapid conversion to pasture, dense second-growth or degraded forest because of timber harvest, agriculture, and settlement. Deforestation is a progressive process that seldom results in the total removal of forest cover. In many instances, removal of mature forests results in a mosaic landscape composed of cleared areas and forest fragments. Although short-term effects of fragmentation are clearly evident, there is less data to address biological impacts that will be manifest over many years. Possible long-term impacts of deforestation (over multiple generations of plants and animals) will depend heavily on hereditary factors and evolutionary dynamics within and among modified forest patches. |
A LandSAT image of the Biological Dynamics of Forest Fragments Project study area in 1995. Dark green is intact forest, light green young second-growth forest and red is pasture or bare soil. Isolated forest fragments of one, 10 and 100 hectares are visible as dark-green squares within light green areas. Trees in 10 hectare plots of continuous forest are available as controls.
The overall goals of this study were to compare patterns of pre-deforestation gene flow to post-deforestation gene flow in replicated forest fragments and continuous forest controls. This project has two main components.
First, to determine if a change has taken place, it is necessary to have estimates of some initial pattern that can be compared with some later pattern measured after forest fragmentation. The initial pattern is the degree of population genetic structure in adult trees since these individuals established well before the relatively recent forest isolation events. InC. alta, data from nuclear microsatellite loci estimate FST as less than 0.1 while data from chloroplast insertion-deletion polymorphisms estimate FST as about 1.0. This finding suggests that in historic C. alta populations, pollen gene flow was extensive while seed gene flow was extremely restricted (the results are indistinguishable from seed gene flow 400 times less than pollen gene flow in a island model).
First, to determine if a change has taken place, it is necessary to have estimates of some initial pattern that can be compared with some later pattern measured after forest fragmentation. The initial pattern is the degree of population genetic structure in adult trees since these individuals established well before the relatively recent forest isolation events. InC. alta, data from nuclear microsatellite loci estimate FST as less than 0.1 while data from chloroplast insertion-deletion polymorphisms estimate FST as about 1.0. This finding suggests that in historic C. alta populations, pollen gene flow was extensive while seed gene flow was extremely restricted (the results are indistinguishable from seed gene flow 400 times less than pollen gene flow in a island model).
Bee pollinated Corythophora alta has small hooded flowers that each last a single day but are produced in small numbers for many months (right). The tree has large woody fruits (left) that contain meaty seeds, a feature common to members of the Brazil nut family (Lecythidaceae). Seeds are presumably dispersed by primates, birds and rodents.
Second, the degree of population structure after forest fragmentation was estimated by examining patterns of mating among trees in both forest fragments and continuous forest. The paternity analysis results suggest that rates of on-plot versus off-plot gene flow are not different between fragments and continuous forest. In addition, rates of self-fertilized seeds were about 10 percent in both fragments and continuous forest. These paternity estimates were obtained with an extension of the standard exclusion model that allows for genotyping error in paternity assignment and incorporates selfing when computing the probability of a random match to the inferred paternal haplotype.
Corythophora alta occurs at a density of 1 to 4 individuals per hectare in the BDFFP area. Trees are tall and sampling is slow and dangerous. Notice the woodsman on the tree trunk clipping branches with a pruning pole to obtain a leaf sample for DNA extraction. Many of the DBFFP woodsman are expert tree climbers and scale trees using only a web belt around their feet. This method of tree climbing does not damage trees. I am greatly indebted to them for sharing their knowledge of the forest and their hard work.
The ultimate goal is to determine if deforestation alters historic patterns of genetic organization and the rate at which deforestation may cause genetic changes in tropical tree populations. Such quantitative measures of the impact of tropical forest fragmentation on the evolutionary dynamics of genetic variation are needed to predict the long-term impacts of deforestation. Estimating the effects of forest fragmentation on gene flow will help us to predict the genetic consequences of further habitat degradation and to intelligently design tropical forest reserves.
This work has received generous support from the Biological Dynamics of Forest Fragments Project, the Smithsonian Institution, the National Zoological Park, Friends of the National Zoo and the National Geographic Society. "This material is based upon work supported by the National Science Foundation under Grant No. 9983014. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation."
This work has received generous support from the Biological Dynamics of Forest Fragments Project, the Smithsonian Institution, the National Zoological Park, Friends of the National Zoo and the National Geographic Society. "This material is based upon work supported by the National Science Foundation under Grant No. 9983014. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation."