Perennial Helianthus Breeding (last updated 10/1/2023)

The following is intended to document my breeding progress with perennial Helianthus (sunflowers).  It will be updated as progress is made.

Due to the differing ploidies between species it is important to document and track ploidy of each parent in a cross to predict the most likely resulting ploidy in offspring so that multiple generations of breeding work can be accomplished without running into dead ends with mismatched chromosome counts.  Ploidies of 2n=2x, 2n=4x and 2n=6x are generally fertile to varying degrees while ploidies of 3x and 5x are generally infertile.

The following is a list of Helianthus species, hybrids or selections currently being used as foundation stock in my breeding efforts:

• Helianthus ×laetiflorus 'Lemon Queen' (2n=6x) - heavily branching with exceptionally heavy flower production of pale lemon yellow flowers.

• Helianthus maximiliani PI 592333 (2n=2x) - from wild seed collected in Manitoba, Canada.  I kept only one seedling from this batch which was consistently very early blooming compared to its siblings as well as compared to H. maximiliani seedlings from the other wild populations.

• Helianthus tuberosus 'Nora' (2n=6x) - early flowering type with clumping tubers.

• Helianthus annuus PI 650738 (2n=2x) - exceptionally large seeded selection from the USDA collection. Single headed, non-branching. Note: The following pictures depict this accession in a severely stunted state of growth as I planted the seeds very late and kept them in pots to ensure they would be sufficiently delayed to have bloom overlap with perennial types as well as for ease of moving around to facilitate pollination as needed.

• Helianthus annuus 'Primrose' PI 490319 (2n=2x) - dwarf plant with lemon yellow flowers. Some branching.  Small seeded.

• Helianthus annuus PI 649797 (2n=2x) - mixed colors claimed to range from lemon yellow to purple(?), either solid or with color rings.  However, grow out revealed that the base color is golden yellow (often referred to as orange) rather than lemon yellow with many individuals carrying various concentrations of red pigment over the golden yellow base color.  Unsurprisingly, none approached "purple" in color.  Multi-headed.  Small seeded.  This will be the first breeding material I have acquired that carries the gene for golden yellow/orange which is believed to be recessive to the normal shade of sunflower yellow just like the lemon yellow is recessive.  

• Helianthus annuus 'Garden Statement' (2n=2x) - produces a double layer of pale yellow ray flowers giving the heads a full appearance.  Seed size appears larger than wild types, but still much to small for use as an edible seed for snacking.

• Helianthus pauciflorus ssp. pauciflorus 'Homanic's Childhood' ('Miss Mellish'?) (2n=6x?) - double layer of long yellow ray flowers averaging 40 per head giving a very full appearance.  This clone is possibly the same as the European cultivar called 'Miss Mellish' (RHS Award of Garden Merit recipient).  I originally recieved this as H. maximiliani, but determined due to its leaf veining that it was closer to H. giganteous.  However, as I've gotten more first hand experience as well as book smarts I've come to the conclusion that this is quite certainly a selection of H. pauciflorus ssp. pauciflorus.

• Helianthus salicifolius 'Low Down' (2n=2x) - a reduced height selection with great flower production. died

The following are seed batches being grown out with more culling needed before individual select clones can be listed:

• Helianthus decapetalus PI 649970 (this species includes both 2n=2x & 2n=4x populations) - Population from Québec, Canada. Interested in crossing with a various H. annuus selections to produce H. ×multiflorus type hybrids with greater variability than existing selections.  I have individually potted ten different seedlings from this batch to evaluate individually for their utility in breeding.  I noted variability in receptiveness to setting seeds after pollinated by H. annuus, but due to growing the original batch in a community pot I was not able to identify individual seedlings at that time.  Now that they are potted individually I will better track any which perform better in their own right and any which are more receptive to hybridizing attempts.  Seedlings numbered as HDF004, HDF005, and HDF008.  Note: In looking for info on whether populations in Québec, Canada were diploid, tetraploid or both I found only one reference that indicated a tetraploid specimen had been sourced from Québec, Canada.  I found no reference to any diploid specimens (or other specimens of known ploidy for that matter) coming from Québec, Canada.  
• Helianthus giganteus PI 547178 (2n=2x) - From a wild population in Wisconsin, United States containing some individuals with purple leaves as well as pale yellow flowers.  This batch has all had uniform flower color, but with some variation in petal shape.  The stems are suspiciously smooth and waxy so guessing it may be an intergrade with H. grosseserratus.  Remaining seedlings numbered as HGF004.  Overall the rhizomes are quite compact and will probably be quite garden friendly without unwanted spread.

