By Jonathan O. C. Kubesch*,**, Frank Reith*, Dillon P. Golding*,***, Jake Sanne*, Forrest Brown*, Derek Hilfiker*, Joseph D. House****, Jenna Beville*, and Peter Arnold*,*****
*Virginia Tech School of Plant and Environmental Sciences, Blacksburg, VA
**Country Home Farms, Pembroke, VA
***Hoot Owl Hollow Farm, Woodlawn, VA
****Indiana National Guard, West Lafayette, IN
*****Arnold Classic Farms, Chestertown, MD
The public is familiar with red (Trifolium pratense) and white clover (Trifolium repens) growing throughout the Kentucky Commonwealth. However, North America, from Oregon to Florida, is home to a plethora of native clover species. Buffalo clover (Trifolium reflexum) is one of several clover species native to the eastern U.S.A. (Kubesch et al., 2022; Kubesch, 2020). This species demonstrates annual to short-lived perennial life histories, and has potential as a horticultural or agronomic crop (Quesenberry et al., 2003; Kubesch, 2020).
Current efforts to increase native clover populations involve laudable efforts regarding site management, as well as conservation horticulture (e.g Littlefield, 2022). After a site is prepared for planting, plugs are produced. Conservation horticulture work currently executes the following procedure:
- Germinate seeds on filter paper in petri dishes (Figure 1)
- Transfer seedlings to cell pack trays
- Pot up plants into small pots (Figure 2)
- Plug individuals into spaced nurseries or maintain on benches for seed production
In restoration and agronomic contexts, seeding clover has a logistic and resource advantage over plugging clovers. Seeding clover can reduce the need for intensive planting efforts, reduce soil disturbance, and ease transportation of unique plant material. Seeding approaches require a basis for setting a seeding rate and dates. Often, clovers are timed for planting between Valentines’ Day and St. Patrick’s Day in the Upper South. Introduced red and white clovers are commonly frost seeded every several years into cool-season pastures (Kubesch et al., 2020).
Seeding clovers can also take advantage of physiological mechanisms that improve seed establishment. In the field, frost seeding involves defoliation of an existing grassland stand, broadcasting clover seed onto the stand, and letting freeze-thaw cycles incorporate the seed into the soil surface. Compared to many native and introduced grasses, clover seed coats allow the seed to survive freeze-thaw incorporation into the soil surface. Quesenberry et al (2003) reports that buffalo clover has a similar seed weight to introduced clovers. A common rate of pasture frost seeding is 4 lb/A red clover and 2 lb/A white clover (Kubesch et al., 2020).
Optimizing rather than maximizing seeding rate is desirable given the limited seed availability of buffalo clover as well as the desire to increase planting area in restoration attempts. Managers want to get a good stand with as little seed as necessary. In addition to generating stand densities that justify direct seeding over plugging, an optimal seeding rate should generate ground cover that conserves soil as well as meets existing criteria for composition.
The present experiment sought to determine whether a 2 lb/A or a 4lb/A seeding rate can optimize buffalo clover establishment relative to white and red clover. This objective was measured through emergence as well as cover assessments. The hypothesis of this study was that the higher seeding rate will achieve the aforementioned targets comparable to, or greater than, red clover and white clover.
Materials and Methods
The present experiment ran from November 2022 to February 2023. Two improved varieties of red (c.v. Dynamite; Grassland Oregon, Salem, OR) and white clover (c.v. Domino; Grassland Oregon, Salem, OR) were compared to a seed increase of a single plant of buffalo clover (plant material VT-2022-1; see Kubesch et al., 2022). A pot study approach was employed given the investigative focus on early establishment conditions. The 4 lb/A and 2 lb/A rates are common for introduced clovers and provide a starting baseline.
Seeding rate treatments were converted from lb/A to seeds/pot using a published seed weight (Table 1; Quesenberry et al., 2003):
lb/A * A/43560 ft2 * 0.111 ft2/pot * seeds/lb = seeds/ pot
Table 1. Clover species, uncoated seed weights, and seeding rates in a seeding experiment. *
Seeds from the three species were sprinkled over moist potting mix (Vigoro All-Purpose Potting Mix; 0.07-0.04-0.03) and then watered in with tap water. The pot received water 1-2x a week in order to maintain potting mix moisture. Pots were maintained on a light bench at room temperature (18 degC) and 24 h light in 350 Smyth, Blacksburg, VA (Figure 3).
