Jonathan O.C. Kubesch*, Kelly Winklelpleck*, Connor Doyle*, Lindsey Barbini*, John H. Fike*, and Michael Johnson**
*Virginia Tech School of Plant and Environmental Sciences
**Clarks River National Wildlife Refuge
Summary
Buffalo clover (Trifolium reflexum) is a true native clover of eastern U.S. provenance. Previous research suggests that buffalo clover has seed yields and weights similar to introduced clover species (making it suitable for increase). However, such research is limited, with little comparison of plant material from across the northern range of the species. In fact, there is overrepresentation of southern accessions from Georgia, Texas, Mississippi, and Florida. This overrepresentation stems from historical seed collections coming from Coastal Plain collections. In order to better represent accessions from the northern range of the species (both in current research and for future studies), this project sought to 1) conduct a comparison of buffalo clover accessions from across the range against red clover (T. pretense); and 2) bank seed from the Clarks River population for use in conservation and research. Plants were grown from seed to seed and studied in a common garden experiment with efforts taken to maintain genetic purity. Seed was then banked with the Southeastern Grasslands Initiative. Variability in phenotypes across the lifecycle were observed among northern and southern buffalo clover accessions, which suggests that the species complex has further structural variance. Southern accessions required less time to bloom and flowered without vernalization. Clarks River plant material has not yet flowered and appears to require vernalization.
Introduction
The native clovers of eastern North America are relics of the complex mosaic landscapes present prior to European settlement (Gillett and Taylor, 2001). Of these species, buffalo clover is an annual to biennial species with known morphological and reproductive variability (Vincent, 2001).
Kentucky clover (T. kentuckiense) is a recent addition to the flora (Chapel and Vincent, 2013). Species delimitation between Kentucky and buffalo clover has been determined using primarily floral characteristics; however, vegetative characters in a common garden experiment might further resolve the relationships within the species complex.
Previous work with Kentucky clover and running buffalo clover in greenhouse and field research suggest that these native species face fewer propagation challenges in comparison to other rare plants (Kubesch, 2018; Kubesch 2020). Historically, Dr. Norman Taylor at the University of Kentucky maintained all 3 clovers native to Kentucky in greenhouse collections (Kubesch, 2018; Daniel Boone, personal communication).
The objectives of this project were to collect seed from the northern edge of the buffalo clover range and compare buffalo to Kentucky clover. It was hypothesized that Kentucky clover would be more morphologically similar to buffalo clover accessions from Kentucky and Ohio than to accessions from the Coastal Plain.
Materials and Methods
Seed Collection and Processing
Seeds were collected from the Clarks River population on 2 July 2020, with the goal of collecting 10-15 viable seeds from each individual. Plants were selected for the collection if they had more than one bloom, in order to minimize potential effects on the wild population. Flowerheads were collected whole in the field and kept separate by individual in packets.
Seeds were processed on 28 July 2020 at the University of Tennessee, Knoxville. For each individual, the seed head material was poured onto a course screen (#10, 2mm screen). And then rubbed to release seed from the calyx. The loose petals were removed from the course screen and pooled into the catch pan The remaining material was poured onto the fine screen (#20, 0.84mm). Debris which passed through the screen was placed in a manila envelope. Seed and larger debris went into a plastic jewelry bag. This process was repeated for each individual.
Germination
Germination procedure followed the general procedure of Quesenberry et al. (2003). Shallow plastic germination boxes (initially disinfected with ethyl alcohol and air-dried) were used for germination efforts. Seeds were placed on saturated germination paper (38# seed germ 2003335; Anchor Paper Co, St. Paul, MN) within the boxes beginning 25 January 2021. For each of the Clarks River plants, 18 seeds were used. Boxes were maintained on a lab bench at ambient temperature (approximately 76 °F) with fluorescent lights on continuously.
During the germination period, paper moisture was maintained by adding water with a syringe. On 10 February 2021 seeds were transferred to new germination boxes containing fresh paper. Boxes received an aqueous solution of butanolide which was thought to potentially improve germination. Seeds were scarified on 17 February 2021 using sandpaper (#M, 120-medium grit smooths) in accordance with the Sustar (2017) protocol. Germination was recorded when a radicle was visible.
Greenhouse growth
Seedlings were transplanted into cell packs in 48-cell trays. The growing media was Miracle Gro potting mix (Miracle Gro Lawn products Inc., Marysville, OH). Trays were maintained on a common bench in a Virginia Tech greenhouse and moisture was initially maintained through bottom-watering. In order to provide initial nutrients, a soluble fertilizer was bottom watered into the trays on March 5, 2021 and March 23, 2021 (Buddha Bloom 0.5-2-1 ; 20 mL/gal). Pest control was necessary as some damage from palmetto bug roaches (Sci name?) was observed. Glue traps were set around trays and insecticides were sprayed around the bench but not on or near the plants.