• Helianthus grosseserratus PI 547186 (2n=2x) - From a wild wild population in Wisconsin, US noted as having thick purple stems and compact leaf arrangement.  Bloom color is a rich golden yellow.  The rhizomes of each seedling were a mix of short and long with most clustered near the original stem and a couple extending quite far in each cluster.  Culled all but one (seedling number HGRF003).

• Helianthus occidentalis PI 664646 (from Ohio, US), PI 673322 (from Illinois, US), PI 673323 (from Illinois, US) (2n=2x) - colony forming species with leaves low to the ground and floral stems mostly without leaves.  Would be very interesting to cross these three populations to select for garden worthy traits such as heavy flower production of the showiest flowers.  Experimental hybrids with this species could also prove very interesting due to how much it differs from most commonly grown sunflowers.

• Helianthus radula Ames 33054 (from Alabama, US) (2n=2x) - a very unique sunflower typically blooming without ray petals and with unique pigmentation.  Would be an interesting wild card to throw into hybrid attempts to see what sort of genes could get unveiled through novel combinations.  Also tried PI 664736 (from Georgia, US) and PI 673173 (from Mississippi), but they failed to overwinter.  The surviving Ames 33054 is not looking great and after two growing seasons still has not bloomed.

• Helianthus annuus 'Mongolian Giant' (2n=2x) - large flower heads and very large average seed size.
• Helianthus debilis 'Vanilla Ice' (2n=2x) - very pale yellow

The following is a list of Helianthus species, hybrids or selections which are of interest to acquire for use in planned crosses in the years ahead:

• Helianthus mollis (2n=2x) - colony forming with attractively fuzzy appearance.  Known to be hybridizable with H. occidentalis and H. giganteus (therefore probably also possible with (H. maximiliani and H. grosserratus).  This species tends to produce "partial hybrids" when crossed with H. annuus. 

Current list of hybrids based on my breeding efforts:

• TNxAPF101 (H. tuberosus 'Nora' × H. annuus 'Primrose' PI 490319) - Earliest bloomer by far in this batch.  First year seedling produced abundant tubers; far more than needed for re-planting. They were of good size and tightly clustered both to each other and to the original stem.  It failed to set any seed despite producing many flowers, but this may just be a numbers game since its perennial parent typically has very low seed set. 
  Taste: (tasted in February 2023) Mild carrot flavor and sweetness.  No undesirable flavors detected.

• TNxAPF102 (H. tuberosus 'Nora' × H. annuus 'Primrose' PI 490319) - Nice (but not abundant) looking tubers close to original stem.
  Taste: (tasted in February 2023) Mild flavor with little sweetness.  No undesirable flavors detected.

• TNxAPF104 (H. tuberosus 'Nora' × H. annuus 'Primrose' PI 490319) - Very smooth tubers on short rhizomes and not clustered to each other at the base. 
  Taste: (tasted in February 2023) Mild with pleasant sweetness.  Maybe even a touch juicy?  No off flavors detected.

• TNxAPF106 (H. tuberosus 'Nora' × H. annuus 'Primrose' PI 490319) - Slightly knobby tubers, but not enough to make cleaning difficult.
  Taste:  (tasted in February 2023) Mild, sweet, juicy and actually refreshing!  Delicious!  Crisp, but seems less dense than its siblings so I suspect that it has a greater ratio of moisture to dry matter compared to them.

