August 2010 Archives

Thanks to the recent success of Syngenta Flowers' outstanding pelargonium variety 'Calliope Deep Red', there is renewed interest in interspecific pelargoniums.  Breeding interspecific hybrids in any genus is challenging, and pelargoniums are no exception to this rule.

Here at GardenGenetics, our pel breeding is focused on what an artist would call "the negative space".  Because we are new, and because we do not have our own production capacity, we cannot effectively compete with the established pel breeders from around the world.  Those breeding companies occupy "the positive space".  G2 is breeding for the unoccupied market niches --- more precisely, the under-occupied market niches --- "the negative space".  This focus has us working in pel novelties, and in particular, with pel interspecific hybrids (ISHs, for short). 

Breeding ISHs is always a challenge.  Making the initial crosses can be arduous.  Nicking is often quite poor, resulting in low seed set.  Pod abortion is frequent (although embryo rescue is always a possibility if you've got access to a TC lab).  Endosperm development in the ISH seed is frequently abnormal, giving you shriveled and deformed seed.  Shriveled seed, of course, may or may not germinate, but then again, plump ISH seed simply may not germinate either.  If ISH seed does germinate, germination is likely to be poor.  Seedling vigor may be weak.  ISH seedlings may emerge and then just die. 

But if you persist, sometimes you succeed.  Here is an ISH seedling between P. xhortorum and P. tonganense.  Not a bad looking plant, and quite abundant in its flowering.  But from a cutting, it is slow to establish.  In flower, the florets are small, and borne upon relatively flower stems.  Like all single-flowered pels, the florets shatter and in this particular ISH, leave you with an abundance of spent stems. 

From a breeding perspective, this ISH also displays some of the classic problems with ISHs.  It is mostly pollen-sterile.  The pollen is creamy-white rather than yellow-orange, and the individual pollen grains are mostly shriveled and unfilled (and don't absorb the typical cytological microscopy stains). 

This particular ISH is mostly ovule-sterile, too, although if you make enough hybridizations, you may eventually get lucky and recover a few viable seeds.  We are now in the 3rd BC generation (to xhortorum) out of this ISH, and have just now begun to recover some fertility.  Pollen production remains weak, although most of the BC progeny are now adequate seed parents. 

Why are ISHs typically sterile?  Well, when you try and combine two different genomes, even if you successfully get them to combine (i.e., pollen from one species successfully fertilizes the ovule of a second species), subsequent cell division may not occur normally.  Why?  Because cell division, whether to create duplicate cells (i.e., mitosis) or germ cells (i.e., pollen or ovules), individual chromosomes need to double, pair up, and divide precisely.  In an ISH, it is typically chromosome pairing which is the problem.  Improper pairing occurs, and as the improperly paired chromosomes try to separate during division, things go awry.  Pieces of chromosomes disappear.  Sometimes whole chromosomes disappear.  The usually highly ordered sorting of chromosomes into the product cells does not always occur, so that the product cells may have extra pieces of genetic material or may be missing important pieces of genetic material.  As you might imagine, this genetic chaos can have a very negative effect on the survival of one of the product cells. 

Here's a shot of one of the progeny from a BC3 to xhortorum.  It is beginning to resemble xhortorum, but some of the classic characteristics of tonganense are still apparent.  Lots of flowers, small florets, single petals, tends to shatter and generate that forest of spent flower stems.  Somewhat lax (trailing) habit.  Many growing points, generating a plant with heavy basal-branching. 

This plant's heavy-flowering, many-budded, many-branched characteristics make it an attractive parent plant --- if you can get it to make viable seed or pollen. 

Here is a seedling of the BC2 (previous generation BC to xhortorum) but crossed to diploid (2n=2X) P. peltatum.  This plant has serious problems with chlorophyll synthesis.  We think that the problems are mostly a function of genomic incompatibilities (the chromosomes between xhortorum and peltatum just don't match up very well).  A branch of this plant may eventually "sport" and become more vigorous.  This plant --- currently very sterile both as a seed and a pollen parent --- may eventually make seed, if we attempt enough crosses onto it.  But this kind of abnormal seedling is very common when you make interspecific hybrids.  If there is future value in this seedling, it can only be if and when you take it to a future generation. 

