Wild Flowers on Marginal Lands

Our farm is about 151 acres in size , half of which is wooded . This was a dairy farm for many decades, but since our purchase my wife Grace Tinderholt and I have been examining wildflower seed production and other horticultural endeavors as alternatives. Wildflower seed markets are very strong right now, and we have been working on methods of establishment and harvest for several years. The flower patches have been expanding as has the income we receive from them, and we are ready to add several more acres in 1996. Ours is an entirely family run operationand we are currently expanding our wholesale marketing.In all our farm endeavors we tried to shy away from expensive chemical inputs, occasionally using Roundup and chemical fertilizers.

Goals :
We undertook this project primarily to expand farming into our hilly pastures without washing them away.This project was started to examine the establishment success of wildflower plugs in the uplands that were treated with four different pre-treatments or no treatment.
Treated strips were to be sprayed with Roundup
and either left alone or or subsequently rototilled, rototilled without chemical pre-treatment, or burned with a flame-weeder torch(an addition since the proposal).Comparisons of plug survival were also to be made between spot sprayed and or untreated spots ion wetlands and along fencelines, and on burned plots in the wetlands.
Wildflower plugs were chosen due to
a. the limited supply of stock seeds available for most wildflower species and Wildflower seeds from local ecotypes were chosen because of the higher prices they fetch in the market.
b. because they produce greater amounts of viable seedlings per ounce of seeds, and
c. because plugs are more easily transplanted than bare root seedlings and
d. because plugs have a higher survival rate.

Likely producers may gather these seeds from local highway banks, railroad right of ways,etc, usually a few.ounces will do. Past experience with wild lupine showed that one ounce of seed could produce about 50-100 plantsif seeded directly and 500-800 seedlings if grown in Flats first. A wide variety of seeds were chosen for dry, mesic and wet locations , including some rare and thus expensive varieties (upto $50 an ounce) and other more common varieties.

Flats from silvicultural trade were chosen since they are much more deeper(3" or more), and two varieties of flats - styrofoam and polyethylene (both small and large plug sizes)- were used to see if the longer lasting latter variety had any advantages. Flats with and without legs was also tried.

Use of peta, vermiculite and perlite as a grwing medium did away with the need for sterilization.The seeds were stratified, hand planted and chemically fertilized; legumne seeds were innoculated. On average 15-30 minutes per flat of 240 cells was required to hand mix soil and and to fill and plant the flat. Broadcast seeding was fast but wasteful.Flats were grown both incold frames and in the open depending on the weather.
Some species couldn't be grwon for several reasons:

1. Butterfly weedand Pale Puyrple Cone FlowerRot -> due to Rot during stratification.
2. Prairies Larkspur, Milkweed --> due to no germination
3. Large Flowered penstemon, Napea dioica,Wild PetuniaHoary Vervain--> Low germination.
4. Grass Leaved Penstemon, Golden Alexander, Heart leaved Golden Alexander--> due to slow growth
5. Obedient Plant, Canada Milkvetch--> ran out of time, trial in 1996.

Nine species were grown:
1. Anise Hyssop(Agastache foeniculum)
2. Blue Vervain(Verbena hastata)
3. Narrow Leaved Coneflower(Ecinacea Augustifolium)
4. Prairie Coreopsis(Coreopsis palmata)
5. Rattlesnake Master(Eryngium yuccifolium)
6. Rosin weed (Silphium integrifolium)
7. Swamp Milkweed (Asclepias incarnata)
8. Wild Lupine(Lupinas Perennis)
9. Wild Quinine (Partheniumintegrifolium)

A late spring and a hot drought in June delayed work at the Wisconsin and Iowa sites.Instaed of plugs dormant transplants were planted in mid-Fall using the following species:
1. Compass Plants (Silphium laciniatum)
2. Large flowered Penstemon(Penstemon grandiflorus)
3. Prairies coreopsis(Coreopsis palmata)
4. Rattlesnake Master(Eryngium yuccifolium)