• LLQxALSF202 (H. ×laetiflorus 'Lemon Queen' × H. annuus PI 650738) - This one has a bit of a paler yellow color than its siblings.  Although less attractive in color, lighter may come in handy if crossed with a red pigmented form in the future.  Most notably this seedling has reflexed ray petals which tilt towards the stem instead of straight out or slightly cupped as is more common in sunflowers.  This trait could be very interesting to work with moving forward.

• LLQxALSF205 (H. ×laetiflorus 'Lemon Queen' × H. annuus PI 650738) - 

• LLQxALSF209 (H. ×laetiflorus 'Lemon Queen' × H. annuus PI 650738) - This may be my favorite from this F2 batch.  In addition to having lots of buds, the ray petals are wider than its siblings giving it a fuller appearance. 

• 

• LLQxALSF210 (H. ×laetiflorus 'Lemon Queen' × H. annuus PI 650738) -  Among the better of the F2 seedlings of this cross, this one has notably side facing flowers vs. some of its siblings with up facing flowers. 

• LLQxGF301 (H. ×laetiflorus 'Lemon Queen' × H. pauciflorus ssp. pauciflorus) - I really liked the compact rhizome structure of this one.

• LLQxGF302 (H. ×laetiflorus 'Lemon Queen' × H. pauciflorus ssp. pauciflorus) - Like its sibling I really like the compact rhizome structure from this one.

• LLQxSELF101 (H. ×laetiflorus 'Lemon Queen' × Self) - 

• LLQxSELF103 (H. ×laetiflorus 'Lemon Queen' × Self) - This one has deeply pigmented purple stems which I love!  In fact the pigment bleeds through to the edges of older leaves and although I'm not sure how consistent this trait will be it does look stunning.

• PHCxDVIF1 Helianthus pauciflorus ssp. pauciflorus 'Homanic's Childhood' × H. debilis 'Vanilla Ice' - I have not seen any references to these to species crossing before so this is exciting!  I got a fair amount of seedlings.  As of mid August only two were well on the way to forming flowers while the rest had zero sign of bud initiation.  In addition to the two I saved with buds I saved one without buds because it differed from the others I culled in that it was not making long rhizomes while the others were.  Below are comparison pictures showing the bract differences between the hybrid and the H. pauciflorus parent (which has the more compact bracts around the flower head).

• MEFxGF101, 102 Helianthus maximiliani PI 592333 (early flowering clone) × H. giganteus PI 547178 - Out of many seedlings I culled down two two which stood out as slightly more floriferous, though variation overall was not significant.  These all were slighly more similar to their seed parent than their pollen parent in height, bloom time (though not quite as early) and overall appearance, but notably did show increased bud count from their pollen parent.
• MEFxGRF101 Helianthus maximiliani PI 592333 (early flowering clone) × H. grosseserratus PI 547186 - A single seedling on which the flower color and shape looks very much like its pollen parent, but with an intermediate plant height between both parents.

• TNxDVIF1 Helianthus tuberosus 'Nora' × H. debilis 'Vanilla Ice'  - I was only able to produce a small number of seeds from this cross, of which one has germinated so it appears to be a viable cross!  To my knowledge H. tuberosus and H. debilis have not been crossed with each other before so my seedling may be the first time these two species have ever combined genes with each other.  Destroyed by a bird...  It's amazing how persistent this bird has been returning to my greenhouse two years in a row to eat my seedlings.  Fortunately the other crosses which were affected had extra seedlings to spare, but unless another seed ends up sprouting from this pot this cross may be a fail for this year.  This bird is like those people who pay a lot of money to eat endangered species...   

• LLQxTNF1 Helianthus ×laetiflorus 'Lemon Queen' × H. tuberosus 'Nora' - One seedling produced from this cross.  So far I'm not seeing any obvious traits from H. tuberosus so hopefully it's not just the result of self pollination.  I'll know soon when it blooms.

• LLQxDVIF1 Helianthus ×laetiflorus 'Lemon Queen' × H. debilis 'Vanilla Ice' - One seedling up. 

