Final report for ONE22-428
Project Information
There have been multiple new biomass varieties of switchgrass and big bluestem, but these new varieties have not been tested in on-farm conditions. By partnering with the Association of Warm Season Grass Producers (AWSGP), we tested seven switchgrass cultivars (Blackwell, RC Chippewa, Flame C3, Independence, Timber, Cedar Creek, and Cave-in-Rock) and one improved big bluestem cultivar (Empire) across three on-farm sites near Huntington Mills, PA. Plots (30 feet by 120 feet) were drilled in spring 2023 (except for Empire big bluestem, which was hand broadcast). The primary measure of success was fall biomass yield which was collected in the 2nd and 3rd year. We also collected mid-summer height and seedling survival during the 2nd and 3rd year post-establishment. Generally, seedling survival was sufficient (>1 plant per linear seeded foot), except for the experimental cultivar Flame C3 which had poor seed quality and the big bluestem cultivar Empire due to uneven seed dispersal due to hand broadcasting.
The plots had slow initial establishment and 2nd year biomass was <2 tons per acre, but 3rd year biomass approached typical production levels (>4.5 tons per acre). To provide a baseline of typical commercial performance, we collected fall biomass yield from two nearby established fields. One field of mixed big bluestem and upland switchgrass averaged 5.5 tons per acre. A fully established field of lowland switchgrass (Timber cultivar) averaged 7.1 tons per acre. Across the three on-farm sites, the best two cultivars were Timber (8.9 tons per acre) and Independence (8.0 tons per acre) while the two worst were Blackwell (2.9 tons per acre) and Flame C3 (2.8 tons per acre). July height and plant density were strongly connected to fall biomass for the Lowland switchgrass cultivars (Flame C3, Timber, Independence and Cedar Creek). This project provides good evidence that winter tolerant Lowland cultivars such as Independence and Timber can produce >40% greater biomass relative to typically used upland switchgrass and big bluestem. Producers also reported that the lowland switchgrass cultivars are less prone to lodging and therefore easier to harvest compared to typical mixes.
We should note that these yield estimates are from hand-harvested fall biomass. Most producers in the region mechanically harvest in spring. Both mechanical harvesting and spring harvesting can result in biomass loss so typical harvest yields may be substantially lower than our hand-harvested yield measurements.
As a secondary project, a AWSGP member seeded switchgrass into a grain sorghum field. Anecdotally, there were challenges during switchgrass seeding and sub-sections of the field were planted too deep, which resulted in patchy establishment. However, when seeded at the correct depth, there was clear evidence of switchgrass survival in the sorghum canopy. Sorghum suppressed the grassy weeds (foxtail, often) which can compete with switchgrass seedlings. Despite the challenges, the partnering producer reported that they would attempt this method again. This secondary study was a simple proof of concept that switchgrass can be established successfully with a cash grain nurse crop.
This project aimed to: (1) Evaluate improved native grass cultivars across three on-farm sites. (2) Measure 2nd and 3rd year seedling counts, 2nd and 3rd year summer sward height and 2nd and 3rd year fall biomass yield. (3) Communicate results to interested parties through online communications, a presentation at the Association of Warm Season Grass Producers annual meeting, and through formal publications.
Widespread adoption of native grass agriculture can diversify the landscape and rural economy. Specifically, native grasses can improve environmental outcomes while creating new markets for sustainable bio-products. Increasing the footprint of perennial crops on the landscape provides greater habitat for a range of insect, mammal, and bird species. Further, deep rooted perennials can increase soil carbon and reduce soil loss and nutrient run-off. For farmers, native grasses can be profitable by providing high-yield, low-input biomass or forage for livestock, particularly when grown on lands unsuitable for annual crops. Promising new incentives for farmers are also being proposed through carbon offsets credits or nutrient credits. In Pennsylvania specifically, nutrient credit programs such as Chesapeake Bay Nutrient Credit Trading program provides a model where farmers are paid for improved land management practices.
Despite these opportunities, perennial native grass adoption has been slow. This is driven by a chicken-and-egg issue where farmers are unwilling to adopt a crop that lacks an established marketplace. Further, without a robust community of farmers testing new methods and varieties, the best practices in native grass agronomics are uncertain. The Association of Warm Season Grass Producers is seeking to remedy both of these issues by collaborating to recruit buyers of grass biomass and to share insights among farmers regarding best practices. A recent opportunity has arisen because multiple member farmers have reported fields with low yields due to a combination of older varieties, weed build-up and disease pressure. These fields were established using a “first generation” upland switchgrass cultivar Cave-In-Rock, along with a mixture of big bluestem and other grassland species. Improved varieties could provide greater yield and disease resistance, but many new varieties have not been tested in the region. Establishment and yield data on improved native grass varieties can provide current farmers with valuable information and could make native grass production more appealing to prospective farmers.