Due to the late planting the experiments were changed to a comparison of spring mowed versus burned plots for Fall establishments which will be assessed in 1996.Cooperators were were behind schedule for making hay for these same reasons by the time planting occurred in late June-July.
Planting for upland plants was done on two hillsides and one fence line on our farm, and the wetland plants( Swamp milkweed and Blue Vervain) were planted in one wetland area here and one in a neighbouring place. The fenceline was planted in sprayed and unsprayed plots with Rosen Weed. Due to the late planting the survival will be assessed in Spring 1996.
After laying out blocks in wertland and upland areas , the rows to be sprayed were chosen at random and hand sprayed to wet witha 3% Roundup solution in early June. After twop weeks the upland plots were mowed(4 weeks foir the wetland area to a height of 4 " to facilitate planting and tilling) and tilling and burning commenced.
Tree planting spuds were used to plant the upland plants in late June and the same occurred for wetland plants in mid -July.
Plants were counted in mid-September to determine survival.Narrow Leaved Coneflower and Rattlesnake masterwere furthewr evaluated in early November to determine leaf counts and maximum leaf width. in order to determine growth differences between treatments. After controlled burns in 1996 all species will be assessed for long term survival and flowering/seed production.

People : Besides Grace Tinderholtand myself, Don and Vera Magnuson took part by allowing us to plant in a wetland. Arlene Fitzsimmons babysat plants and planted some in a garden. Dennis Kutka helped count plantsand provided upland plots. Camp EWALU in North-east Iowa provioded upland plots and three people , John Schwoch, Tiffany Paul;son, Aaron Torgerson, carried out much of the field labor of tilling and planting.

Results
In spite of the late start the survival rates were good. Kruskal Wallis ANOVA was used in analysing the data to determine differences in treatments since the raw data was non-normal. For most species there were no differences in survival among the treatments used was better than fire. There were noblock effects in the upland areas, but differences among blocks was significant for the wetland plains. The blocks in more stiff reed canary grass had many fewer plants andmost were undergoing active grazing by slugs.Though the average survival for wetland plants was lower than the upland plants because of this grazing,survivalin the non-grazed plots was comparable to the uplands,as detailed below:

Species Mean Median
Survival Survival
1. Swamp
Milkweed 66 72
2. Blue Vervain 73 72
3. Perennial Lupine 81 80
4. Narrow Leaved
Coneflower 92 95
5. Wild Quinine 95 100
6. Rattlesnake
Master 95 97
7.Prairie Coreopsis 86 83

No treatment effects upon survival could be seen for the Narrow Leaved Coneflower or Rattlesnake Master,but this was not true for growth assessements. In both casesthere wassignificant differences among treatments. Simply mowing the grass did not control competition and the most intensive treatment, sprayiong and tilling always had the highest valuesfor growth for each species. For the Coneflower this treatment allowed for significantly better growth than did theintermediate grass control of simply sparying ortilling alone. ForRattle snake Master, spraying spraying and tilling gave comparable results to the compounded treatments.

The relative ineffectiveness of fire for withholding grass growth in mid and late summer was unexpected as was the impact of grass competition even for grass 9" tall. As we continue to work with plug reared seedlings.
we willmonitor the long term effects of the of the pre-treatment grass control and attempt tocompare in a replicated the relative benefitsof an organic approach (burning followed by mowing to 4-6" , and spraying with Roundup.

Conclusion
1. Plug reared wildflower seedlings are fairly easily established, even under rigorous conditions.
2. Competition for light during establishment is more important than previously envisioned.
3. The best method tested in 1995 was a single pass with Roundup prior to mowing and planting.
4. Tilling was hard and gas and time intensive.
5. Burned rows were the easiest to plant sincethey were thecleanest beds.
6. A single pass produced excellent grass control.
7. Maybe hybriids might work best.
Thus planting methods that minimize soil erosion are readily available.
Analysis after twomoreseasons would probably yield insights into economic returns associated with the different techniques.

Recent bulletins put out by the NRCS Plant Materials Center in Bismark showed that they are using plugs to rear seed producing wildflowers and that machines are usefulin these establishments. Also the existing cover and additional seed producing species certainly provided wildlife cover and minimal erosion. We anticipate pretty if not newly valued marginal land on our fam and other surrounding farms 1996.

Outreach
We contacted local radio and t.v. stations, Universities ,Newspapers, and Community Bulletin Boards but had only five visitors on our field day. Results being pending at that stage only the visible trends and survival data were discussed. In December I delivered a Talk at the National Resources Research Institute on this issue. A brief summary of the results will be presented at the second field day to be held in Fall. Subsequently I plan to publish a short paper on this topic.