Breeding and Development of Perennial Sunflower and Perennial Flax Varieties for Low-Input Agriculture

Final Report for GNC06-071

Project Type: Graduate Student
Funds awarded in 2006: $9,700.00
Projected End Date: 12/31/2007
Grant Recipient: University of Minnesota
Region: North Central
State: Minnesota
Graduate Student:
Faculty Advisor:
Donald Wyse
University of Minnesota
Expand All

Project Information


This project is part of a long-term program that will bring perennial sunflower and perennial flax into agricultural systems to allow farmers to diversify their operations, improve profits, improve environmental quality, and reduce inputs of labor and supplies. In the last year and a half, we developed perennial flax breeding populations that will be evaluated in several organic and conventional field sites in Minnesota and North Dakota this year. A major breakthrough was achieved in our perennial sunflower populations, in which we retained perennial habit after back crossing with an elite annual inbred, an accomplishment never before recorded.


Perennial crops provide a combination of benefits to farmers and agriculture that cannot be conferred by annual crops. Existing perennial crops, such as alfalfa, protect against soil erosion and nutrient loss and improve water use efficiency. Fall tillage is not necessary during the multi-year lifetime of the crop stand. Cover cropping or polyculture may also be used successfully with perennials to reduce the need for summer tillage and herbicides. Perennial crops provide living ground cover for longer periods during the growing year because the plants can emerge from dormancy early in the growing season. These plants take up soil moisture, decreasing year-round tile line flow and concurrent leaching of nitrogen from the soil (Randall et al., 1997). Nitrogen from tile lines contributes to the contamination of surface waters in Minnesota and elsewhere (Randall and Iragavarapu, 1995).

Sunflower and flax have valuable seed oils that make them profitable for farmers to grow. In particular, sunflower has high-oleic and mid-oleic (also known as NuSun®) oil profiles, which are important seed oil profiles for the production of no-trans-fat vegetable oil. For the last several years, demand for NuSun® oil has exceeded supply. This demand is mostly from processed food manufacturers such as Frito Lay. Flax oil is also in demand as a health supplement, and has one of the highest concentrations of Omega-3 oils found in plants. Consumption of Omega-3 oils has been linked to improved heart health. Additionally, the black-and-white seeded (confectionary) sunflowers are grown for the production of snack-type sunflower seeds. This is a particularly high-value market.

The combination of environmental and economic benefits from perennial crops is important for producers who are looking for ways to reduce inputs and environmental harm while increasing profits. We believe that sunflower and flax are ideal candidates for perennialization because both the sunflower and flax genus contains perennial species that are similar to, or can intermate with, the crop-type species.

This project is a seminal study into the feasibility of using perennial relatives of existing annual crops to develop perennial varieties with the agronomic and seed characteristics similar to the domesticated relative. Surveys of germplasm and progeny of interspecific crosses indicate that variability exists for yield components, which will allow us to select for high yield in populations with perennial habit. Piper (1992) showed no significant correlation between vegetative vigor and yield in the perennial species Cassia marilandica in 3 out of 4 years. Thus, we expect that there will be little or no negative association between perennial habit and yield in perennial varieties.

Perennialization of domesticated sunflower was first proposed by Ščibria (1936), who noted that populations of H. tuberosus X H. annuus could be candidates for development of a perennial sunflower. While hybrid populations have been formed for disease resistance breeding in the past, no one has published specifically on the mode of inheritance of perennial habit in such populations. There has been casual mention of rhizome/tuber production in some F1 populations, while in other populations there was no production of rhizomes or tubers (Cedeno et al., 1985; Kräuter et al.,1991). The populations with noted tuber production tend to have H. tuberosus as the female, indicating that perennial habit may have at least partial cytoplasmic inheritance. Previous investigators have found that the F1 hybrid plants are tetraploid (2n=4x=68) and the BC1F1 plants are triploid (2n=3x=51) (Cedeno et al., 1985). Low fertility was found in the BC1F1 plants, which was likely due to the triploid condition of the genome. Further backcrossing gives rise to progeny with greater meiotic stability through progressive loss of chromosomes, and this has led to fertile, diploid populations after several backcrosses (Cedeno et al., 1985; Hulke and Wyse, 2008; C.C. Jan, personal communication, 2008). The development of resistance to white mold, downy mildew, and phomopsis diseases has been derived from wild perennial species in this way (Pustovoit et al., 1976; Skoric, 1985).