Native grass establishment is currently risky due to the risk of failed or weak grass stands. Native grasses establish slowly and produce negligible biomass during the first year and have reduced yields during 2nd and 3rd season. Breeding programs have focused on improving native grass germination rates and seedling vigor. The strongest example of this is the program REAP Canada (Resource Efficient Agricultural Production) which has adopted a single-year breeding cycle based on vigorous upright growth during the establishment year. This program has developed multiple cultivars including RC Chippewa, a selection from the Cave-In-Rock.
A second major source of improved varieties is through the improvement of lowland switchgrass ecotypes. Broadly, switchgrass occurs as two primary ecotypes: upland and lowland. Lowland ecotypes tend to grow in the southern United States, while upland ecotypes are native to the northern United States. The majority of farmers in the northeast region use an upland ecotype. Agronomically, lowland ecotypes have greater yield potential and disease resistance. However, lowland ecotypes are prone to winterkill and poor vigor when grown in the northern United States. Recent breeding efforts, however, have created multiple winter-tolerant lowland cultivars. These lowland ecotypes can increase biomass yield by up to 50%, but realistic on-farm comparisons have not occurred. Further, different breeding programs have released multiple lowland populations which have not been compared in the Northeast. This study proposed to evaluate the improved lowland switchgrass cultivars Timber, Independence, and Cedar Creek. Timber is a cultivar generated from multiple populations native to the Northeast and was released by the Cape May Plant Materials Center. Independence is a release from the University of Illinois and was created from individuals selected for high biomass and vigor from the cultivar Kanlow. Cedar Creek was created in Wisconsin from a Kanlow-derived breeding population which was selected for multiple cycles of winter survival as far north as plant hardiness zone 4. An experimental population called Flame C3 was also developed from a diverse mix of southern populations selected for survival in Wisconsin.
Although more research effort has been dedicated to switchgrass, multiple breeding programs have also focused on improved big bluestem cultivars. Big bluestem has been successfully planted with switchgrass by farmers in the region and improves yield stability across growing seasons and fields. This study will evaluate the new cultivar Empire big bluestem which was selected for height, vigor, and late flowering.
This study aims to provide important cultivar performance data for farmers and may also provide information to reduce the risks of sward establishment. Combined, this will improve confidence of farmers in the Association of Warm Season Grass Producers and can also encourage other farmers with marginal or erodible land to establish native grasses.
Cooperators
- - Producer
- - Technical Advisor (Researcher)
- - Technical Advisor
- (Researcher)
- - Producer
- - Producer
Research
In late May of 2023, the original sward was renovated using standard methods. Briefly, the swards were mowed and sprayed with glyphosate at the suggested rate prior to re-seeding. After seeding, the field was cultipacked to improve seed to soil contact. The six varieties were planted using a John Deere 750 no-till grass seed drill at a rate equivalent to 4 lbs PLS per acre. Seed for most varieties were donated by Ernst Seeds of Meadville, PA and two varieties were donated by active researchers. Due to limited seed in four cultivars, the experimental design was revised to be an unbalanced incomplete block design. Four cultivars were planted across two replicates per site (Independence, Flame C3, RC Chippewa, and Timber) and four cultivars were planted across one replicate per site (Blackwell, Cave-In-Rock, Cedar Creek, and Empire). Plot size was 30’ by 120’. First and second year weed control included of 2,4-D at the suggested rate as needed to control of broad-leaf weeds, and summer mowing to reduce weed competition.
Establishment success was measured by the following: In July of 2024 and 2025, mean fall sward height and seedling number (per linear seeded length) was recorded at four quadrats randomly placed along of a transect running through the center of each plot. Biomass accumulation was measured post-frost during the winter of 2024-2025 and 2025-2026. Fall biomass yield was hand clipped at standard mowing height (4”) and from four quadrats randomly placed along the center of each plot (2.25 square feet each, nine square feet total). Forage samples were stored in a cool dry area (68 degrees F and 40% relative humidity) and air dried to a constant weight (3 weeks) prior to recording the total weight per plot. A subsample was placed in a forced air drying oven at 100 degrees F for 24 hours to determine residual dry weight (dry weight 89% in 2024-2025 and 80% in 2025-2026). Samples were weighed using a hanging scale with a precision of 0.05 lb. The percent of switchgrass plants with exposed seed heads (percent heading) was also recorded in July of 2025.