On 20 January 2023, pots were fertigated with a balanced fertilizer solution (Jack’s Professional Water-Soluble fertilizer; 20-10-20 plus micronutrients; J.R. Peters, Inc; Allentown, PA) at a ¼ tbs/gallon rate. This fertilizer was added to maintain micronutrient availability often lacking in peat-based potting mix. A general insecticide was lightly dusted on the soil surface following fertigation to prevent aphid and spider mite outbreaks (Marathon 1% granular; OHP, Inc; Mainland, PA). Adjacent experiments suffered from aphid and spider mite infestations, and the move was preventative.
Emergence, quantified from stand density, corresponded to the number of green seedlings that were visible above the soil surface in each pot. Emergence was assessed regularly at 7, 56, and 70 days after planting (DAP). Emergence contrasts with germination in that emergence omits seedlings that produce a radicle and focuses on individuals that push through the soil surface. Emergence is a conservative estimate and a field-level assessment. Tracking emergence during the first 56-70 DAP determines when seedling counts or emergence might best correspond to alternate success metrics, such as cover.
Cover corresponds to ground cover. In the present experiment, cover was assessed regularly in the DAP. For ecologists and agronomists, cover assessments often determine the botanical composition as well as management actions. In managed grasslands, cover thresholds can guide land managers to change grazing, overseed species, or completely re-establish a stand. Additionally, the desirable 30% dry matter content threshold for clovers in cool season pastures corresponds to 70-80% visual cover (Belesky and Green, 2014). In the present experiment, success was thus described as meeting or exceeding a 75% cover threshold by the end of a typical 56-70 DAP establishment period (Hall and Collins, 2018).
Data analysis
The present experiment was designed as a randomized complete block design with a factorial arrangement of clover species and seeding rate. Individual pots received an assigned species*seeding rate combination. Blocking was perpendicular to the cold end of a light bench room to account for small differences in room temperature.
Emergence and cover were analyzed as univariate responses to ANOVA. Comparisons were made at individual time points rather than across the time points. Sampling was adjusted such that measurements were taken at the same DAP for all species. Assumptions of the analysis followed ANOVA assessments. Fisher’s LSD offered a pairwise comparison of treatment means. Data analysis was conducted in SAS v9.4 (SAS Institute, Cary, NC).
Results and Discussion
Seeding rates used in this experiment reflect field-scale frost seeding rate recommendations for red and white clover. The present study did not incorporate freeze-thaw cycles, which could create seed-soil contact differently than in the watering-in approach. However, watering events were made with the intent to effect sufficient seed-soil contact for all species. Extension demonstrations from the University of Kentucky suggest that this pot study could have ensured seed-soil contact by placing seed at a drill planting depth (⅛-¼” below soil surface) rather than sprinkling seed over the soil surface.
Seed weight data came from roughly 20 years of developing literature. This standard approach from practitioners could result in differences between planned seeding rates and actual seeding rates. Currently, reported buffalo clover seed weights in the published literature (e.g. Quesenberry et al., 2003) might differ significantly from seed weights among available plant material (e.g. USDA-GRIN accessions). A recent publication reports a slightly lower seed weight of approximately 534000 seeds/lb across the geographic range of buffalo clover (Sanne et al., 2023). This global average summarizes moderate differences in seed weight among accessions. Sanne et al. (2023) also calculated that a pure stand rate for buffalo clover would be 9 lb/A, assuming a pure live seed basis. In an agronomic context, clovers are generally overseeded on a proportional basis, but further study on higher seeding rates (5-9 lb/A) is advisable.
As a percentage of sown seeds, emergence in the first week did not differ among seeding rates or species. At the 56 DAP emergence evaluation, no differences were detected in emergence among seeding rates or species (P=0.3466 for species; P=0.6140 for seeding rate; and P=0.6847 for the interaction of species and seeding rate).
No statistical differences were detected among the treatment combinations at 70 DAP (P=0.6695 for the interaction term). Emergence across species came to 31%. Notably, all plantings produced stands which consisted of less than 70% of the seeds planted (Figure 4). Technical standards for red and white clover in the Upper South expect a minimum of 12-16% emergence to pass cost-share stand evaluations (USDA NRCS, 2013). Buffalo clover can meet this threshold, which might favor eventual cost-share conservation work.