Breeding of perennial flax species was recently proposed by our research group and researchers at The Land Institute, Salina, KS. Many perennial relatives of flax exist that have omega-3 fatty acid profiles similar to that of annual flax (Yermanos and Beard, 1964). These species are split into two ploidy groups, 2n=2x=18 and 2n=2x=30. The latter group includes both the annual domestic flax and perennial relatives. Considerable cross-fertility has been noted between many of the 2n=18 species and between some 2n=30 species (Gill, 1987). Crossing between the two genomic groups is not likely to be possible because of the large genomic differences.

Our group initiated breeding of perennial sunflower and perennial flax prior to the start of this SARE project. The perennial sunflower breeding program was initiated in 2003 from crosses between 16 wild H. tuberosus accessions collected near Rosemount, MN, and RHA 265, a H. annuus fertility restorer line with high oil content. The resulting F1 plants were used as parents in a traditional backcross program using HA 434, an annual, high-oleic acid inbred line, as a recurrent parent. The BC1F1 plants produced very few seeds when backcrossed, resulting in a population bottleneck. None of the BC1F1 plants produced regrowth after the winter of 2005-2006, indicating that perennial habit was not retained through a backcross. New populations were formed for this SARE project to attempt to get around this problem, as described below. The perennial flax breeding program began in 2001, with observation plots of 13 perennial Linum species. Those that overwintered well were used as parents in crosses during the summer of 2004. Crosses were successful among 2n=18 perennial species, but no progeny were obtained in crosses between 2n=18 and 2n=30 species or among 2n=30 species. Because it appeared that interspecies crosses between the crop-type annual and perennial species of either group were impossible, we planned to directly improve populations of 2n=18 species in this SARE project.

Project Objectives:

Objectives of current research:
1. To produce and improve interspecific populations of sunflower to begin the process of moving the perennial habit genes from H. tuberosus into the crop-type annual genome of H. annuus.
2. To begin the process of improving the perennial species of flax (Linum spp.) for agronomic characteristics using recurrent selection.

Short term outcomes:
1. Determine the inheritance of perennial habit in annual x perennial sunflower populations while developing additional populations, as needed.
2. Develop perennial flax populations for line evaluation at multiple locations.

Intermediate term outcomes:
1. Develop perennial sunflower inbred lines.
2. Develop perennial flax open-pollinated varieties (OPVs).


Click linked name(s) to expand/collapse or show everyone's info
  • Carmen Fernholz
  • Jerry Miller
  • Donald Wyse


Materials and methods:

Perennial Sunflower
Germplasm resources as of project start date: Tuber stocks of 18 H. tuberosus plants were collected in 2001 from UMORE Park, Rosemount, MN. Nine of the plants were collected from cultivated areas and were named JA1 to JA9. Nine more plants were collected from undisturbed areas approximately 8 km (5 miles) from the first site of collection. These were named JA10 to JA18. All plants were moved to plots at the St. Paul Agricultural Experiment Station, St. Paul, MN, where they still exist as a living plant collection. All plants are believed to be native to the area in which they were collected in east central Minnesota. In addition, seeds of H. annuus inbred lines CMS HA 89, HA 89, CMS HA 434, and HA 434 were obtained from Gerald Seiler and Jerry Miller, USDA-ARS, Fargo, ND, USA, over the course of the project. In 2005, reciprocal crosses were made between the 18 H. tuberosus genetic stocks and two sunflower maintainer inbreds (HA 89 and HA 434). The CMS versions of these inbreds were used when the inbred was the female parent, to ease cross-pollination. This resulted in 4 types of populations: CMS HA 89 X JAx, JAx X HA 89, CMS HA 434 X JAx, and JAx X HA 434, where x = 1 to 18 (denoting each of the H. tuberosus plants). The CMS lines contribute the French male-sterile cytoplasm of Leclercq (1969), which is the same cytoplasm used in elite annual hybrids grown by farmers today. The JA lines contribute H. tuberosus cytoplasm.

