Species for Wildflower Seed Production
Anne Marie Johnson and Ted Whitwell
Department of Horticulture, Clemson University
Adobe Acrobat version of this document.
Large scale wildflower plantings, such as in roadside beautification
programs by state transportation departments, are becoming increasingly
popular in the southeast. These highway plantings are usually established
by direct seeding from seed harvested on site, or more commonly, purchased
from growers in the western and southwestern U.S. Western states such as
California have a climate conducive to quality seed production, but some
species native to the southeast are not successfully grown in these major
seed producing areas. For example, Bidens aristosa
, a southeastern
annual, is not available from any of the large wildflower seed companies,
yet this plant is easy to grow in the southeast and is a very attractive
roadside wildflower species.
Previous studies have presented information on germination and storage requirements
for a limited number of wildflower species that are currently available
in the seed trade. There is little information available on seed yields,
germination capacities, seed maturation, and harvesting methods for species
that may have potential for seed production in the southeast.
The objective of this study was to evaluate 29 wildflower species for commercial
seed production in a southeastern environment. Wildflower species were selected
for evaluation based on regional performance and potential for wildflower
sod production (Johnson and Whitwell, 1994). Production potential for these
species was based on seed yield, germination percentage shortly after harvest,
and seed maturation rate.
Materials and Methods
Research plots were located in coastal
plain region of South Carolina on a Goldsboro fine-loamy, siliceous, thermic
Aquic Paleudults. Soil fertility levels were: pH=5.0, 0.0 N-105 P-84 K-lbs/A,
and 0.8% organic matter. The planting area was cleared of existing turf
and weeds by a non-selective herbicide application, and deep tilled (5 inches)
and fumigated with methyl bromide. In December, plots received an application
of 6.3 N-2.4 P-9.1 K lbs/A as a commercial slow-release nitrogen based fertilizer.
Overhead irrigation was used during the study to maintain adequate soil
Seeding date was 7 Oct. Five (summer annuals) of the 29 species were planted
the following March to improve establishment. Seedling rates were based
on supplier's recommended rates for wildflower meadow establishment (Wildseed
Farms of Eagle Lake, TX). Seeds were mixed with contractors sand at 2 parts
seed: 1 part sand for even dispersal. Seeds were broadcast using jars with
perforated lids onto respective 10.8 ft
Maturation, Harvesting, and Yields .
For each species, data
were collected for flowering date, seed maturation rate, seed number and
weight, dates of optimum seed harvest, and germination percentage. Flowering
date was recorded when the first inflorescence opened. Seed maturation was
determined by observing the rate at which seeds matured over time. Species
seed maturity was based on seed color, seed softness when pinched, and percent
of mature seed structures present compared to immature seed structures present.
The rating scale was 1=85 to 100% of seeds matured at one time, 2=60 to
84% of seeds matured at one time, 3=seeds matured over a period of weeks,
seed heads remained intact until plant senescence, 4=seeds matured and shattered
over a period of 2 to 4 weeks, and 5=seeds matured and shattered over a
period of 1 to 2 months as inflorescence production continued.
The number of seeds produced was determined by observing the amount and
maturation condition of seed-bearing structures present in each plot. Seeds
in a square foot area per plot were collected, cleaned and counted, or sub
sampled and counted, depending on the number of seeds present. Optimum date
of seed harvest was based on maximum amount of mature seed present at a
given harvest time. Seeds were stored for two weeks in paper envelopes at
78F at 58% RH to dry before counting. Total seed yields (lbs) were calculated
based on seed germination and weight of 100 seeds.
Germination Tests .
Seeds were cleaned, counted, and stored
in darkness in paper envelopes at 78F at 58% RH. Undersized and green seeds
were removed. Seed germination tests were conducted four to six weeks after
harvest. Seeds were surface sterilized in a 5% sodium hypochlorate solution
followed by three rinses in deionized water. Seeds were placed on moistened
Whatman's #5 filter paper in petri dishes and sealed with paraffin wrap.
All treatments were exposed to continuous fluorescent light for two weeks
at 73F. Following germination tests, remaining seeds were stored in a sealed
metal can at 45F and 33 % RH.
