 |
 |
 |
|
|
|
|
|
Evolutionary biology research can be broadly separated into two categories - "pattern" research that attempts to elucidate specific aspects of the history of life on earth (e.g., patterns of relationship, trends over time, historical causes of geographic distributions, etc.) - and - "process" research that focuses on the causes of evolutionary changes (e.g., specific selective factors that operate on organisms, constraints that limit and shape evolutionary change, etc.). I am very "process" oriented in my interests, what interests me is why things happen more so than the particulars of what exactly came about. Given that focus, I study evolutionary biology using both theoretical and empirical approaches. In particular, the broad question that I am interested in is what influences the evolvability, the capacity to evolve, of organisms and species. Evolvability can be influenced by characteristics of populations (e.g., larger populations can maintain more genetic diversity and may be more prepared for circumstances that require rare or new alleles) or individuals (e.g., the degree to which organisms can buffer their development against environmental stress may allow them to better form adult morphologies that confer higher fitness). I am interested in both aspects of evolvability, three examples of studies I have conducted are described below. I do not feel bound to any specific system or technique, I am open to using the appropriate tools to study whatever interesting questions I encounter. In addition to the theoretical and empirical work described on this page I am conducting comparative studies of allometry using measurements from the literature. I enjoy nothing more than debating a question with a collaborator or student and devising a study to resolve or address the issue. Some examples of issues I have studied follow. The lab is equipped with infrastructure necessary for both types of research. For theoretical work there are a number of computers appropriate for running large scale computer simulation studies (e.g., the first example study described below). I use these computers extensively, but they are available to any of my students (and department members) interested in conducting computational research. For empirical work there is general bench space, an incubator array, Drosophila optimized workstations and microscope stations for examining morphological traits. This space is ideal for examining phenotypes in individuals and populations and designing experiments around the gathering of such data. An example study of population evolvability A growing pattern of evidence suggest that the genetic sequences of regulatory regions are evolving much faster in non-vertebrates than in vertebrates even as function is preserved. There seems to be a difference in the evolvability of the genetic sequences of these regulatory regions. In addition to many other aspects of their physiology and biochemistry these groups differ in their typical population sizes, invertebrates usually have larger populations. I hypothesized that this difference in population size may account for the difference in evolutionary rate seen. In particular, I investigated a mathematical and computer simulation model of compensatory evolution (where double mutants are slightly advantageous, but the intermediate single mutant state is deleterious). This model predicted much faster rates of double mutation fixation in populations of very large size, contrary to previous results indicating that larger populations are much less able than smaller ones to cross such fitness valleys. Since non-vertebrates have larger average population sizes than vertebrates this model serves as a possible mechanism for the observed disparity between observed rates of enhancer sequence evolution.
Two Example studies of individual evolvability Natural selection acts upon phenotypic variation in populations by favoring individuals with certain phenotypic values. It is well known that the phenotypic value that an individual possesses is the combined result of an organism's genetically determined target phenotype and any environmental factors that may alter development (e.g., less food creates smaller phenotypes regardless of genotype). The relative importance of an organism's developmental stability (DI, the ability of an organism develop a target phenotype given the appropriate genotype despite environmental perturbations) is unknown. With two collaborators I estimated the magnitude of this DI variance using published fluctuating asymmetry (FA, the non-consistent differences between the left and right sides of an organism) data. Phenotypic variance due to DI (as estimated from FA) was compared to overall trait variance to provide a sense of how precisely organisms are able to achieve their selectively optimal phenotypes. The relative trait variance in natural population due to DI showed mean (15%) and median (6%) values that indicate that DI is a non-negligible factor that contributes to trait non-optimality in populations. An organism's DI therefore contributes in a real way to it's evolvability by altering phenotypic noise that reduces the efficiency of selection.
Given that DI influences evolvability, does selectable genetic variation for DI exist? Can organisms be selected to change this aspect of their evolvability? Consider bilaterally symmetrical organisms; if DI is high we expect the sides to be more mismatched in size (higher FA) whereas if DI is low we expect the sides to be better matched in size (lower FA). The average FA of a population can only be altered with selection if genetic variation for the underlying cause (DI) is genetically variable. To examine whether this is the case I performed an artificial selection experiment to increase and decrease the FA of wing vein distance traits in Drosophila. All four selected lines responded to the selection, revealing heritability estimates for FA of 0.01 in the up lines and 0.006 in the down lines. Converting these FA heritability estimates into DI heritability estimates provided estimates ranging from 0.3 to 0.5 for the heritability of DI. Since DI is heritable it is a selectable trait - albeit a trait that is selected not to form a phenotype directly, but to form it more consistently - and therefore a target of natural selection. This selection acts not to change a phenotype itself, but to change the reliability with which that phenotype is realized and, by extension, the efficiency of selection to change it in the future, an aspect of evovability.
|
 |
 |
 |
 |
 |