Methods used during project: The F1 plants of the 4 population types listed above were grown in the greenhouse during the winter of 2005-2006 to attempt a new backcross procedure. Because the typical backcross procedure of crossing the tetraploid F1 with the diploid annual resulted in a population bottleneck, we attempted a backcross program using tetraploid annual sunflower lines instead of diploid lines. The tetraploid lines HA 89(4x) and HA 434(4x) were produced by treating HA 89 and HA 434 plants with colchicine according to the protocol of Jan and Chandler (1989). F1 plants were then crossed with either HA 89(4x) or HA 434(4x), whichever matched the annual parent in the original cross, to produce BC1F1(4x) seed. (The “4x” superscript indicates that the plants are tetraploid.) In the summer of 2006, the F1 plants were cut back and transplanted to the field for observation, and the BC1F1(4x) plants were started in the greenhouse and moved as seedlings to the field. During the summer of 2006, the BC1F1(4x) plants were sib-mated within full-sib families and self-pollinated. Some plants were also backcrossed with the 4x inbred stocks, if they appeared to produce rhizomes during the growing season. The nurseries of F1 and BC1F1(4x) plants were screened for surviving plants following the winter of 2006-2007, which indicated presence or absence of perennial habit in our environment.

Perennial flax
Germplasm resources as of project start date: Perennial flax accessions were obtained from collections at the North Central Regional Plant Introduction Station, Ames, IA, and Black Hills State University, Spearfish, SD. These accessions were planted at the St. Paul Agricultural Experiment Station, St. Paul, MN, in 2001. The following accessions could survive the winter in St. Paul: PI 263511, PI 383683, PI 440472, PI 440473, PI 502405, PI 502406, PI 502407, PI 502408, PI 522296, PI 522305, PI 522306, PI 545704, Ames 19140, Ames 22537, Ames 22586, Ames 25334, Ames 25406, and Ames 25646. The accessions with good survival were random mated with other accessions to form two distinct RM1 populations in 2005. One population has over 45% of the parental base originating from a line of L. baicalense with large seed size. Another is a population that includes germplasm from almost all of our surviving accessions, and is considered our broad-based population.

Methods used during project: The plants with the fastest germination and best seedling vigor of the broad-based and large-seeded perennial flax populations were random-mated once more to form the RM2 population during the summer of 2006. This was necessary because seed germination and seedling vigor were poor in the RM¬¬1 population, and we wished to evaluate agronomic traits on a larger population. The RM2 population was transplanted into the field at St. Paul in the fall of 2006. The plants that overwintered successfully were grown to maturity during the summer of 2007, and the plants harvested and threshed individually. Each plant was given an identification number, corresponding to the parent plant in the field. The seeds have been divided in anticipation for replicated field trials to be planted in spring 2008.

Research results and discussion:

(Listed according to projected outcomes)
Short-term outcomes
1. Determine the inheritance of perennial habit in annual x perennial sunflower populations while developing additional populations, as needed. The F¬1 plants with the French (annual) cytoplasm and H. tuberosus (perennial) cytoplasm were observed during the summer of 2006 and survival was determined after the winter of 2006-2007. Nearly 100% of the F1 plants produced tubers in the field and survived winter successfully, indicating that there was no maternal or cytoplasmic effects on perennial habit. Further, the formation and behavior of the perennial organs were similar to the perennial parent, indicating the perennial habit is due to a dominant, gain-of-function set of genes. Crosses with the perennial parent produced more progeny; however, this was expected because the parent with the higher ploidy level generally produces more seed in interspecific crosses. The BC1F1(4x) populations produced in 2006 included about 1000 plants, which is nearly a ten-fold increase in population size over typical triploid BC1F1 populations. Of these 1000 plants, 8 plants were perennial as determined by regrowth after the winter of 2006-2007. These plants will be maintained for further study in our permanent collection. Some plants in the population did not survive in the field, but appeared to produce rhizome-like sprouts. These plants may have had some, but not all, of the genes necessary to be a successful perennial in our environment. (H. tuberosus, the donor of perennial habit, is a Minnesota native plant, and is extremely winter hardy in our environment.) It appears from the infrequency of perennial progeny that there are several important perennial habit genes. The data also suggest it will be possible to carry these genes through successive backcrosses if care is taken to maintain an adequate population size. These results will be published this year in the Proceedings of the 17th International Sunflower Conference, Cordoba, Spain. We believe that we have reached our short term objectives of obtaining information on the inheritance of perennial habit, producing new breeding populations, and publishing our findings. We have exceeded our own expectations by making a breakthrough: production of a BC1F1 population with fully perennial individuals.

