Research projects in
the Barrett lab

Here is some of the research currently being undertaken in the Barrett lab with background and the particular projects that I have funding to support. Students or PDFs that are interested in these topics and who would like to work in my lab should contact me by e-mail.

Evolutionary transitions from outcrossing to selfing

The transition from outcrossing to selfing is very common in flowering plants and has important demographic, genetic and evolutionary consequences. We have worked on this problem using two main systems (Turnera, Eichhornia); both involve the evolutionary breakdown of heterostyly (58, 72). Most of our focus in recent years has been on the neotropical, annual, aquatic Eichhornia paniculata, where we have evidence of multiple independent transitions to selfing (115, 259). This species is an excellent study system as it is diploid, easily crossed and we can grow 2-3 generations a year in our glasshouses at Toronto. We have extensive collections of E. paniculata from throughout its range ( Brazil, Jamaica, Cuba, Central America). Work by Brian Husband, Martin Morgan and Josh Kohn in the 90s demonstrated that the shift to selfing in E. paniculata results from the joint action of genetic drift and natural selection and represents one of the few cases that meet several of the key conditions in Sewall’s Wright’s shifting balance theory of evolution (97, 100, 102, 124).



Crosby pic

Tristyly breakdown pic

E paniculata

Floral and mating system diversity in Eichhornia (Pontederiaceae) species (143). Eichhornia paniculata – A) a population in a Cuban rice field; B) inflorescence; C) outcrossing and selfing flowers from Brazil and Jamaica, respectively. Style morph frequencies in natural populations indicate the breakdown pathway to selfing from the 1:1:1 equilibrium in trimorphic populations. This results from the joint interaction of genetic drift and natural selection. The pathway involves three stages with sequential loss of morphs from trimorphism to dimorphism and finally to monomorphism (72, 259).


Current work is focused on the molecular population genetic and genomic consequences of transitions to selfing in E. paniculata (242, 263). During his Ph.D., Rob Ness in collaboration with PDF Mathieu Siol developed extensive genomic resources for this species, including de novo assemblies of the floral transcriptomes of outcrossing and selfing genotypes. This was conducted using short read sequences from the Illumina Solexa GA II available in our department at CAGEF . These resources have provided us with ~10,000 single copy EST’s. We have also identified 269 genes associated with floral development, 22 of which are differentially expressed in selfing lineages. These represent a set of potential candidate genes for investigating the evolution of the selfing syndrome.

I would like to recruit a PDF and a Ph.D. student to continue with this project to investigate the genetic architecture of the selfing syndrome by identifying genomic regions responsible for traits governing the transition to selfing, as well as identifying the genes governing tristyly. Also, by looking at genome-wide polymorphism data it will be possible to investigate the reduced efficacy of selection that should occur in selfing populations as a result of their lower effective population sizes. Our recent review of the population genomics of plant adaptation (266) illustrates that there is considerable scope for investigating a range of questions in molecular population genomics and evolution concerning how selection can shape genomes in plant populations. Eichhornia paniculata provides a valuable system for addressing many of these questions in the context of mating-system variation.



Floral diversity in insect-pollinated Wurmbea (Colchicaeae) species (227). Populations of these species are either monomorphic or dimorphic for gender (243).

Gender strategies and the evolution of sex ratios

I have a long-standing interest in gender strategies, and in trying to understand the costs and benefits of combined versus separate sexes. Earlier work involved studies of Aralia (16, 18) and Wurmbea (179, 227) but more recently we have concentrated our efforts on the clonal, aquatic plant genus Sagittaria, particularly S. latifolia. This species is common in eastern N. America and populations are mostly either monoecious or dioecious and occupy different wetland habitats. Although plants of the two sexual systems are fully inter-fertile, there is little gene flow between them over most of their range because of ecological and life-history differentiation. We have exploited the intra-specific variation in sexual systems of S. latifolia in an effort to understand the ecological and genetic mechanisms responsible for the evolution and maintenance of dioecy (177, 184, 198, 205, 216). Sarah Yakimowski has recently developed microsatellite markers for this species (257) and this has allowed us to investigate the breakdown of sexual system differentiation at range limits where populations are composed of females, males and hermaphrodites (267).

Herm Sagittaria Fem Sagittaria Male Sagittaria

Inflorescences of hermaphrodite, female and male plants of Sagittaria latifolia (Alismataceae)


I would be interested in supervising future student projects on Sagittaria and these could involve investigations of: 1) the genetic basis of life history differentiation in monoecious and dioecious populations using QTL-mapping, 2) the ecological mechanisms responsible for sexual system differentiation through reciprocal transplant experiments; 3) the role of clone size, geitonogamous selfing and inbreeding depression in the spread of unisexuals in monoecious populations.