Plants were potted up to 4” pots from the cell packs on 29 and 31 March 2021 using the same potting mix. Plants received 200 mL of a P and K fertilizer to ensure sufficient availability for growth (HPK; roots organic, 0-4-3; 20 mL/gal). Plants developed central crowns and produced foliage. Plant vigor and growth increased following the potting up. Plants were watered to saturation 2-3x per week as needed.
Pest control on the plants was not considered until 13 April 2021, when mite damage was identified on some of the red clover leaves (confirmed by Hunter Hammock). Damage was not considered to pass the threshold, defined as plant fatalities.
Plants were fertilized again on 30 April 2021 with soluble fertilizer (HPK; Buddha Bloom). In order to further supplement micronutrients, Osmocote Plus was applied on 7 May 2021 (Osmocote Plus; 15-9-12 with Mg, S, B, Cu, Fe, Mn, Mo, and Zn; The Scotts Company, Marysville, OH). Plants also were fertilized with a Maxigro fertilizer solution (Maxigro; 10-5-14; General Hydroponics) on 30 April and again on 9 July 2021.
Mite damage was severe by 9 July 2021; thus, plants were moved from the greenhouse to the Virginia Tech Urban Horticulture Center. Sevin insecticide was sprayed (A.I.? rate?) after watering plants to saturation on 16 July 2021.
On 11 August 2021, surviving native and red clovers were potted up to larger quart pots in general purpose media (Classic 300, Nursery Supplies Inc, Chambersburg, PA; BMS300/Elite 300S, ITML; PROMIX BX Mycorrhizae General Purpose, Premier Horticulture Ince, Quakertown, PA). On 11 October 2021, ½ tsp of Osmocote Plus was applied to all plants. Oxalis weeds as well as dead clover foliage were removed from all pots.
Reproduction
Buds, blooms, and post-blooms were monitored starting on 31 March 2021 for the TX400 accession. Buds were defined as inflorescences still green. Blooms were defined as inflorescences with open petals in color. Phenotyping of flower color occurred at the bloom stage. Post-bloom referred to inflorescences that were past bloom and started to brown for seed formation.
Seed production
Seed collection began when a set list of criteria were achieved, using recommendations from clover breeder, Dr. K.H. Quesenberry, University of Florida. Head harvest commenced when a post-bloom inflorescence had a brown calyx, seeds appeared visually yellow, and the peduncle was brown back to its junction with the main stem. Heads, separated by parent plant, were placed into coin envelopes. Seeds were stored at ambient temperature and humidity for potential after-ripening. Seed collection from individual plants began 28 April 2021 for the TX400. Red clover started blooming on 4 May 2021, though TX400 already passed bloom.
Seed banking and provenance tracking
Original seeds that were not used were stored in order to prepare additional batches of plant material if needed. The seed produced during the greenhouse period was processed and prepared using the sieving method as described above.
Reporting is described below.
Data analysis
Red clover plants appeared to have fewer, larger leaves throughout the grow-out period. The Kentucky and buffalo clovers all produced many, smaller leaves during the period, prompting collection of a series of morphological measurements on 23 April 2021. Four metrics were recorded. Canopy height was measured as the distance from soil level to the highest leaf of the main canopy. Canopy width was measured at the widest point across the canopy. Petiole length was measured on a random, mature leaf as the distance from the crown to the leaf. Leaves on the reproductive stem were not used for petiole length measurements. Leaf count was measured as the number of living leaves at any developmental stage. Senesced leaf material was excluded from leaf count. 128 plants of the 208 total plants present were measured, as the additional material was mainly red clover.
These vegetative measurements were analyzed at the species and accessions level by ANCOVA using SAS and DandA macro. The plants were treated as a completely randomized design with tray number as a covariate to account for spatial position in the greenhouse. Correlations among the various responses measured were made using PROC CORR. Principal Components analysis (PCA) was conducted on the responses for dimensional reduction. The first 2 components contributed to 88.62% of the variation. Graphical exploration of Factor1 and Factor2 was used to see potential clusters at the species and accession levels. ANCOVAs on the individual responses were also analyzed. For species-level analysis, Fisher’s protected LSD was used, and for accession-level analysis, Tukey’s HSD was used.