Analysis used a fixed-effects model estimating the effects of years, grass varieties, sites and blocks on July plant height, July seedling density, and winter biomass yield. A random variety-by-farm interactive effect model was also used to estimate the yield stability of different cultivars among farms. Further, plant height and seedling density were regressed against biomass yield to determine relationships between these easy measures and fall biomass performance.
This research project officially began August 1, 2022 and was completed in the winter of 2025. A pre-planting herbicide WSG Variety Trial Spray plan and plot map was developed in the fall of 2022 by the researchers and participating producers.
One of the three original warm season grass producers committed to the project declined to participate in the project for personal reasons. However, another agricultural landowner was added as a third farm. This turnover resulted in all three farm sites occurring within one mile of each other. However, the soil conditions and weed pressure varied between sites. All sites were in an upland (elevation of 1,100 feet above sea level) area with rocky (channery) soil. Soil samples were collected and sent to the University of Massachusetts: Amherst Extension Soil and Plant Nutrient Testing Laboratory for analysis and all soil variables were non-limiting with the exception of slightly acidic soil in Site 3.
|
Variable |
Site 1 |
Site 2 |
Site 3 |
|
Soil Type |
Mardin channery silt loam |
Oquaga channery silt loam |
Oquaga channery silt loam |
|
Weed pressure |
Heavy |
Moderate |
Minimal |
|
Slope |
3-8% |
8-15% |
15-25% |
|
Depth to bedrock |
>60" |
20-40" |
<20” |
|
pH |
6.3 |
6.6 |
6 |
|
Cation Exchange Capacity (meq/100g) |
13.6 |
12.9 |
15.4 |
|
Phosphorus (ppm) |
14.7 |
22.9 |
15 |
|
Potassium (ppm) |
171 |
177 |
207 |
|
Calcium (ppm) |
1615 |
1638 |
1301 |
|
Magnesium (ppm) |
220 |
140 |
231 |
|
Sulfur (ppm) |
12.8 |
11.6 |
9.8 |
|
Boron (ppm) |
0.2 |
0.2 |
0.1 |
|
Manganese (ppm) |
13.6 |
6.1 |
9.5 |
|
Zinc (ppm) |
1.2 |
1.4 |
2.3 |
|
Copper (ppm) |
0.2 |
0.5 |
0.3 |
|
Iron (ppm) |
2.5 |
1.3 |
1.6 |
|
Aluminum (ppm) |
19 |
24 |
35 |
We were unable to source all of the original entries proposed for the study. Specifically, we had to remove RC Crazy Horse big bluestem and RC Big Rock switchgrass because these were unavailable at planting time. RC Big Rock was replaced with RC Chippewa, which is a switchgrass population selected for the same traits with slightly earlier flowering time. RC Crazy Horse was replaced with an experimental cold tolerant lowland ecotype switchgrass population called FLAME C3. Also, seed was limited for both Empire big bluestem and Cedar Creek switchgrass. Therefore, the experimental design was altered to be an incomplete block design where Cedar Creek and Empire were replicated only once per site. The remaining plots were replaced with a single replicate per site of two well established upland switchgrass cultivars to serve as 'check' varieties: Cave-In-Rock and Blackwell. This resulted in four entries containing two replicates per site (RC Chippewa, Independence, Timber, and FLAME-C3) and four entries with one replicate per site (Cedar Creek, Empire, Blackwell, Cave-In-Rock).
The planting occurred in spring of 2023 using a John Deere 8300 seed drill at a rate of 4 lbs. per acre. Switchgrass establishment appeared to be successful, but there was strong first year weed pressure. Big bluestem was broadcast seed and establishment was uneven across the plot. Because of this, big bluestem measurements were omitted from seedling density measures. Flame C3 seed, which contained a high proportion of chaff and debris, had poor establishment and clogged in the seed drill. Weed pressure, particularly yellow foxtail, was aggressive in year 2 (2024) and an additional 2024 herbicide application was required.
