The tools of the seed collector include gloves, a good
hand clippers, a large supply of grocery bags of various
sizes, and a collection of plastic buckets. Buckets
can be purchased at building supply stores or hardware
stores. Also, cast-off buckets from grocery stores or
restaurants may be available gratis. To keep the hands
free for collecting, a rope or sturdy strap should be
used. An alternate is to fasten the bucket to a belt
loop with a caribiner.
The seed collecting technique varies with the species.
A few tips are given here.
• For species with large seed heads such as those in
the sunflower family: gather the stalks of the seed
head together and cut as close to top as possible, and
then drop into the bucket. As the bucket fills up, press
the seed heads to the bottom. When the bucket is jammed
full, transfer the collected material to a large bag
and return to the collecting site.
• For species making structures with loose seeds, such
as shooting star or prairie cinquefoil: cut the flower
stalk below the lowest seed container, carefully place
the stalk into the bucket and then turn upside down
so that the seeds drop in the bucket. Knock the stalk
against the bucket to release more seeds, and then discard
• For species whose seeds strip well (grasses, some
asters and goldenrods: pull the seed head over the bucket
while still attached to the plant, and then strip the
seeds directly into the bucket.
• For species that make berries: pull the seed head
over the bucket and strip off the berries.
• For milkweeds or other species that make seed pods:
cut each pod just below the stalk and drop into the
bucket. Later, remove the seeds and let the “fluff”
fly away (a windy day helps).
• For carrot family species: gather the seed head together
over the bucket and cut the stalks above where they
join, so that each umbel breaks apart.
rigida) flower head
goldenrod seed head
collecting technique. The bucket is fastened to
the belt, keeping its position low and allowing
the hands to be free. The seed head is pulled
over the top of the bucket and then cut.
right: close up photograph of seeds, showing the
characteristic "fluff" that most members
of the Aster family possess.
drying and cleaning
Seeds should be well dried, either in the sun or in
a low-humidity environment. A basement or storage room
with a well-ventilated gas heater is ideal. The seeds
should be spread out on tarps, sleds, or specially built
trays. If space is available, they can be spread out
on a concrete floor.
The seeds should be turned occasionally, and allowed
to stand until they are all dry. Even if the seeds seemed
dry in the field, they will have enough moisture so
that they will need drying. Feel the seeds with the
bare hand to determine if they are dry. If the seeds
are not dry, they will become moldy. Seeds are rarely
released in pure form from plants. Most species have
seed coats, or come off the plant attached to “fluff”
that is involved in seed dispersal. Some species bear
their seeds in fruits. Each species has its own peculiarities,
although asters, goldenrods, and other members of the
Asteraceae all produce seeds of similar character. Purification
involves extensive manipulation by breaking up the crude
seed heads, screening, sifting, and (perhaps) blowing
chaff away in a fanning mill. In most cases when planting
a savanna restoration, highly pure seeds are not needed.
Once the seeds are dry, they should be cleaned. The
goal here is to remove all stems, seed coats, and other
chaff. Unless they are to be used for a special research
study, the seeds do not need to be highly purified,
but large sticks and debris should be removed. How clean
they should be will depend on the manner by which they
will be planted. If a mechanical planter is to be used,
then the seeds need to be cleaner than if they are to
be hand broadcasted.
trays for drying seeds. The rack in the background
is on wheels so that it can be pushed out of the
rain. The hammermill is used to release seeds
from their pods or seed heads.
screening compass plant seeds.
The hammermill is very useful for processing seeds.
It breaks up the seed heads and releases the seeds.
Once released, the seeds can be further purified
by screening. Although expensive, a hammermill is
essential for any substantial restoration project.
Nonprofit organizations can often combine resources
and share such a piece of equipment. Individuals
may find it possible to trade volunteer labor at
a larger restoration project for use of the hammermill.
of milkweed pods
Once cleaned, the seeds should
be stored in brown paper grocery bags. Plastic bags
are not recommended because they will not allow moisture
to escape. Be sure to label each species with name,
as well as date and location collected.
Species vary widely in the size and shape of their seeds
and the restorationist should become familiar with the
structures of all species being used.
storage until planting time, seeds should be stored
in clean, heavy-duty bags. These bags can be purchased
at any agricultural feed mill. If they are to be
stored for the winter, they should be hung up so
that rodents cannot reach them.
Commercial seed sources
Commercial seed companies specializing in prairie and
savanna restoration may also be a source of local genotypes.
Good companies will know the sources of the seed stock
that was used to establish their seed production beds,
and will specify that information on the label. Restorationists
should obtain advice about good seed companies in their
area from experienced local people.
Seeds can be planted either in the spring or fall. Fall
planting is ideal, since the seeds then overwinter under
conditions that mimic those in nature. However, it is
essential that planted seeds reach bare ground, which
usually means after a burn. Although fall burning is often
excellent, it depends upon favorable weather conditions,
and may be thwarted by an early winter.
If planting must be delayed until spring, hang seeds in
an outdoors protected location such as a barn, screened
in porch, or unheated basement. For large-scale storage,
commercial feed bags, available at an agricultural co-op,
are recommended. Wherever they are stored, they must be
protected from rodents.
As soon as spring burns are finished, the seeds should
live seed and seeding rates
Quality of seed plays a major role in the success of
a planting operation. Seed viability can vary markedly
from year to year, depending on the weather during the
period of flowering and seed setting. Seeds purchased
from commercial sources have the advantage here, because
they are sold on a Pure Live Seed (PLS) basis.
