Kent E. Holsinger

Genetics in geographically structured populations: defining, estimating and interpreting FST

Wrights F-statistics, and especially FST, provide important insights into the evolutionary processes that influence the structure of genetic variation within and among populations, and they are among the most widely used descriptive statistics in population and evolutionary genetics. Estimates of FST can identify regions of the genome that have been the target of selection, and comparisons of FST from different parts of the genome can provide insights into the demographic history of populations. For these reasons and others, FST has a central role in population and evolutionary genetics and has wide applications in fields that range from disease association mapping to forensic science. This Review clarifies how FST is defined, how it should be estimated, how it is related to similar statistics and how estimates of FST should be interpreted.

A Bayesian approach to inferring population structure from dominant markers

Molecular markers derived from polymerase chain reaction (PCR) amplification of genomic DNA are an important part of the toolkit of evolutionary geneticists. Random amplified polymorphic DNA markers (RAPDs), amplified fragment length polymorphisms (AFLPs) and intersimple sequence repeat (ISSR) polymorphisms allow analysis of species for which previous DNA sequence information is lacking, but dominance makes it impossible to apply standard techniques to calculate F‐statistics. We describe a Bayesian method that allows direct estimates of FST from dominant markers. In contrast to existing alternatives, we do not assume previous knowledge of the degree of within‐population inbreeding. In particular, we do not assume that genotypes within populations are in Hardy–Weinberg proportions. Our estimate of FST incorporates uncertainty about the magnitude of within‐population inbreeding. Simulations show that samples from even a relatively small number of loci and populations produce reliable estimates of FST. Moreover, some information about the degree of within‐population inbreeding (FIS) is available from data sets with a large number of loci and populations. We illustrate the method with a reanalysis of RAPD data from 14 populations of a North American orchid, Platanthera leucophaea.

Polytomies and Bayesian Phylogenetic Inference

Bayesian phylogenetic analyses are now very popular in systematics and molecular evolution because they allow the use of much more realistic models than currently possible with maximum likelihood methods. There are, however, a growing number of examples in which large Bayesian posterior clade probabilities are associated with very short branch lengths and low values for non-Bayesian measures of support such as nonparametric bootstrapping. For the four-taxon case when the true tree is the star phylogeny, Bayesian analyses become increasingly unpredictable in their preference for one of the three possible resolved tree topologies as data set size increases. This leads to the prediction that hard (or near-hard) polytomies in nature will cause unpredictable behavior in Bayesian analyses, with arbitrary resolutions of the polytomy receiving very high posterior probabilities in some cases. We present a simple solution to this problem involving a reversible-jump Markov chain Monte Carlo (MCMC) algorithm that allows exploration of all of tree space, including unresolved tree topologies with one or more polytomies. The reversible-jump MCMC approach allows prior distributions to place some weight on less-resolved tree topologies, which eliminates misleadingly high posteriors associated with arbitrary resolutions of hard polytomies. Fortunately, assigning some prior probability to polytomous tree topologies does not appear to come with a significant cost in terms of the ability to assess the level of support for edges that do exist in the true tree. Methods are discussed for applying arbitrary prior distributions to tree topologies of varying resolution, and an empirical example showing evidence of polytomies is analyzed and discussed.

Mass-Action Models of Plant Mating Systems: The Evolutionary Stability of Mixed Mating Systems

In self-compatible plants the selfing rate is directly related to the relative amounts of self and outcross pollen deposited on receptive stigmas. As a result, the fraction of ovules self-fertilized by a particular plant is a function not only of the fraction of the pollen it produces that is deposited on its own stigma but of the amount of pollen exported by other plants in the population and the probability that exported pollen successfully reaches its receptive stigmas. Thus, the reproductive success of different genotypes is both frequency- and density-dependent. In dense populations in which the probability of successful pollen export is relatively high, complete outcrossing is evolutionarily stable, even in the absence of inbreeding depression. In sparse populations, genotypes that increase the selfing rate in the population may increase in frequency, but genotypes that cause complete selfing can never become fixed, unless reproduction is pollen-limited. Whenever self-fertilization can evolve, mixed mating systems are evolutionarily stable. This model predicts that selfing rates should be both density-dependent and frequency-dependent. The limited data currently available are consistent with both of these predictions, although there are several complications in attempting to apply the mass-action model directly to biotically pollinated plants. These results suggest that the probability of successful outcrossed reproduction plays at least as important a role in determining the mating system of plants as inbreeding depression.

INBREEDING DEPRESSION DOESN'T MATTER: THE GENETIC BASIS OF MATING-SYSTEM EVOLUTION'

Models of mating‐system evolution commonly assume that inbreeding depression is independent of the genotype at loci determining the mating system. Because of the association that develops between genotypes at different loci in inbred populations, an individual that is heterozygous at a mating‐type locus is more likely to be heterozygous at a fitness locus than is a randomly chosen individual. A modifier model for the evolution of self‐fertilization in plants demonstrates that inbreeding depression is not an adequate descriptor of the relative fitness of inbred and outbred progeny. If inbreeding depression is primarily a result of segregation at overdominant loci, intermediate rates of self‐fertilization may be favored, even if the inbreeding depression is less than 30%. Indeed, in some cases, mutants that cause some outcrossing can be introduced into completely selfing populations when the inbreeding depression is as little as 1%. If inbreeding depression is primarily a result of the expression of recessive lethals in inbred progeny, selfing can evolve in an initially random mating population, even when the inbreeding depression is over 70%.

