are most widely recognized for their beautiful, vibrantly
colored flowers that appear in the early spring (Fig. 1). Known
as chasmogamous flowers, these are different from the bud-like
cleistogamous flowers that are produced later in the season.
Chasmogamous flowers have been referred to in the literature,
painted through the centuries, and even appear today on everyday
items such as glassware, linen, and clothing. Despite the
popular attention to these flowers, scientists have not
completely understood their purpose.
People have often assumed that chasmogamous flowers were
produced to attract insects, resulting in outcrossing between
two different plants. This is also suggested by the flower
name itself - chasmogamous - which is Greek for "open
marriage" (i.e., a union between two plants). The renown
violet expert, Dr. Ezra Brainerd once suggested that in
several stemless species, "the structure of the flower,
in style, stamens and petals, shows a most ingenious
arrangement to prevent self-pollination" (Clute, 1907).
However, chasmogamous flowers in some non-violet species were
recently found to self-pollinate. Because I was studying Viola
at the time, I was interested in examining chasmogamous
flowers to determine if selfing could also occur in some
stemmed members of the genus. It was already known that
selfing was not possible in flowers of V. riviniana,
V. odorata, and V. canina (Knuth
1908; Beattie 1969; Banasinska and Kuta 1996).
I studied the downy yellow violet, Viola pubescens
(Fig. 2), because little was known of its floral biology even
though it is so common in northeastern beech-maple forests.
Over four years, I found that the chasmogamous flowers were
only occasionally visited by insects. These included bee
flies, skipper butterflies, bumblebees, carpenter bees, and
halictid bees (Fig. 3). To determine if chasmogamous flowers
of V. pubescens were able to self-pollinate, I
covered plants in the forest with mesh cages (to deter
pollinators). The majority (59%) of flowers on these plants
still produced seeds, indicating that the chasmogamous flowers
were able to self-pollinate. Selfing occurred at the end of an
unvisited flower's life as the stigma slowly bent down and
touched pollen grains that had fallen out of the anthers and
were resting on the bottom petal. Known as delayed selfing
(because it only occurs after opportunities for insect
visitation have passed), this mechanism was previously unknown
in V. pubescens, but has been found recently in
another stemmed violet, V. canadensis (Culley,
unpublished data). There also is a brief reference to this
type of selfing in V. arvensis in a 1902 book by
Anton Kerner von Marilaun.
Once I realized that the chasmogamous flowers of V.
pubescens were able to self-pollinate, it was
important to find out how frequently this occurs in nature.
With seeds I had gathered from forest plants, I used a
laboratory technique to determine the rate of selfing over a
two year period. I discovered that over half (60%) of the
seeds produced in chasmogamous flowers the first year resulted
from self-pollinations. This was contrary to the frequent
assumption that the elaborate chasmogamous flowers are largely
outcross-pollinated by insects. However, only 7% of the seeds
originated from selfing the following year. This difference in
selfing rates was probably due to fewer visits by pollinators
in the first year, which resulted in delayed selfing within
the unvisited flowers (that is, higher selfing rates).
(Figure 2.) Cleistogamous (CL) Flower Of The
Downy Yellow Violet, Viola pubescens
So why do chasmogamous flowers of Viola pubescens
retain the ability to self-pollinate, even though they are so
vibrantly dressed in an attempt to lure insect visitors? This
is especially puzzling considering that the flowers of many
other violet species, such as those Dr. Brainerd studied, are
not able to self-pollinate. The answer may lie in the fact
that the insect visitors of V. pubescens are
very unpredictable. In four years of observations, I only saw
pollinators visiting the flowers during one season. Thus, the
plants may not always be able to count on insects being
available to pollinate them. Because the showy chasmogamous
flowers require so much energy (nutrients, etc.) to make, it
would be best for the plant to produce seeds through selfing,
rather than waste the entire flower if pollinators were
absent. Thus, V. pubescens and similar violet
species have an advantage over other plants that require
insect visitation to produce seeds.
The next question that remains to be answered is to determine
why chasmogamous flowers of other violet species cannot
self-pollinate, especially given that it is such an advantage.
Perhaps these plants have adequate pollinators each year, or
there may be other reasons. Hopefully, this question will be
answered in the next few years, as the popularity of Viola
continues to spread and more people are intrigued by its
fascinating floral biology