Purchased seed has two major drawbacks: it is expensive;
and it may not be local genotype.
The variable quality of seed is one reason that restoration
sites should be planted more than once. Each year is
different, and quality of seed varies, so some species
may become established one year and not another.
Determining seed viability is not complicated but requires
technical knowledge which the seed testing laboratory
provides. In the Midwest, there are commercial seed
testing laboratories as well as testing laboratories
associated with agriculture universities.
Suppose the percent viability of the seeds is not known?
The Natural Resources Conservation Service (NRCS) specifies
practices which may be applied as part of conservation
management activities and are permitted for landowners
supported in part by Federal grants. These practices
are also applicable even if government funds are not
For instance, if the seed used was hand collected from
a local source, the NRCS specifies a minimum seeding
rate of 50 seeds per square foot. An additional provision
is that the seeding rate should be adjusted so that
no single species amounts to more than 20% of the seeds
in each square foot. NRCS also specifies that at least
25 seeds per square foot must be native grasses or sedges
and at least 5 species of grasses and 15 species of
forbs and legumes must be seeded. These criteria are
applied by NRCS for both prairie and oak savanna restorations.
Note that the number of species given in the previous
paragraph is a minimum and great effort should be expended
to plant more species. If only small amounts of hard-to-get
species are available, then they should not be placed
in the general seed mix, but hand planted separately
in suitable parts of the savanna.
Seeding by weight or volume
instead of numbers
Many restoration plantings use weight of seeds instead
of number. For instance: “Use about 10 to 15 pounds
of mixed forbs and grasses per acre using at least twice
the weight of forbs compared to that of grass seed.
Use a 10-pound rate for certified seed and a 15-pound
rate for hand-collected seed.”
Whether weight or numbers, the more seeds the better.
Some people use seeding rates of 25 or even 50 pounds
of seed per acre. Research has shown that the higher
the seeding rate, the higher the species diversity in
the final planting.
Packard also suggests seeding
rates based on volume of seed. For instance, one
cup (16 ounces) per 100 square feet or “one kitchen
garbage bag to 18,000 square feet (a bit less than half
an acre). Such a seed mix would consist of 50 percent
seed (with the heads broken up but chaff retained) and
50 percent perlite.” Instead of perlite (an insulating
product available at lumber yards) sawdust can be used
as a seed carrier.
Seed mixes for savanna restoration
Because of its heterogeneity, the savanna being planted
never provides a single type of habitat. It is convenient
to think of the planting unit as a mixture of four types
of areas: prairie (0-10% canopy), open savanna (from
10-30% canopy), closed savanna (from 30-70% canopy),
and woodland (greater than 70% canopy). Each type of
area should be marked and an appropriate seed mix used
for each one.
the planting, walk the area with the head to the sky
and locate areas with very few trees (prairie), scattered
trees (open savanna), woodland areas with occasional
openings (closed savanna), and woods without any openings
(woodland). If a recent (post-restoration) air photo
is available, it can be used to locate these various
An oak savanna
species list that categorizes principal native species
according to their canopy preferences is provided here
as a PDF.
The Pruka and
lists provide brief overviews, and the Bader
list provides much greater detail.
These lists are just representative of what can be used.
A lot will depend upon the region of the Midwest, and
the species for which seeds are available.
The quantity of seed planted will depend in the first
instance on the amount of seed that has been collected,
or purchased. Again, recall the dictum that the higher
seeding rate and the more species the better.
stages of making a seed mix. The various species
are measured out and added to the mix.
the seeds. Sawdust has been added as a carrier.
This especially helps to keep the very fine seeds
from settling out. It is important that the seeds
be well mixed.
seed mix is distributed to buckets to carry
to the field.
can either be hand-broadcasted (as is being done
here), or planted with a mechanical tractor-pulled
device such as a fertilizer spreader or drill
planter. For hand broadcasting, a light snow is
very helpful (but not essential).
even distribution of seeds is shown in this pattern
on the snow. Most of the material visible is sawdust,
but by getting down on the snow it is possible to
see single scattered seeds.
a large planting acreage, mechanical equipment might
be used. This rig uses a fertilizer spreader that
has been adapted to use for seeds. A device in the
bottom of the hopper keeps the seed mix stirred
while it is being distributed.
Calculating seeds per square
The recommended planting rate by the NRCS is 50 seeds
per square foot. With crude or partially cleaned seed,
how does one estimate the amount to use?
Here is a procedure adapted from Kurtz. It is for the
amount of seeds per unit weight of a single species,
and assumes the species has been partially cleaned.
1. Select a representative sample of the partially cleaned
2. Remove a small amount and separate the seed from
remaining straw, flower parts, fluff, etc., generating
two samples for each species, seed and debris.
3. Weigh the cleaned seed and the debris separately
on a sensitive scale. The weight unit used can be grams
or grains, or whatever the scale reads.
4. Divide the seed weight by the sum of the seed weight
and the debris weight and multiply by 100 to give percent
of the original sample that is seed.
5. Thus: with 15 grams of seeds plus 80 grams of debris,
the percent seeds is 15.8%: 15/(15+80) * 100
6. Record this percentage on the species list and go
on to the next species.
7. When making the final seed mix, adjust the amount
of each species to account for its purity.
8. With a little experience, it may be possible to “estimate”
the purity by simple visual inspections.
High accuracy is not needed in this work.