Due to poor initial results, a second germination test was conducted for
. B. aristosa
seeds were moist-stratified
in damp, whole sphagnum moss and stored in plastic freezer bags inside sealed
plastic containers at 45F and 33% RH for 120 days. Seeds were surface sterilized
then sown in river sand at a 0.25 inch depth. Flats were placed under mist
irrigation for three weeks. Average greenhouse high temperature was 82F
and low was 63F.
Results and Discussion
An arbitrary categorization of species potential for commercial seed production
was established based on seed maturation rating and percent germination.
Germination and seed maturation habits varied from species to species. Species
with the highest germination percentages and most even maturation period
were Hesperis matronalis
, Monarda citriodora
, and Silene
species were considered to have excellent seed production potential with
seed maturation ratings of 1 or 2 and germination percentages greater than
75% and were given a category I rating.
Species placed in category II had moderate seed production potential based
on seed maturation ratings of 1, 2, 3, or 4, and germination percentages
between 45% and 75% (Table 1
These species were Bidens aristosa
, Centaurea cyanus
., Coreopsis tinctoria
, Gypsophila muralis
, Ipomopsis rubra
, and Rudbeckia hirta
. Though seed maturation
rates varied from 1 (85 to 100 % of species mature at one time) to 4 (seeds
mature and shatter over a period of 2 to 4 weeks) in this category, this
was considered acceptable for harvesting purposes.
Species placed in category III had low seed production potential with a
seed maturation rating of 5 and/or germination lower than 45% (Table 1
). Those species were Achillea
L., Coreopsis lanceolata
L., Cosmos bipinnatus
Cav., Cosmos sulphureus
Cav., Eschscholzia californica
Foug., Gaillardia pulchella
Nutall., Papaver rhoeas
L., Phlox drummondii
Hook., Salvia coccinea
L., and Verbena tenuisecta
Though several species in this category exhibited determinate flowering
and had a seed maturation rating of 1 (85 to 100% of seeds mature at one
time), germination percentages were extremely low. These species may require
conditions for breaking dormancy other than those provided in this study,
such as alternating temperatures and a longer germination period. In addition,
species in this category with a maturation rating of 5; the best harvesting
date was difficult to determine due to the extended shattering period.
There was no evidence of a correlation between the seed maturation rating
and germination percentage.
Species which had acceptable germination for commercial seed production
may indicate that seeds were non-dormant at that time and germinated well
in moist conditions at 73F in light. Acceptable germination was set at greater
than 45% to include native species with potential for production.
Seed yield varied from species to species. Species with the lowest yields
were Chrysanthemum leucanthemum
at 31 lbs/A, Verbena tenuisecta
at 42 lbs/A, and Papaver rhoeas
at 55 lbs/A. Species with the highest
yields were Ipomopsis rubra
at 5112 lbs/A, Hesperis matronalis
at 2658 lbs/A, and Monarda citriodora
at 1426 lbs/A. From a seed
production standpoint, high yields are desirable. However, low yield species,
such as Verbena tenuisecta
have potential for seed production because
of high retail prices ($85 per lb.).
Several species were eliminated from the study due to poor establishment
or insufficient flowers and seed development. Species eliminated from the
study were Chrysanthemum maximum
Ramond, Echinacea purpurea
(L.) Moench., Lobularia maritima
(L.) Desv., Oenothera missouriensis
Sims, Ratibida columnaris
(Sims) D. Don, Rudbeckia amplexicaulis
(Vahl) Cass., and Solidago rugosa
In a separate study examining seed cleaning techniques and machine harvesting
of Silene armeria
, Rudbeckia hirta
, and Coreopsis tinctoria
, machine harvesting resulted in significantly lower yields than hand harvesting.
Also, in order to accommodate the variations of seed sizes and weights,
adaptations to existing seed cleaning equiptment or specialized equiptment
Results of this evaluation indicate that several species had maturation
characteristics and germination percentages suitable for commercial seed
production in our southeastern environment. Native species not typically
found in the wildflower seed trade, such as Bidens aristosa
have potential for commercial seed production.
A quality wildflower seed source in the southwest would be advantageous
for Departments of Transportation and other organizations requiring wildflowers
that are well adapted to regional conditions. A regional seed source would
also provide purchasers with southeastern native species less commonly found
in the seed trade and more easily grown in the southeast.
Last Updated 2/1/97