2. Develop perennial flax populations for line evaluation at multiple locations. During the course of this SARE project, we have been able to improve the seed germination and vigor of the seedlings using a recurrent selection approach, in which only the plants with good vigor and fast germination were selected from the populations and random mated. Ideally, we would have evaluated these factors together with yield and agronomic traits, but the seedling vigor and seed dormancy factors would have affected the observations on yield and agronomic traits, rendering the data sketchy at best. By selecting for seedling vigor and fast germination first, we obtained two populations that had better germination and still maintained a great deal of the initial genetic variation of the population. This is what we would expect according to breeding theory. Genes at locations distant to the genes responsible for seed dormancy and seedling vigor factors are largely unaffected by selection for seed dormancy and seedling vigor. Thus, we can improve one very important characteristic first, and focus on agronomic and yield characteristics second without the confounding nature of seedling vigor differences among the lines. We will begin the first screening for agronomic and yield factors starting this spring. Lines were derived by harvesting seed from individual plants that survived the winter of 2006-2007. These lines are called half-sib lines, because the female parent is a single plant, but the male parents are all of the other plants in the population. By planting these lines out in the field in a randomized, complete block design at multiple environments, we can determine the total contribution of genetics to total field variation. Further, we will obtain information on which plants have the best genetics. We can then random mate the best plants to produce populations with better characteristics. We believe that we have reached our short term objectives by generating populations with improved seedling vigor and less seed dormancy than previous populations. The lines selected for agronomic evaluation starting this year (2008) were produced exclusively from plants that were able to survive the winter of 2006-2007; therefore, the lines are expected to have improved winterhardiness over most of the parent lines.

Intermediate-term outcomes
1. Develop perennial sunflower inbred lines. The short term nature of this project, about a year and a half, did not allow us enough time to produce perennial sunflower inbred lines of high quality. We include this as an intermediate-term outcome to indicate our goal to produce desirable lines for release to farmers and seed producers. To this end, the last year and a half has been critical to our current understanding of perennial genetics in the genus Helianthus, and by achieving BC1F1 populations with perennial habit, we have made a giant leap forward in making perennial sunflower a reality to farmers. As soon as possible, we hope to provide breeding material to outlying experiment stations and organic farmers for evaluation.

2. Develop perennial flax open-pollinated varieties (OPVs). Again, the short term nature of this project, about a year and a half, did not allow us enough time to produce perennial flax OPVs of high quality. The last year and a half did provide us with critical advancements in perennial flax variety development, however. We have achieved populations with better seedling vigor and less dormancy, which makes development of yield trials more predictable and uniform. Further, with the help of our contributing organic farmer, Carmen Fernholz, and additional experiment stations in Minnesota and North Dakota, we will be able to begin data collection on replicated, half-sib line evaluations at multiple environments in 2008. These evaluations will provide us information to (1) improve the existing populations by random mating the best plants with the best plants (to begin a new cycle of recurrent selection) and (2) give us the opportunity to selectively mate plants from the population that produce variety-caliber progeny (to produce a variety). The second outcome may or may not be possible in the current round of recurrent selection. This will be decided based on the data collected from the field.

Cedeno, R, MS McMullen, and JF Miller. 1985. Cytogenetic relationship between Helianthus annuus L. and H. tuberosus. p.541-546. Proceedings of the 11th International Sunflower Conference, Mar del Plata, Argentina. 10-13 Mar. 1985. International Sunflower Association Toowomba, QLD, Australia.

Gill, KS. 1987. Linseed. ICAR, New Delhi, India.

Hulke, BS and DL Wyse. 2008. Using interspecific hybrids with Helianthus tuberosus L. to transfer genes for quantitative traits into sunflower, H. annuus L. Proceedings of the 17th International Sunflower Conference, Cordoba, Spain. In press.

Kräuter, R, A Steinmetz, and W Friedt. 1991. Efficient interspecific hybridization in the genus Helianthus via “embryo rescue” and characterization of the hybrids. Theoretical and Applied Genetics 82:521-525.

Piper, JK. 1992. Size structure and seed yield over 4 years in an experimental Cassia
marilandica (Leguminosae) population. Canadian Journal of Botany 70:1324-1330.

Pustovoit, G.V., V.P. Ilatovsky, and E.L. Slyusar. 1976. Results and prospects of sunflower breeding for group immunity by interspecific hybridization. p.193-204. Proc. 7th Int. Sunflower Conf., Krasnodar, USSR. 27 June – 3 July, 1976. Int. Sunflower Assoc. Toowomba, QLD, Australia.

