Originally, I planned to complete preliminary work on a bunch of different new crops that I began breeding seriously this year so I could write up one big article on each species, but I think progressive updates on all of them is a more useful way to show the mindset and strategies needed to feel your way into the unknown. If you missed the start of this work check out the Four Flower Gauntlet.
The first crop to report on is the sword bean. In autumn 2023 I hybridised three different species, and in spring 2023 I planted out those F1 seeds. They grew into a wall of green on the bamboo trellises, flowered profusely, and podded sporadically while the pod sucking bug pressure was high over summer. Now the weather has cooled they have started podding more heavily, and seed should be ripe by late winter. It is not uncommon to see lower fertility levels in wide hybrids, so I am happy with the amount of seed that is forming. Breeding crops for seed production is easy since any plant which produces more seed will leave more offspring if you pool all the seed together. I planted three trellises, each with seed from one of the three mother species (but with unknown pollen parents) and slashed the vines that tried to climb between them, with the hope of keeping track of female parentage. The vines got away from me in autumn, so it looks like I will just be throwing all the hybrid seed together and selecting based on seed traits and production levels. Their genetics will be all over the place for a few generations, so there is no point applying too much selection in the next generation. I plan to cover as much of my weedy vegetable garden as possible with bamboo trellises this spring to do a mass planting of F2 seeds.
Crop number 2 is the rosella (Hibiscus sabdariffa). I have grown two distinct strains for years and wondered why I never saw spontaneous hybrids between them. This year I started bagging and emasculating flowers, though autumn was so consistently rainy it was almost impossible to get a window for this work. I managed to emasculate, bag and hand cross the two strains but all three tries failed. I also did the same with a few more flowers to wide cross to other Hibiscus species, but they all failed as well. This made me step back and question my method. Emasculating these flowers required scraping away the masses of stamens on the sides of the ovary, with the damage causing a fair bit of sap release. Perhaps this was causing the flowers to drop, regardless of pollination status. So I did a follow up experiment where I emasculated a few flowers, but then hand pollinated with pollen from the same strain. One of these failed (since the stigma didn’t mature properly) but the other set a pod, so it looks like emasculation is sometimes damaging the flower. There could be another way to prevent selfing, or maybe it isn’t even an issue. To test the latter possibility I also tried bagging a flower without emasculation. This was to check if emasculation was even necessary, since the stamens are quite far away from the stigma and no obvious self pollination mechanisms exist (some flowers move their stamens to the stigma if they aren’t pollinated by a certain time as a back up plan). Unfortunately I didn’t keep track of the result, so I will have to repeat this experiment next year. Negative and positive controls can be really useful, especially before you invest hundreds of hours into hand pollination. At any rate, the season was mostly a bust due to the dry spring and late start, followed by constant rain during the flowering period. I will have to try again next year, but my follow up experiments have given me hints to improve future efforts.
Crop number three is the lima bean. These have been a complete bust. Flowers proved impossible to emasculate or hand cross, though the latter might be possible next year with practice. The trellises are now overgrown with cucamelons and neighbouring sword beans. I will plant more blocks of these next year away from such competition and try again.
Crop number four is quinoa and huauzontle. I got a handful of new varieties this year, including a few grexes, which can often kickstart local adaptation and breeding efforts. Each strain was bulked up by planting a few seed in a small pot, then transplanting the seedlings without disturbing the rootball into bigger pots for grow out. This approach avoids wasting large pots on seed might be dead on arrival (only one variety fell into this category this time). I transplanted a few spare seedlings into my veggie garden, but they drowned during the nonstop rain. I also sowed a few seed of each variety into tube pots, repeating the process every week or so. This method forces the seedlings to flower at a relatively small size, and staggers their pollen production and stigma receptivity for easier direct crossing without turning to pollen storage. I also followed the method of a research paper and removed the terminal male buds of all the quinoa seedlings. This meant the basal female flowers could only be pollinated by untrimmed huauzontle seedlings. The successively sowed seedlings matured at smaller and smaller sizes, a consequence of their daylight sensitivity being incompatible with my latitude. Quinoa sowed at the very end of summer flowers at about 40 cm tall, while that sown in late autumn only grows a few inches. I plan to grow out half of this seed this winter to see if I can bulk it up (and keep an eye out for any daylength insensitive plants), so that next autumn I can plant out both F1 and F2 seed from crossing quinoa and huauzontle. Precociousness can be useful to grow multiple generations in one season. I might even be able to push out an F3 this spring if I am lucky.