• DVIxOP Helianthus debilis 'Vanilla Ice' × OP (including H. annuus PI 649797, H. annuus 'Garden Statement' and self as pollen sources)

Seeds produced in 2023:

• MEFxGRF101 × MEFxGF101/102 - These should combine traits from H. maximiliani, H. grosseserratus and H. giganteus.  
• LLQxALSF209 × siblings + OP
• H. annuus red mix - 2nd and 3rd generation crosses between PI 649797 (red), 'Garden Statement' and 'Primrose'.  Culled all standard yellow seedlings with no red pigment and allowed open pollination between the best individuals with red pigment, but also one pale yellow individual with no red pigment because it had such nice form with great branching and flower production.
• LLQxSELF × siblings

Hybridizing goals for the 2024 season:

• Continue crossing annual forms to develop a strong red breeding line for use in later crossings to perennial lines
• Possibly cross one of the LLQxSELF clones with paler flowers than the original LLQ with red flowered H. annuus  
• Cross H. tuberosus 'Nora' with H. ×laetiflorus again, but this time use 'Nora' as the seed parent.

Overall Breeding goals:

• Reliably perennial
• Tightly clumping - not aggressively spreading
• Attractive and overall healthy appearance
• Highly attractive to pollinators
• Select an interfertile subpopulation of clones that produce large enough seeds to be worthwhile as a shelling seed for snacking.
• Select clones with large enough flower heads that they could aesthetically pass as annual sunflowers.
• Select heavily floriferous clones with a long bloom time.
• Select highly ornamental clones with varying flower colors.
• Select clones with varying statures including dwarf, medium and tall.
• Identify novel traits which may not have been visible in parent plants.
• Identify clones with good tasting tubers (not a high priority as there are already excellent tasting clumping selections of H. tuberosus, but while there is not much room to improve upon existing selections in that regard it would still be good to know which clones selected for other traits ALSO happen to make good edible tubers).
• Fertility.  Selections with large enough seeds for shelling will also need to produce full heads of seed without requiring hand pollination to be practical. 
• Leafhopper resistance!  I've noticed that leafhoppers seem attracted to most (if not all) of my Helianthus.  However, some of the plants react poorly to the leafhopper feeding on their developing leaves as they become very misshapen and unattractive.  Meanwhile others simply don't react and continue to develop normally shaped leaves.   

Notes on Sunflower Genetics:

  There has been a fair amount of research into the genetics of cultivated annual sunflower (H. annuus) including color of ray flowers, shape of ray and disc flowers and color of disk flowers as well as many other traits of less immediate importance for my breeding work.  This research should to some extent be applicable to hybrids between H. annuus and various perennial sunflowers, but due to some perennials being of higher ploidy than the annuals, the inheritance of dominant and recessive genetic traits is complicated.  For example if a dominant gene is paired with a recessive gene the dominant would usually express, but if there are three copies of the same recessive gene paired with that one copy of dominant gene it may not be able to fully suppress the otherwise recessive trait which appears to result in expression of traits intermediate between the simple dominant or recessive trait.

  Some basic traits for diploid H. annuus follow:

• Standard yellow ray flower color is dominant
• Pale lemon yellow ray flowers is recessive, but there is debate over whether it is controlled by a single gene or two separate genes both needed to be present together. Based on my observations I'm leaning towards it being controlled by a single recessive gene.
• Orange ray flowers are recessive to standard yellow
• Anthocyanin contributing a red-purple color in ray petals is considered dominant in that if the corresponding gene is interited from only one parent the anthocyanin will be present in ray flowers.  However, it will not fully cover the base color of lemon yellow, yellow or orange.  Instead it will express in the base portion of each ray petal (to varying extents - even varying between flowers on the same multi-headed plant), which results in two-tone flowers.  When the gene for anthocyanin in ray flowers is inherited from both parents the offspring will express a greater concentration of anthocyanin in the ray flowers to the extent that they entire color may appear red in varying shades.
• Purple disc flower color is dominant over yellow disk flower color
• Branching growth form is controlled by genes on three different chromosomes.  This could get infinitely complicated at higher ploidy levels.
• Stem diameter and height are controlled by genes located on many different chromosomes which would explain the incredible size diversity already present in cultivated diploid annual sunflowers.  Higher ploidy populations would get rediculously difficult for predicting outcomes among offspring, but fortunately if these are selectively maintained as perennial genotypes clones of desirable sizes could simply be clonally propagated even if their progeny would likely be wild cards in terms of size.
• Flower head diameter is very complicated with genes controlling expression being found on over a dozen different chromosomes
• Ray flower number and length are controlled by multiple genes. 
• Domesticated lineages of annual sunflowers have a reduced self-incompatibility for pollination meaning that they are more likely to set some seed without cross pollination than wild types.  This is also controlled by multiple genes, but overall appears to be somewhat recessive so producing self pollinating hybrids with perennial species would likely require multiple generations of crossing between hybrid siblings before any individuals are found to express this trait.  While complicated, this trait would likely be automatically selected for simply by breeding individuals with consistently higher seed counts per head as various degrees of self compatibility would ensure a higher rate of pollination and seed set.