Results and Discussion
Seed Collection and Processing
Of 9 individuals, one collected had no seed whatsoever. All other plants had multi-colored seed present, namely yellow and black. Seed yields appeared comparable to distributions from S-9 USDA GRIN. More of the florets were fertile than originally expected. This fruitfulness exceeded the target collection goals.
Germination
Across the plant material brown exudate and leachate was seen on germination paper. The presence of exudates preceded the emergence of the radicle. For the TX400 and GA287, the seeds germinated immediately without scarification. Only one seed germinated from Clarks River-1 prior to scarification. All seeds imbibed immediately following scarification.
The lack of a scarification requirement in the southern accessions was noteworthy. Dormancy mechanisms may differ across latitudes.
Greenhouse growth
During transplanting there was more root and nodulation on Cincinnati and Clarks River plant material than on the T. kentuckiense. This may have represented different N needs. Without inoculum, the active rhizobial associations observed were either from the growing media or from the environment. This result has been confirmed by Quesenberry and colleagues in Florida.
Visual inspection suggested that GA-287 resembled T. kentuckiense more so than the Cincinnati and Clarks River T. reflexum.
The growth habit of GA-287 and T. kentuckiense resembled a shuttlecock. Some Clarks River plants behaved similarly, though not to the same degree from a visual inspection. For the vegetative measurement, the first 2 components contributed to 88.62% of the variation (Figure X). Factor1 was weighted positively towards canopy height, canopy width, and petiole, but negatively towards leaf count; Factor2 was inversely weighted. Plants with higher values for factor1 were larger with fewer leaves, whereas increasing Factor2 values represented smaller plants with many leaves.
At the species level, T. kentuckiense and T. reflexum were similar for Factor1 but distinct for Factor2 with T. kentuckiense having more, smaller leaves (Figure 1). The species were distinct for canopy height and leaf count, but similar for canopy width and petiole. Canopy height and petiole length are related measures of aboveground growth. T. kentuckiense had the most leaves of the 3 species. In all cases, both of the native species differed from red clover.
At the accession level, patterns of similarity broke down among T. kentuckiense and the accessions of T. reflexum. Factor1, canopy width, and petiole width did not differ among these native clovers. Factor2 and canopy height were similar among T. kentuckiense, GA287, and Clarks River. GA287 and T. kentuckiense also had similar leaf counts. Plant morphology was similar for accessions from Georgia, western Kentucky, and central Kentucky., However, the Cincinnati accession was excluded from what?? based on similarity to the nearby T. kentuckiense site. In plotting Factor1 and Factor2, GA287 and T. kentuckiense were graphically closer to one another than any of the other accessions. Clarks River was graphically closer to GA287 and TX400 than to the Cincinnati population (Figure 2).
The distinctness of T. kentuckiense from T. reflexum was apparent with a few vegetative measurements, although there appears to be some similarity to GA287. Possible explanations for this similarity include:
- GA287 is actually a population of T. kentuckiense.
- T. kentuckiense species status might be more questionable than previously thought.
- Vegetative characteristics in the field might not be reliably demonstrated by common garden experimentation.
The original delineation of T. kentuckiense from T. reflexum was developed from a combination of vegetative and reproductive characteristics of herbarium specimens (Chapel and Vincent, 2013).
Clarks River plant material started demonstrating disease and pest symptoms on 7 May 2021. One plant had a white crown. Red petioles were present on all 3 species, suggesting nutrient deficiencies. Spider mite webs were observed on 9 May 2021.
Pest load increased from May onward, and heavy mite damage was evident by 28 June 2021. Given limitations on pesticide use in the greenhouse, plants were moved to an open-air hoophouse for subsequent treatment and care. Summer propagation in the hoop house reduced pest pressure. Mite damage resulted in a number of fatalities, most notably for Cincinnati, Clarks River, and Kentucky populations. TX-400 plants fully senesced by 13 July 2021.
GA-287 was the most successful plant under the conditions provided. Clarks River and Kentucky accessions were chlorotic for a time following the move from greenhouse to hoop house. Following the move to quart pots, plant growth improved and nutrient deficiency symptoms became less pronounced.
Competition from Oxalis from the Miracle Gro media warrants the use of other media or weed control measures in future studies. The Oxalis competed with the clovers for nutrients and made management challenging as it was difficult to avoid damaging clover crowns when manually weeding.
By 22 October 2021, fall coloration was prominent on all clover species; this fall coloration is purpling from anthocyanin expression, as well as mild chlorosis.