These demographic observations suggest that early seedling mortality between germination and full establishment can reduce populations meaningfully across all three species. Buffalo clover’s comparable emergence to red and white clover is encouraging, especially given that the buffalo clover germplasm available has not been selectively bred for agronomic vigor, as has been done for red and white clover.
Cover provides a quick assessment of botanical composition as well as soil conservation measurement (Symstad et al., 2008). Visual cover assessment has been used in eastern grasslands to monitor stands, and can be layered into biomass measurements on the same space (Tracy and Bauer, 2019).
Cover did not differ among pots of any species or seeding rate at 56 DAP after planting (P=0.1641 for species, P=0.1455 for seeding rate, P=0.0870 for species*seeding rate). No statistically significant differences were present at 70 DAP (P=0.1247), and cover was 35% across all species. The window from 56- 70 DAP is when a manager might consider using livestock, removing exclosures, or defoliating plants. In an agronomic context, most forage stands are not hayed or grazed until 70 DAP (Hall and Collins, 2018). In this experiment, the mean cover of all species*seeding rate combinations ranged from ~10-75% (Figure 5). All species failed to meet the target cover criteria at the start of this critical management period.
Criteria must precede the term success. The present experiment suggests that different criteria are present for cost-share and conservation functions. For a cost-share context, emergence targets would be sufficiently met. Meeting this target early can result in cost-sharing early on. From a density perspective, Vogel and Masters (2001) suggest that >20 plants m2 suggests a fully successful stand, 10-20 plants m2 suggests an adequate stand, and <10 plants m2 suggests an unsuccessful stand. Across all species, the mean 70 DAP stand counts scaled to 33-167 plants m2. This result suggests that these stands were fully successful in terms of density. The study sought to meet or exceed 75% cover by the end of a typical establishment window (56-70 DAP). The failure to meet this target suggests that higher seeding rates may be necessary to get populations established in the field or longer establishment periods may be needed. In a frost seeding situation existing vegetation needs to be controlled and regularly harvested in order to allow clover seedlings to establish (Tracy et al., 2014). This experiment failed to meet cover targets in the absence of competing vegetation (Figure 6).
Red and white clover seeding into established stands requires effective competition control with existing grasses, rather than broadcasting or drilling methods (Tracy et al., 2014; Schuluter and Tracy, 2012). Scaling this experiment up to field trials should build on these findings. Seeding method matters less than reducing initial grass competition. Subsequently, managers should maintain a low standing forage mass through a clipping every 3 weeks from April to October. Further research might evaluate the optimal seeding rate for buffalo clover when frost seeded into the field. A field trial of multiple accessions might also offer value to future authors, as such trait diversity within buffalo clover may result in more specific seeding rate recommendations for different accessions.
Conclusions
This experiment sought to determine whether a 2 lb/A or a 4lb/A seeding rate can optimize buffalo clover establishment relative to white and red clover. This objective was measured through emergence as well as cover assessments. The hypothesis of this study was that the higher seeding rate would achieve the aforementioned targets comparable to, or greater than, red clover and white clover. Undomesticated buffalo clover had a similar establishment to cultivars of red and white clover. Contrary to the hypothesis, a 2 lb/A seeding rate of buffalo clover has similar emergence and cover as other species*seeding rate combinations. However, given that treatments did not consistently meet cover thresholds, higher seeding rates across all species seem necessary.
Buffalo clover seeding rate experiments are not present in the academic or technical literature. Optimizing a seeding rate to meet field objectives and efficiently with seed resources is critical for ecologists and agronomists. Buffalo clover might serve as a native alternative to introduced clovers in agricultural grasslands. The optimum seeding rate for meeting all field objectives was not found between 2-4 lb/acre.
Acknowledgements
The authors would like to thank the owners of Arnold Classic Farms, Hoot Owl Hollow Farm, and Country Home Farms for supporting this research. Special thanks to Dr. S. Ray Smith at the University of Kentucky for his support of clover provenance tracking as well as his recent talk to Kentucky producers regarding frost seeding. Kubesch thanks his wife, Sarah Grace, for continuing the native clover research program in Pembroke, VA. This pot study was conducted at Virginia Tech with the support of Dr. Ben Tracy. The light bench used came generously from Velva Groover.
References
References for this article are hyperlinked throughout for ease of access. If any links fail to open, please contact the corresponding author: Jonathan Kubesch, M.S. (jakubesch@gmail.com).