Summer height measurements and seeding counts were collected for 3.2 linear feet along a planted row at four points at each plot on July 8th, 2024. On December 12th, 2024, fall biomass was collected at four random 2.25 square meter quadrats in each plot. In 2025, height and seedling counts were collected on July 2nd 2025 and fall biomass was collected on October 28th. During both fall biomass collection dates, biomass samples were collected (four quadrats from random points) from two established production fields: one of mixed big bluestem and upland switchgrass and a second of established Timber lowland switchgrass. The average yields of these fields will be used as benchmarks to estimate variety performance. Specifically, the mixed big bluestem and upland switchgrass field produced 3.27 tons per acre in 2024-2025 and 5.23 tons per acre in 2025-2026. The Timber lowland switchgrass field produced 5.55 tons per acre in 2024-2025 and 7.14 tons per acre in 2025-2026.
A fixed effect model indicated that grass variety had highly significant effects on seedling establishment success (p<0.001). The mean number of seedlings per linear meter was 3.69, which is less than ideal but sufficient for a successful stand. The seedling counts per 3.2 feet were different among test fields (Site 1: 3.77, Site 2: 4.47, Site 3 2.17), with lower germination in the marginal site with shallow soils. Among varieties, Cave-In-Rock, Blackwell, and Cedar Creek had the greatest seedling number (>6) and Flame C3 had the lowest seedling number (<2). Generally, the upland switchgrass ecotypes had greater seedling establishment success relative to the lowland switchgrass ecotypes (4.75 seedlings per 3.2 feet vs 2.74).
Plant vigor was also measured through July height measurements. There was a significant difference (p<0.001) between grass varieties. Similar to seedling counts, the marginal Field 3 had reduced plant heights (Site 1: 45.0 in, Site 2: 42.8 in, Site 3: 38.21 in). The highest mean seedling heights (>35 inches) were Cave-In-Rock and Independence. The similar height of the upland cultivars is deceptive in mid-summer since upland cultivars are initiating flowering and do not continue to gain height past late July. By contrast, the late flowering lowland switchgrass typically continue to accumulate biomass until fall.
Big bluestem was omitted from July height analysis because the Empire plots had not initiated stem growth. The lack of stem development in early July was an interesting observation because Empire had produced a relatively large quantity of leaf material which is a promising trait for producing livestock forage. Un-replicated forage samples of typical big bluestem (which had produced extensive stem material), Empire big bluestem, timothy, and orchard grass were collected and sent to the University of Ohio Sustainable Agroecosystems Laboratory for Near Infrared Spectroscopy determination of forage quality. This testing indicated that Empire big bluestem maintained a relative forage quality (RFQ) of 114.7 relative to 79.7 for typical early July big bluestem (crude protein 10.8% vs 9.0%; acid detergent fiber 37.2% vs 44.4%; neutral detergent fiber 62.0% vs 73.0%). The Empire RFQ values were also close to samples of unharvested July timothy (115.6) and orchardgrass (71.4) in nearby fields. Although Empire was developed for biomass production, this delayed stem development could make it a useful forage cultivar and future research could confirm this observation.
The native grass swards were not fully established in the 2nd year. Mean fall biomass harvest for the trial plots in the 2nd year was 1.19 tons per acre. This is poor relative to the 2024-2025 yields of 3.27 dry tons per acre and 5.57 dry tons per acre in nearby established fields. In the 3rd year post establishment (2025-2026), there was a significant (p=0.011) variety effect, with Blackwell and Flame C3 producing the least biomass and Timber and Independence producing the most biomass. There was a large difference in biomass between the sites (Site 1: 5.45 tons per acre, Site 2: 5.30 tons per acre, Site 3: 2.88 tons per acre). In the figure below, the vertical orange line represents the yield of an established production field of big bluestem and upland switchgrass. The vertical green line represents the yield of an established Timber lowland switchgrass field.
Separately, a secondary experiment was tested switchgrass establishment underneath an annual crop of grain sorghum. There were challenges during switchgrass seeding and sub-sections of the field were planted too deep, which impacted stand success. However, there was visual evidence of switchgrass seedling establishment within the sorghum canopy, which is encouraging since switchgrass is often unidentifiable during 1st year establishment. The producer reported 46 bu./ac harvest for sorghum, but attributed the low yield due to drought. There were small control plots which were not seeded with a sorghum nurse crop and these were dominated by warm season grass weeds. The producer reported that they would repeat this strategy for establishing new native grass swards. Given the challenges establishing native grasses in the primary study, this is a promising area for future research.