We have also been working on the problem of sex ratio bias in dioecious populations. David Field and Melinda Pickup, two PDF’s currently in the lab, are preparing a review paper concerning the ecological and life history correlates of biased sex ratios in dioecious species. In addition, we have been working on Rumex species, a group of wind-pollinated herbs characterized by female biased sex ratios. We are interested in determining the mechanisms responsible for this female bias, which is less common in flowering plants than male bias. Working with R. nivalis, former PDF Ivana Stehlik demonstrated that bias results from both microgametophyte selection and gender based differences in mortality (223, 226, 233, 249). This work was greatly facilitated by the development of sex–specific markers. We are now exploring sexual dimorphism and the mechanisms governing female bias in R. hastatulus, an annual species native to the southern USA ranging in distribution from Texas to North Carolina. Specifically, Melinda Pickup is investigating three main questions: 1) What is the extent of geographical variation in sexual dimorphism; 2) to what extent does the local mating environment influence the degree of female bias in progeny sex ratios; 3) what is the relative importance of plant density and local sex ratio variation on progeny sex ratios.

I am interested in recruiting a Ph.D. student and a PDF to begin collaboration with my colleague Stephen Wright in EEB on sex chromosome evolution in R. hastatulus. This species is an excellent model system because it is annual, very easy to grow and cross, and we also have a broad sampling of geographical variation. The species is of particular interest because it possesses two distinct mechanisms of sex determination. In populations from the western portion of the range females are XX and males are XY, whereas in eastern populations females are XX and males are XYY. Phylogenetic evidence indicates that species in the genus with two Y-chromosomes are derived from those with a single Y. We are interested in investigating various aspects of the molecular population genetics and molecular evolution of X- and Y-chromosomes in this group. This will be achieved by obtaining sex-linked markers using genomic approaches and high throughput sequencing. Specifically, we will investigate the hypothesis of Y-chromosome degeneration and reduced effective population size of the Y-chromosome resulting from suppression of recombination and the fixation of deleterious mutations. This will be undertaken by comparing patterns and amounts of nucleotide diversity and molecular evolution across regions of the X and Y chromosome, and with autosomal genes. We are also interested in investigating the relation between sex-linked genes and the degree of sexual dimorphism, and the extent to which female bias in populations may influence patterns of diversity. Finally, obtaining a suite of sex-linked markers will also provide an outstanding resource for future experimental studies aimed at dissecting the gender-based life-cycle selection that results in female bias.



Rumex nivalis and R. hastatulus (Polygonaceae), dioecious wind-pollinated species with female-biased sex ratios (223).


Floral biology, pollination and mating

Many plant species possess floral polymorphisms and understanding their evolution and functional significance has been of considerable interest to evolutionary biologists since Darwin’s pioneering studies on the topic summarized in “Forms of Flowers” (1877). We have investigated a range of problems associated with floral polymorphisms including heterostyly (39, 92, 103), stigma-height dimorphism (187, 196, 215), enantiostyly or mirror image flowers (190, 200, 201), heteranthery (254, 270) and various sexual polymorphisms (168, 229, 247, 262, 269) using diverse approaches. These have included comparative and phylogenetic methods (136, 199, 214, 241), floral manipulations and marker genes (97, 104, 130), and experiments focusing on the interaction of flowers and their pollinators (114, 119, 254). My own recent reviews on the topic of plant sexual diversity (188, 262) identify many unanswered questions that would repay detailed investigation and students interested in projects on reproductive biology are always welcome in the lab.


Flowering plant species with heteranthery; the occurrence of two distinct anthers within bee-pollinated flowers that differ in form and function. Heteranthery is commonly associated with enantiostyly, buzz pollination and nectarless flowers (254, 267).


Floral diversity in Wachendorfia (Haemodoraceae). The species exhibit dimorphic enantiostyly in which populations are often composed of plants with styles that are deflected to either the right or the left (185). This condition promotes cross-pollination (190).


Floral diversity in Narcissus (Amaryllidaceae). The genus exhibits four distinct forms of stylar variation – stylar monomorphism, stigma-height dimorphism, distyly and tristyly (214).