Randall, GW, DR Huggins, MP Russelle, DJ Fuchs, WW Nelson and JL Anderson. 1997.
Nitrate losses through subsurface tile drainage in Conservation Reserve Program, alfalfa, and
row crop systems. Journal of Environmental Quality 26:1240-1247.

Randall, GW, and TK Iragavarapu. 1995. Impact of long-term tillage systems for continuous corn on nitrate leaching to tile drainage. Journal of Environmental Quality 24:360-366.

Ščibria, NA. 1936. Hybrids between the Jerusalem artichoke (Helianthus tuberosus L.) and the
sunflower (Helianthus annuus L.) C.R. (Doklady) de l’Académie des Sci. de l’URSS 11:193-

Skoric, D. 1985. Sunflower breeding for resistance to Diaporthe/Phomopsis helianthi. Helia 8:21-24.

Yermanos, DM, and BH Beard. 1964. Fatty acid composition of the oil of wild species of flax. p. 24-25. In: Papers presented at the 34th Annual Flax Institute of the US, Nov. 12-13, 1964.

Participation Summary

Educational & Outreach Activities

Participation Summary:

Education/outreach description:

Hulke, BS and DL Wyse. 2008. Using interspecific hybrids with Helianthus tuberosus L. to transfer genes for quantitative traits into sunflower, H. annuus L. Proceedings of the 17th International Sunflower Conference, Cordoba, Spain. In press.

Hulke, B, and D Wyse. 2006. Introgressing genes for perennial habit into sunflower (Helianthus annuus L.) via wide crosses [CD-ROM computer file]. Abstracts 2006 ASA-CSSA-SSSA Int. Annu. Meet. ASA-CSSA-SSSA, Madison, WI.

June 2008 – A lecture on perennial sunflower breeding will be presented to scientists at the 17th International Sunflower Conference, Cordoba, Spain (proceedings noted above in publications).

July 2006 – Spoke to the Northern Seed Trade Association regarding perennial sunflower and flax breeding. Formal tour of the plots was included. (Handouts included in Appendix).

November 2006 – Presented a poster (noted above in publications) at the 2006 Crop Science Society of America annual meeting on perennial sunflower breeding.

Throughout project – Many informal plot tours at St. Paul were given to interested parties.

Project Outcomes

Project outcomes:

The data produced from this project and the preliminary work done before this project have been instrumental in learning about the inheritance of perennial habit in sunflower and developing breeding strategies for perennialized crop-type sunflower. Some of the F1 sunflower populations are also being studied for potential as a feedstock for cellulosic ethanol production. Further, we have determined that there is variation for seedling emergence, vigor, and some agronomic traits in perennial flax populations. We will add to this knowledge with field evaluations being conducted in 2008 and in future years. Because this project is of a longer term nature than most SARE projects, it is hard to determine at this time what the economic impacts will be. If these crops are successfully perennialized, the potential is there for the perennial versions of these crops to replace the annual versions on the rural landscape. Consistently high demand for the value-added products of these crops over the last several years, combined with the potential for enhanced erosion control and nutrient scavenging, could make perennial sunflower and flax more desirable than the annual versions of the crops. One reason that sunflower is avoided by producers today is because it is a relatively high-maintenance crop. A perennial version would require less maintenance. Short term, the impact is largely scientific, although the work is drawing interest from many producers who ask for tours of our plots. Over the longer term, as perennial varieties are released, the impact could include changes in crop rotation and other production practices. These crops will also have conservation value.

Economic Analysis

Economic analysis is beyond the scope of this study.

Farmer Adoption

No farmer recommendations at this point. We believe that this project has affected the mindset of some farmers in our state regarding the potential of new crop research. The exact number is unknown, but we estimate the number of farmers that have witnessed or participated in this project exceeds 50. This will ease producer adoption problems in the future, as the farmers are already aware of this work through field days and presentations.


Areas needing additional study

There is an urgent need to keep these breeding programs going. All of the research thus far has led us to believe that perennial sunflower and flax can become a reality with a modest research investment. Specific areas of future research include:
(1) using RNA-mediated technologies to find candidate genes for perennial habit in our sunflower backcross populations,
(2) further refining the backcross procedure in sunflower, and
(3) determining the heritability of agronomic traits in our perennial flax populations. We are currently working on recruiting a new graduate student to follow up on this work. Brent (now Research Geneticist, Sunflower Unit, USDA-ARS), Don Wyse, and Robert Stupar (a new member of our team) will co-advise this student.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.