In this study, we aimed to determine which switchgrass and big bluestem varieties would produce the most biomass in on-farm conditions. The two high performing lowland cultivars (Timber and Independence) produced an average of 40% more biomass relative to the next best upland switchgrass cultivars (RC Chippewa and Cave-In-Rock). This difference is similar to the difference observed between two production fields which used an upland switchgrass and the lowland switchgrass cultivar Timber (Timber +60% greater in 2024-2025 and Timber +30% greater in 2025-2026). This provides good evidence that robust lowland cultivars can provide a significant yield benefit for Pennsylvania farmers. However, long term yield measurements could provide more reliable evidence about the profitability of these cultivars.
Education & outreach activities and participation summary
Participation summary:
Initial discussions among the Association of Warm Season Grass Producers, the grantee principal investigator (Wes Ramsey), and Neal Tilhou (the project researcher) defined the intent of the project and prepared a pre-planting herbicide application plan in cooperation with the three producers identified as participants in the project.
The results of this study was communicated to the public through multiple channels. Preliminary results were presented at the Association of Warm Season Grass Producers meeting in 2024, and 2025 ,final results will be reported at the 2026 meetings. Final results will also be posted to the Association website https://www.awsgp.org/ and e-mail list consisting of over 300 members and the results will also be published on the Penn Soil Resource Conservation & Development (RC&D) Council website www.pennsoil.org and social media pages. The Association of Warm Season Grass Producers also regularly presents at Ag Progress Days and can include information about the results of this trial. Also, results will also be used to draft a peer-reviewed paper which would include entirely novel yield data for these grass varieties and this region.
In early spring of 2024, discussions took place between the research participants about additional research concepts that could be incorporated into this project due to some cost-saving measures that occurred within the first year of the project. Due to continued interest among producers in the Association of Warm Season Grass producers, we supported an additional on-farm experiment. These experiments could provide additional specific guidance for more affordable native grass establishment. This second experiment tested switchgrass establishment underneath cash crops, either for hay or grain. There are very few reports of this method and the strategy is, justifiably, considered risky since switchgrass seed is expensive. Tyson Myers was planning to establish native grasses after a grain sorghum crop. After discussion, he established a switchgrass stand underneath a grain sorghum crop, with both planted in spring 2024. He seeded three promising and genetically distinct switchgrass varieties: Timber, Liberty, and Independence. Tyson Myers provided a preliminary planting report with images of the pre-treatment process and planting report and images of the first season growth which are added into the information products section of this report.
Learning Outcomes
We have not carried out outreach beyond the participating producers. Bruce Trumbower reported that 2024 was a rainy year and foxtail grass was aggressive and required two weed and grass control sprayings in May and June and one maintenance mowing in March on the three plots. Plot 1 grew the best, Plot two was second best, and plot 3 was third best by his observation. Timber switchgrass seemed the strongest in his opinion.
In the secondary experiment attempting to establish a switchgrass stand underneath and following a sorghum nurse crop, Tyson Myers reported that "My feelings are that the switchgrass seeded into the sorghum did and is growing successfully. One thing I noticed was that when I seeded the switchgrass I did not have the drill settings correct and actually planted it too deep. So the spottiness across the field may be due to my error. There are definitely areas where switchgrass is taking off in the field, and it did not come from any other source than our "intercropping". In the future if I were trying to establish a stand of warm season grasses I would try this again with my own funds if I wanted to expedite the establishment."
Project Outcomes
During the course of this grant, the primary participating farmer recruited two neighboring younger farmers to lead his grass enterprise. These neighboring farmers have continued to expand perennial grass acreage using the improved cultivars based on these results. They have reported that Timber has improved yields and is easier to harvest in spring (less lodging).
We do not have formal evidence of changes in AWSGP farmer practices due to our results only because the final data was collected 3 months before the end of this grant. Therefore, we haven't had time to communicate the results and receive feedback from farmers.
Our secondary project which established switchgrass under grain sorghum was considered a success for the participating farmer and they expect to use this strategy in the future.
This experiment focused on using large experimental plots in on-farm conditions. Therefore, the results are are highly representative of 'real world' farm conditions. However, the small number of replicates (1-2 per site) and high heterogeneity between sites means that our ability to statistically differentiate between cultures was poor. We believe these tradeoffs were justified and our results should be persuasive to farmers.
Unfortunately, we had establishment issues for the Flame C3 population and the Empire big bluestem population. These varieties, therefore, will have to be evaluated in other conditions to get reliable information about performance. In particular, Empire should be compared to other big bluestem cultivars or in mixture with switchgrass since fields are often seeded as mixtures.