Why might we want to continue working with such an ugly-looking specimen of a plant.  Here is a close-up including a flower from this interspecific hybrid. 

Plant breeding is fun stuff !

Sometimes, curiosity just gets the better of you.

This is Kristin, one of our 2010 summer interns.  Great young horticulturist.  Has done public garden internships at Longwood and Polly Hill.  Graduating from Penn State in December.  Looking for full-time employment thereafter.  Hint hint. 

What Kristin is doing is counting and classifying pollinators on Pentas lanceolata.  We had heard anecdotal tales about how many more pollinators were attracted to the older pass-around heirloom varieties of pentas than the newer seed-propagated varieties, but could find no real evidence let alone real-world data. 

Also, I'm part of the Center for Pollinator Research at Penn State, and a pollinator preference project seemed like something that could be interesting and productive.  So ... one of the many projects we threw at Kristin for the summer was to take some of the pentas plants which we had available, and begin to see if there might be something valid in the garden gossip.  This is the first year, and her data will only be preliminary.  If Kristin's results show that there might be something interesting, we'll set up a proper research study for 2011, complete with proper statistical design.  We'll get our campus colleagues to participate in the design phase, and perhaps bring in an intern from campus to run the larger study. 

It's not nearly as straightforward as I had originally expected.  Pentas has two reproductive flower forms.  A botanist would describe these forms as heterostyly.  Many of the pentas varieties in commerce are pin forms, in which the styles are long and exposed, and the stamens remain deep in the corolla tube.  The alternate form is the thrum form, in which the stamens are borne on long filaments, and the style is short so that the stigma is protected deep within the coroll atube.  Primula vulgaris is the classic example of pin and thrum forms, but the phenomenon occurs widely. 

Why the botany lesson? Well, if pollinators are more attracted to pentas pollen than to pentas nectar, then there may be a preference for the thrum (short-styled) form over the pin (long-styled) form.  A good study design would account for this possibility --- which means locating and maintaining thrum forms.  Unfortunately, these are not widely available in all pentas varieties.  So this year's study is preliminary, and should generate enough information to let us plan the next study a little better. 

But for now, K has a real interest in insects and pollinators, and is doing a great job collecting data.  Once we get the data worked up, and if there is anything interesting in it, we'll be sure to talk about it here. 

In July, 2008, we began to use our research greenhouses.  At last, we were in our own space. 

Two years later, and how are we doing?  Well, the greenhouses are full to capacity most of the year, and overflowing for a few months in the spring.  Our need for overflow capacity in the spring is the reason we are currently putting up a hoop house. 

We're quite happy with the greenhouses.  Our Nexus structure is simply awesome, as is the Argus control system. 

Would we do anything differently?  Of course we would.  I've never done a building project where you would answer that question any other way. 

We made the greenhouse bays 36 ft wide because that was the dimension we used for the very first rough sketch.  Looking back, we could have easily made them wider --- at least 42 ft --- gaining a significant amount of bench space with only a modest increase in cost. 

We would have installed insulated concrete block kneewalls on the perimeter, rather than taking the polycarbonate to the ground.  This would have added a few days to the construction time, and a modest amount to the overall cost but would have a significant impact on energy use.  We grow 90+% of our crops on benches --- light coming in the sidewalls below bench height is irrelevant. 

Actually, considering how good the drainage is on-site, we probably could have sunk the entire greenhouse structure into the ground to bench height. 

We would have brought in a greenhouse-experienced concrete floor specialist.  We used a very competent local firm, but this was their first greenhouse floor.  The floors do not always slope completely to the floor drains, and this can be annoying.  Yes, we poured the floors during the winter --- no real choice there --- and yes, much of the structure is on compacted fill.  But still.  We would prefer to have the floors drain more effectively. 