Crosby pic

In collaboration with Bruce Anderson ( University of Stellenbosch, South Africa) we have recently investigated the origin and function of the specialized bird perch in Babiana, a genus of geophytes native to the Western Cape (220). Caroli de Waal recently completed her M.Sc. thesis in my lab on the reproductive ecology of four bird pollinated Babiana species, two of which possess specialized bird perches. She investigated the pollination biology, mating systems and patterns genetic diversity of these species and found that in B. ringens there was evidence for the loss of perch function associated with increased self-pollination in populations at the periphery of the eastern range of the species. Future work on the possibility of pollinator shifts involving different types of sunbirds visiting populations and the loss of bird pollination and evolution of selfing could be profitably undertaken.


Malachite sunbird (male) visiting flower of Babiana ringens with an intact perch and a plant with the perch removed. Perch removal resulted in a striking reduction in fertility and an increase in the selfing rate (220).

(Figure on left) Bird-pollinated Babiana (Iridacaeae) species endemic to the Western Cape, South Africa. a) B. ringens subsp. ringens; b) B. ringens subsp. australis; c) B. hirsuta; d) B. carminea; e) B. avicularis. Both B. ringens and B. avicularis exhibit a specialized bird perch composed of the naked inflorescence axis. All species are pollinated by sunbirds.


A selection of wind-pollinated species with diverse morphologies: A) Carex pedunculata (Cyperaceae) see (256); B) Gynomonoecious Gunnera peltata (Gunneraceae); C) Leucadendron rubrum (Proteaceae). This species is dioecious with extreme sexual dimorphism (female left; male right); D) Pennisetum clandestinum (Poaceae); note the anthers extended on very long filaments to aid in pollen dispersal. Image by L.D. Harder

The evolution of wind pollination is poorly understood despite the fact that it represents one of the most important evolutionary transitions in pollination systems. A recently completed Ph.D. thesis in my laboratory by Jannice Friedman, now a PDF with John Willis at Duke University, involved the first comprehensive modern treatment of the ecology and evolution of wind pollination. Jannice addressed a range of questions concerning mating and pollination in anemophilous plants using diverse approaches including comparative methods, theoretical models, field experiments, marker gene studies of mating patterns, and glasshouse quantitative genetic analyses. Her work has helped to explain a number of long standing questions including why anemophilous species commonly have unisexual flowers with only a single ovule.

This diversity of projects on plant reproduction that I have supervised reflect my own strong interest in this area and I would be happy to take on students who might be interested in pursuing any of these topics further.





Evolutionary genetics and local adaptation in invasive species

Ever since I worked as a weed biologist in Swaziland and the Lower Amazon Basin, I have been fascinated by the ecology and evolution of invasive plants. Where do they come from and why are they invasive? My early studies concerned two quite different problems: 1) crop mimicry in barnyard grasses (24); 2) the relative importance of clonal vs. sexual reproduction in water hyacinth (10, 11). More recently, in collaboration with my former Ph.D. students Chris Eckert (Queens University) and Rob Colautti (now a PDF with Tom Mitchell-Olds at Duke) we have been investigating the evolution of local adaptation in the wetland invader Lythrum salicaria. By sampling seed collected from representative populations across the invasive eastern North American range and using common garden and quantitative genetic analyses we have demonstrated: 1) abundant heritable variation for most life history traits; 2) reduced variance and increased skew of quantitative traits with northward migration; 3) strong clines in flowering time and plant size associated with latitude and length of growing season; and 4) evidence for genetic constraints at range limits between flowering time and size (265). Further, reciprocal transplant experiments and measurements of natural selection have provided evidence indicating a fitness advantage to plants that flower earlier at a smaller size in northern populations (271). There are still many fascinating questions concerning the invasion genetics of purple loosestrife and I would be interested in collaborating with a student on this system. In addition, I have recently completed a collaborative study with colleagues in China from Beijing Normal University on a world wide survey of genetic diversity in water hyacinth (268). This study highlights the need for more intensive surveys of populations in the native range and for studies of genotype by environment interactions to investigate the ecological tolerances of clones from different regions.

Invasive spp

Invasive species with diverse reproductive systems investigated in the Barrett Lab; clonal water hyacinth (Eichhornia crassipes, Pontederiaceae) in its native (1a, Argentina) and introduced (1b, Vietnam) range, see 21, 268; the wind-pollinated, annual, common ragweed (2a,b Ambrosia artemisiifolia, Asteraceae) native to eastern North America (246); 3, the perennial, distylous, ruderal Turnera subulata (Turneraceae), native to Brazil and introduced to S.E. Asia (5); 4, purple loosestrife (Lythrum salicara, Lythraceae), a tristylous wetland invader native to Eurasia and common in wetlands of eastern N. America (139, 237, 264, 265).