The final crop on the list today is Salvia. I have grown a spectacular crop of chia (S. hispanica) this year. They loved the rain and grew up to 3 m tall (when supported by a legume trellis). This also meant I had a constant supply of chia pollen for many months, which I have put to good use. My straggly potted Salvia columbariae seedling has flowered on and off for months, so I have been crossing it with chia pollen at every opportunity (though I haven’t bothered emasculating the columbariae, which could be an issue since it is self-fertile). Next year I will grow more columbariae seedlings and remedy this oversight if the initial cross doesn’t work. The instinct to do the lowest effort experiment first, then investing more energy if necessary, can be useful with these kind of projects. I also have a pot of Salvia polystachya flowering. This perennial species is very close to chia according to genetic analysis, and I have a single self-infertile clone, so any seed produced should be hybrids. The calyxes are too small to be sure if any are forming, but the way the flowers have been dying off is suggestive of success. The calyxes are also holding very strongly, so fingers crossed.
I have a bunch of ornamental perennial Salvias in my old flower gardens at my parents’ house. I have been taking a bunch of chia flowers (picked before dawn so the bees don’t strip the pollen) and visiting flower spikes marked with coloured wire every second day. Salvia flowers release pollen on their first day, then the stigmas become receptive for a few days afterwards, so pollinating every other day works well enough. In many species the stigma pushes out of the protection of the petal tube as it becomes receptive. Salvia madrensis, with egg yolk yellow flowers, and involucrata in pale pink have showed the least signs of success. S. miniata with scarlet flowers produced a few seeds, though inspection of unpollinated spikes showed the occasional seed set. The seeds are already growing, so I should know soon if they are hybrids. Salvia elegans is showing a similar pattern, with the rate of calyxes swelling with forming seed much higher on the spikes I hand pollinated than on the rest of the plant. Finally the gorgeous magenta S. iodantha is also showing signs of taking the chia pollen. If I manage to get hybrid seedlings from at least two of these species I will be very happy, and this will set me up for creating F2s and back crossing to chia next season. A perennial seed crop Salvia could be interesting in our climate, especially if it is tall enough to compete with weeds with minimal assistance, though diversifying annual chia would also be a good outcome.
So that is about all I have been up to on the crop breeding front. Hopefully this array of projects in motion will give you a sense of the struggles and inevitable failures that come with the territory. If you are thinking of going down this path I recommend starting 3-4 projects in parallel, with the expectation that only one of them is likely to lead anywhere. Experimental crop breeding is a lot like tai chi, where tackling a barrier head on is a waste of energy. Usually there is another roundabout route to a comparable end goal. Flow like water and find existing cracks to widen. Drift like pollen in every direction and eventually you will land on a receptive stigma.
When crossbreeding, is it important which plant is the female or male of the cross (other than that one way may not take)? Or should a female quinoa and male huazontle produce the same result as a female huazontle and male quinoa, assuming they both fertilize successfully?
When crossing a perennial and an annual, how hard is it to get a now perennial version of the annual? I didn’t realize this could be done at all. Are the traits determining annual-ness relatively unimportant to the rest of the plants health and production? Or are you likely to get some kind of biennial mix that then must be selected further?
I planted sword beans this year- first time I’ve ever heard of them. From
What I’ve read they are complicated to eat. I ate a lot of them young green pods as I googled that this was okay. Do you eat them green or just eat the dried bean?