We could have built in more laboratory space from the start.  We could have built in a germination room, as well as a light room next to the TC lab.  We could have added a few more offices.  We could have included a walk-in refrigerator for seed storage.  But all of these things can --- and will --- readily be retrofitted into the existing facility.

All in all, we are very happy with our research greenhouses.  We could go on and on about the things we did correctly, but ... let it suffice to say that we're very happy. 

One of G2's first woody ornamental projects actually began before we started GardenGenetics.  I started collecting lilac germplasm in the early 2000s, with the intent of working with the genus in my backyard.  I came across the classic text Lilacs, by the late Father John Fiala, in a Timber Press catalog; bought it on a whim; and became thoroughly hooked on the amount of variation available in Syringa. 

To a breeder, variation is what fuels the passion.  A breeder cannot make progress without genetic variation. 

So, I began collecting germplasm.  And began working on breeding technique.  And began making some crosses, just for grins.  And then we started GardenGenetics in 2007. 

Now that G2 was a business, I had to develop realistic goals for each project, including lilacs.  Intense traditional "french lilac" fragrance.  Large florets within large flower clusters.  Single flatly open florets for maximum display.   Five clean colors in a reasonably matched series.  Compact plant habit, 5 to 6 ft high, 5 to 6 ft wide.  Tough leathery foliage for powdery mildew-resistance.  Perhaps burgundy fall color (along with purplish overtones on newly emerged leaves). 

My first serious crosses were made in 2009.  The seed was chilled and sown in the winter of 2009-10.  None of the seedlings bloomed in 2010, and were transplanted into one of our research "strips".  With luck, some of these will bloom next year.  Woody ornamental breeding tends to be slow. 

G2's "strips" are places in our fields where the limestone ridges  prohibit us from installing one of our standard quarter-acre plots.  If we can't do a plot, we try and put in a strip --- a five-foot spader-width strip 80 to 100 ft long.  We use strips for our perennial and woody projects, projects which occupy the same space for multiple years. 

In every seedling population, interesting things just happen.  For instance, we've got a variegated lilac seedling.  It behaves like it is chimeral, but if it is, it is a long way from stabilizing.  And, like most variegated plants, the growth rate looks to be relatively slow.  Still, it may make an interesting collector's plant.  Let's see it bloom first.   

We breed lilacs (and any perennial or woody species that we can) in the greenhouse in the winter.  Plants are potted up in the late summer from the field; overwintered at ambient temperatures in the hoop house so the plants can vernalize; and then brought into the greenhouses in February or March to bloom.  Pollination control is much better in the greenhouse; the timing of the projects fits into our schedule much better during the winter than in late spring (when EVERYTHING is blooming and needing our attention). 

Once again, woody ornamental breeding is slow.  First crosses in 2009.  First bloom of the F1s in 2011 (maybe).  With luck, first bloom of the F2s will be in 2013 or 2014.  Three to five years of selection and trialing --- first ready-for-market date is not going to be before 2018, and this assumes that there are superior individuals in the F2s. 

It is far more likely that the best genetic combinations will not occur until we begin making F4 x F4 crosses in 2017 or 2018.  These "F1 hybrids" should have significantly improved uniformity as well as significant heterosis (hybrid vigor).  But we probably won't see them until 2019 or 2020. 

If you find yourself here as a result of being contacted about the 2010 Independent Plant Breeders Conference, THANK YOU !

The 2010 IPBC is being co-hosted by GardenGenetics and Longwood Gardens.  The 2010 IPBC will occur at Longwood from 3 November through 5 November 2010. 

What is the IPBC all about?  It is a place for independent plant breeders to interact with each other; with potential industry partners and licensees; and to learn more about what the horticultural and green industries need in the way of new plant varieties.  Our intentions are to share and to network, freely and openly. 

We are actively seeking sponsors.  Please contact rickATgardengenetics.com for more information.