Most plant species have one type of flower, but several species have two different types of flowers on the same individual:

Chasmogamous Flowers   Cleistogamous Flowers   V. canadensis   V. canadensis   V. pubescens   V. pubescens

Chasmogamous (CH) flowers are often large, showy, and attractive to pollinators.  In contrast, cleistogamous (CL) flowers are very small and resemble buds that form directly into seed capsules.  It is often thought that CH flowers are produced to attract insects and to facilitate outcrossing, while CL flowers are the "back-up  plan" in case pollinators are scarce any particular year.  However, Stewart (1994) has shown that CH flowers may actually be capable of self-pollination in some species.  My research involves determining the factors involved in the production of CH/CL flowers and the genetic consequences of this mating system in violets.

The Viola genus is an excellent group with which to study aspects of the CH/CL system. Nearly all of the North American violets produce both CH and CL flowers, and many species have a widespread distribution. For the purposes of my research, I chose to focus on two stemmed violet species that are found in central Ohio and northern Michigan. Both Viola canadensis (a white-flowered violet ) and V. pubescens (a yellow-flowered violet) are found in the understory of  beech-maple forests. The species are very similar in vegetative morphology and flowering  phenology. To determine the relative advantages (and disadvantages) of the CH/CL mating system in violets, I have conducted a number of studies with the following results:

Flowering Phenology  - CH and CL flowers in V. pubescens are produced at different times and the amount of light may be an important factor in the switchover. CH flowers are produced in the early spring before the forest canopy develops and CL flowers appear later in the season after the trees leaf out. [Int. J. Plant Sci. 163(1): 113-122]
 

CH and CL Seed Production - In 1997, many of the seeds produced in an Ohio population  of V. pubescens came from CH flowers, mainly because more CH flowers successfully  formed and dispersed seed than CL flowers.  On a per plant basis, both floral types produced similar numbers of flowers, and both capsule types had similar numbers of seeds per capsule, in addition to similar abortion and predation rates.  There was no difference in mean seed mass between CH and CL seeds. [Int. J. Plant Sci. 163(1): 113-122]
 

Selfing in CH flowers - CH flowers in both V. pubescens and V. canadensis are able to self-pollinate if left unvisited (through a delayed selfing mechanism).  This has important implications for the population genetics of the species because selfing rates in CH flowers may be higher than previously expected.  A selfing rate study recently revealed that V. pubescens is capable of substanial, but variable selfing rates from year to year (s = 0.07 - 0.60). [Int. J. Plant Sci. 163(1): 113-122]
 

Inbreeding Depression - The level of inbreeding depression was measured in CH  flowers of V. canadensis and was found to be very small (0.10).  Thus, there is little disadvantage to selfing in this species.  In addition, there were some measurable fitness differences between progeny from self-pollinated CH and CL flowers, indicating that floral type differences may exist in CH/CL species. [Can. J. Bot. 78(11): 1420-1429]
 

Genetic Variation - As measured by both isozyme and ISSR  (Inter-Simple Sequence Repeat) molecular markers, the amount of genetic variation is rather high in V. pubescens (e.g., P=77%).   However there still appears to be some genetic structuring, as indicated by a  non-zero Fst (0.29 with isozyme data).  For more information on the ISSR technique, see the ISSR  Resource Website, including a page detailing my primer optimization work. [Heredity 86(5): 545-556]

      Back to Theresa Culley's page

For more information on violets and to meet other Viola enthusiasts, visit the American Violet Society!



---

FAVORITE LINKS:

I also am interested in statistical genetics and population variability estimates.  Here are a few of my favorite programs:

SAS - for all types of statistics (in my opinion, the best out there!)
GDA - Genetic Data Analysis (written by Paul Lewis and Dimitri Zaykin)
BIOSYS-2 - (via FTP);  update of BIOSYS-1.
Mark Miller's Programs - TFPGA, AMOVA-PREP, etc.
PAUP
PHYLIP
NTSYSpc
ISSR Programs - including Vera Ford's programs and RAPDFst (used for dominant marker data).
Phylogeny Programs - a comprehensive list of programs for use by systematists and population geneticists.

Last Updated:  Feb. 12, 2001