There are a variety of crazy notions that are very, very, very unlikely to go anywhere. But, there is at least a glimmer of crazy hope that they might do something interesting. Here are some. I’ve written them up with details applicable to wheat, but the concepts can probably be applied with equal unlikeliness to other crops.
Breeding perennials from annuals without crossing with perennials
I would grow a few thousand seedlings, really densely so they do not take up much space. Then during a period beginning before senescence starts, and continuing until it is well underway, chop down a different group of seedlings each day (or several a day).
The ones chopped down before senescence starts should all sprout back The ones chopped down after senescence has started should all die. Now, the fantasy is that some small fraction of them would have some degree of fuzziness in their triggering of senescence, just enough so that if it was disrupted just right the leaf senescence would continue while senescence of the crown would not.
If there *is* any kind of fuzziness like that, it could be selected for. And then, just maybe, that fuzziness could be coaxed to open up wide enough to allow to start seeing this perennial behavior happen without the interventions.
And there are plenty of other factors that effect the senescence process (here’s a good review, another interesting paper about senescence in corn, and a nifty page on wikipedia), and I might try juggling various other kinds of shocks to see if I can in any way halt the senescence. Senescence is often triggered by stresses: shade, heat, drought, lack of nutrients, pests, etc. This is because if the plant is in danger of being consumed or destroyed, it is much better for it to shunt nutrients to the seeds away from the rest of the sinking ship. So maybe before senescence I could turn the lights down, stop fertilizing, simulate pest pressure, and increase the heat – then once senescence is starting, or just about to start, I could give more light, more fertilizer, stop pesting the plants, and give a more comfortable temperature.
The reason this would be marvelous is that the current approach of getting perennial behavior from direct domestication of or wide hybridization with wild relatives involves a very large tradeoff in many of the agronomically desirable quantities, like yield, flour quality, and threshability. A paper studying over 150 wheat x wheatgrass crosses found that the only perennial lines had *all* of the chromosomes of the wild perennial wheatgrass. If any of those chromosomes were lost, so was the perennial trait.
This means that some of the limitations of current perennial wheats may be very, very difficult to fully breed out, because a significant chunk of the DNA in the resulting hybrids is wild, and essentially needs to be domesticated. We could possibly get around that by starting with a high-performing annual wheat and unlocking latent perennial capacity in its chromosomes, rather than steamrolling over that latent capacity with active perennial capacity from a wild relative.
It is possible to graft grasses. In apple trees, some small quantity of genetic material can pass from rootstock into the grafted material, and maybe grasses are the same. If that is the case, and if that small amount of genetic material made it into the seeds, it might make some interesting results.
This is less of a moonshot because it is crazy, more because it adds another step. Secale montanum (wild perennial rye) is a very interesting perennial species that already has fairly large seeds. A perennial triticale could be developed by crossing triticale with either Secale montanum or one of the perennial rye lines developed from it. The perennial trait from that triticale could then be brought over to wheat.
The reason to try this is that rye is phylogenetically closer to wheat than the wheatgrasses that are normally used as sources of perennial germplasm for wheat. This could potentially reduce the work of re-domestication that would be required to get a usable perennial wheat.