Reproduction
On 31 March 2021, 5 plants from the TX-400 accession started producing green buds and blooming. Bud counts, flower counts, and flower color were monitored on these plants from that time forward. On 27 April 2021, the red clover started producing reproductive stems and preparing for bud formation. On 28 April 2021, seeds were collected from the 1st blooms noted on 31 March 2021.
Seed production
Seed production corresponded to efforts in the 2020 field season. At present, only TX-400 plants have produced seed among this cohort of plants. Seed production in this context might be better measured on the basis of seedlings established as opposed to a set seed yield. The rarity of seed and the germination procedure requires more handling of seed.
Seed banking and provenance tracking
Seeds were maintained for plants on the basis of matrilineal relationships. Brad Morris and Tiffany Fields of S-9 USDA were eager to accept seed from the newly banked seeds of the Cincinnati and Clarks River accessions. Dwayne Estes and Cooper Breedon of the Southeastern Grasslands Initiative were also eager to accept seed from accessions.
Seed was mailed to the Southeastern Grasslands Initiative seedbank on December 1, 2021. A representative sample was generated from the original seed collected and sent as matrilineal relationships. Provenance was maintained by means of written records, including submission of this report, to other related parties.
Courtesy of Cooper Breedon: “The Conservation Seed Bank (CSB) is housed within the Southeastern Grasslands Institute at Austin Peay State University. The purpose is to preserve the genetic integrity of our rare and declining grassland plants and ultimately incorporate the seeds into grassland restoration plans following science-based principles. The CSB is part of our growing native seed conservation program at SGI which strives to increase the availability of appropriately and ethically sourced seed material into grassland conservation projects. Our growing program has been funded by various federal agencies, nonprofits, and foundations and includes the recently launched Seeds of Success, Southeast program funded by the Bureau of Land Management and U.S. Fish and Wildlife Service Partners for Fish and Wildlife Program which has established SGI as the SOS, Southeast hub for a 10-state seed collection effort.” (https://www.segrasslands.org/conservation-seed-bank).
Conclusions
Clarks River buffalo clover proved to provide a useful data point in intraspecies and interspecies examinations. The vigor of the plants suggested that this accession might be further increased in the greenhouse for restoration work at the refuge.
References
Chapel, K. J., & Vincent, M. A. (2013). Trifolium kentuckiense (Fabaceae, Papilionoideae), a new species from Franklin and Woodford counties, Kentucky. Phytoneuron 2013-63: 1–6.
Gillett, J. M., & Taylor, N. L. (2001). The World of Clovers. Ames, IA: Iowa State University Press.
Kubesch, J.O.C. (2018). Edaphic and morphological factors affecting running buffalo clover (Trifolium stoloniferum) ecology. Ohio State University. Columbus, OH.
Kubesch, J.O.C. (2020). Native Clover Conservation in the Bluegrass: An Agronomic Perspective. https://www.knps.org/2020/03/24/native-clover-conservation-in-the-bluegrass-an-agronomic-perspective/.
Quesenberry, K.H., J.M Mullaney, A.R. Blount, R.S. Kalmbacher, and J.G. Norcini. 2003. Characterization of physiological and morphological variability in buffalo clover germplasm. Soil Crop Sci. Soc. Florida Proc. 62: 66-69.
Sustar, E. 2017. The effects of scarification on germination rates of running buffalo clover. Ohio State University, Columbus, OH. Unpublished Research Report.
Vincent, M.A. (2001). The genus Trifolium (Fabaceae) in Kentucky. J. Ky. Acad. Sci. 62(1):1-17.
Acknowledgements
Kubesch would like to acknowledge his wife, Sarah Grace Kubesch, for her role throughout this field and greenhouse experimentation; he is grateful for her encouragement. Seed from Clarks River was courtesy of Clarks River National Wildlife Refuge, Benton, Kentucky, courtesy of the U.S. Fish and Wildlife Service. Additional thanks are due to Tammy Stackhouse and Hunter Hammock for their work in clover conservation.
Jonathan Omar Cole Kubesch, M.S., is a PhD candidate in the Virginia Tech School of Plant and Environmental Sciences. His research focuses on native grass and wildflower conservation for bees and beef cattle. Jonathan received his M.S. from the University of Tennessee, Knoxville in 2020 and his B.S. from Ohio State University in 2018. In his spare time, Jonathan works on conserving native clovers from across eastern North America as well as fitting native plants into agricultural systems. Jonathan married Sarah Grace Kubesch, a 3rd grade school teacher, and together they have 1 dog, 2 cats, 4 ducks, 4 turkeys, and a mess of fish.