So far I've already observed that despite there not being a genetic precedent for intermediate color between lemon yellow and standard yellow in diploid sunflowers I was able to produce an intermediate color when crossing diploid yellow with hexaploid lemon yellow.  The offspring appear intermediate in color concentration.  I hypothesize that this is because the standard yellow genetics from the diploid parent were only able to cancel out one set of the lemon yellow genetics from the hexaploid parent leaving the other two sets of lemon yellow genetics from the additional ploidy unaffected thus producing ray flowers with more yellow concentration than "lemon yellow", but less yellow concentration than "yellow".  While on the one hand higher ploidy counts do complicate breeding by making it far more challenging to predict breeding outcomes on account of having multiple copies of genes for each trait, in practice an advantage may be that higher ploidy could result in greater incidence of intermediate expressions between traits.  This could potentially lead to the development of higher ploidy perennial clones which express all manner of intermediate colors, shapes and sizes which otherwise might not be found in diploid predecessors.  In other words, higher ploidy breeding lines will be a headache for those with specific breeding goals to achieve, but a blessing to those who are breeding just to see what variations might pop up.

Retired/Culled: 

Note: Many seedlings from my trials are not worth keeping record of after culling.  Therefore I will primarily use this cull list for keeping track of plants which had been kept long enough to serve as breeding material prior to culling and since they are part of the pedigree of other seedlings which may end up being kept.

• Helianthus ×laetiflorus 'Lemon Queen' × Helianthus annuus PI 650738 (LLQxALSF102) (4x probable) - Medium vigor with prominent terminal flower and smaller secondary flower buds held on upper nodes just below primary flower, but without any significant branching.  Reddish-purple pigment on stem.  The top few flower heads emerged deformed/lopsided; lower buds forming normally so it have have been due to top three buds being very crowded against each other early in development.  Flower heads open facing sideways which is a great viewing angle.  Some flowers were removed, but of those allowed to mature some viable seed was set.  A moderate number of overwintering tubers were formed, but they appear to be long enough that it could be an aggressive spreader if planted in the open ground.  I do like some of its traits and so have saved and planted seeds from it to search for superior variations. 

• Helianthus ×laetiflorus 'Lemon Queen' × Helianthus annuus PI 650738 (LLQxALSF101) (4x probable) - Medium narrow lance shaped leaves.  Vigorous with branching and flower buds produced from top to bottom of main stem.  Ray petals intermediate in color between parents (soft yellow).  Appears to set a fair amount viable seed even via open pollination, but with no visible difference in size than its perennial parent.  Produced a large number of compact overwintering rhizomes.  Interestingly, this seedling as well as two of its runty siblings all sent up small leaves at ground level in mid-to late winter.  This is not a trait that I have observed on the perennial parent which seems to wait for a bit of warmth before leaves emerge.  We actually endured multiple cold snaps after these basal leaves emerged and they never seemed to suffer any cold damage. I liked this one the first year, but in its second year it sent up so many shoots that it self-crowded to the point of each stem only producing a few buds.  Additionally, it became obsolete on account of a fair number of its offspring showing further improvement in the traits I'm selecting for.