Bayesian approaches for the analysis of population genetic structure: an example from Platanthera leucophaea (Orchidaceae)

We describe four extensions to existing Bayesian methods for the analysis of genetic structure in populations: (i) use of beta distributions to approximate the posterior distribution of f and θB; (ii) use of an entropy statistic to describe the amount of information about a parameter derived from the data; (iii) use of the Deviance Information Criterion (DIC) as a model choice criterion for determining whether there is evidence for inbreeding within populations or genetic differentiation among populations; and (iv) use of samples from the posterior distributions for f and θB derived from different data sets to determine whether the estimates are consistent with one another. We illustrate each of these extensions by applying them to data derived from previous alloyzme and random amplified polymorphic DNA surveys of an endangered orchid, Platanthera leucophaea, and we conclude that differences in θB from the two data sets may represent differences in the underlying mutational processes.

Phylogenetic analysis of chloroplast DNA restriction site data at higher taxonomic levels: an example from the Asteraceae

Chloroplast DNA variation was examined among 57 genera of Asteraceae representing 15 currently recognized tribes. Complete cleavage maps were constructed for 11 six‐base pair restriction enzymes, and a total of 927 cleavage site differences was detected, 328 of which are phylogenetically informative. The data were used to construct phylogenetic trees using both Wagner and Dollo parsimony and the resulting monophyletic groups were evaluated statistically using the bootstrap method. The level of homoplasy in the restriction site data is 54–56% (excluding autapomorphies), most of which is due to parallel site losses. The most parsimonious trees generated by both parsimony methods have nearly identical topologies at lower taxonomic levels, but differ in subfamilial circumscriptions and tribal groupings. Dollo parsimony provides support for the monophyly of two subfamilies, the Lactucoideae (excluding the Barnadesiinae) and Asteroideae, but Wagner parsimony supports the monophyly of the Asteroideae only. This incongruence is due to different assumptions of the two parsimony methods about relative rates of parallel site gains and losses. After eliminating the six most rapidly changing restriction sites or performing successive approximation, Wagner parsimony produces trees with the same subfamilial groupings as the Dollo trees. We conclude that the Dollo tree with two monophyletic subfamilies is the best estimate of phylogenetic relationships in the Asteraceae because this method more accurately reflects the evolution of restriction sites. We also demonstrate that in spite of high levels of homoplasy in chloroplast DNA restriction site data at these higher taxonomic levels, it is possible to make statistically supported estimates of phylogenetic relationships.

Dispersal and plant mating systems: the evolution of self-fertilization in subdivided populations

Intermediate rates of self‐fertilization can be evolutionarily stable when the progeny of self‐fertilization events are less successful migrants than those of outcrossing events, unless self‐fertilization reduces an individuals contribution to the pollen pool by an amount equal to the rate at which it self‐fertilizes. This result holds regardless of whether pollen or diaspores are more widely dispersed. The differential migration of selfed and outcrossed progeny may be a result of differential establishment with comparable rates of dispersal, or it may be a result of differential dispersal rates. In the first case, detailed predictions concerning the evolutionarily stable selfing rate can be made. In the second case, only qualitative predictions are possible in the absence of specific assumptions about how the migration rate is affected by the average selfing rate in each subpopulation.

THE EFFECT OF TOPOLOGY ON ESTIMATES OF AMONG-SITE RATE VARIATION

Among-site rate variation, as quantified by the gamma-distribution shape parameter,a or α, and the ratio of transition rate to transversion rate (Ts/Tv) influence phylogenetic inference. We examine the effect of topology on estimates of these two parameters in 12S rRNA sequences from nine species of mice belonging to the generaOnychomys andPeromyscus by generating 100 random topologies and estimating these parameters using parsimony and maximum-likelihood methods for each of the random topologies. The parsimony-based estimate ofTs/Tv from the well-corroborated topology falls within the distribution of estimates based on random topologies, whereas the maximum-likelihood estimate ofTs/Tv based on the well-corroborated topology lies well outside the distribution of estimates derived from random topologies. TheTs/Tv ratio derived via maximumlikelihood estimation is three times the parsimony-based estimate, suggesting that parsimony-based estimates are severe underestimates even when the correct topology is used. Both parsimony- and likelihood-based estimates of the gamma-distribution shape parameter (α) are sensitive to topology because the best estimates based on the well-corroborated topology are well outside the distributions of estimates derived from random topologies for both methods. We show that the reason for topology dependence is the presence of long internal branches in the underlying topology.

Ecological and evolutionary consequences of biological invasion and habitat fragmentation

There is substantial evidence that environmental changes on a landscape level can have dramatic consequences for the species richness and structure of food webs as well as on trophic interactions within such food webs. Thus far, the consequences of environmental change, and particularly the effects of invasive species and the fragmentation and isolation of natural habitats, have most often been studied in a purely ecological context, with the main emphasis on the description of alterations in species abundance and diversity and trophic links within food webs. Here, we argue that the study of evolutionary processes that may be affected by such changes is urgently needed to enhance our understanding of the consequences of environmental change. This requires an approach that treats species as dynamic systems with plastic responses to change rather than as static entities. As such, phenotypic plasticity on an individual level and genotypic change as a population level response should be taken into account when studying the consequences of a changing world. Using a multidisciplinary approach, we report on recent advances in our understanding, identify some major gaps in our current knowledge, and point towards rewarding approaches to enhance our understanding of how environmental change alters trophic interactions and ecosystems.