Improving heirloom crops by selection (legumes, grains)

Question

Short version

I want to know of methods (and devices) of selecting the biggest/heaviest bean and grain seeds from my garden, to re-sow the next season. I would like to hear from people that have done this, how they went about this, and if the results were satisfactory.

Longer explanation

I have a few different crops of legumes (green beans, mung beans, peas - might try chickpeas, lentils etc. in future) and grains (maize, might try quinoa in future) that I have grown from heirloom seeds. The main idea with heirloom seeds is to have a perpetual cycle since one saves a few seeds at harvest, to sow again the next season.

Obviously, heirloom seeds are originally obtained from a commercial supplier or, better still, friend - which means the seeds were originally produced elsewhere where the climate, soil and other conditions may have been slightly different and perhaps your own garden might now be slightly less optimal. (I have still obtained satisfactory results until now, so this is not a big worry.)

On the other hand, it should be possible to improve your crops for your own local conditions by selecting seeds with the most desirable traits for resowing the next season, for at least a few generations. (My grandparents and parents were commercial papaya growers that grew their own seedlings for each year's plantation, so I have known about the concept - and practical application, even if not very advanced as judged by modern scientific methods - from an early age :-). )

With some crops (e.g. tomatoes and squash families) I would probably look at the fruits yielded and then take seeds from the most desirable ones. However, for legume and grain crops the part you eat is exactly the same as the seed, so my thinking is that the bigger the seed size, the more desirable.

Obviously, one would not take seed from only one fruit or one plant. Having seeds from a nice overall selection might help to maintain some genetic diversity.

With all this in mind: If I have a container of seeds (maybe 200 - 1000 seeds in total) what is a good way to separate and sort them by size? So that I can keep the biggest (say) 20% and eat the rest. Spreading them on a towel and eyeballing them, picking out some with tweezers, seems to be cheap, but very time consuming, labor intensive, and not very exact. Any better suggestion?

---

Added: I've also recently sprouted mung beans (for eating) and noticed that some kernels did not germinate; those seemed to be smaller than the successful ones (not a pleasant surprise on the teeth...) So perhaps this might be a consideration for removing the smallest (not completely developed) seeds from a harvest too.

Answer 1

Selection

Remember to consider a whole-of-plant perspective when saving seed, as the seed determines the success of the whole plant. If the end goal of a grain crop is the maximum yield for a given area, then selecting based on the size of individual seeds may not give an optimal result.

Growing

You might try to maximise the total weight of seeds produced by a single plant, though a greater number of lower yield plants situated more closely together may do just as well or better. This could lead you to select based on yield from a standard sized plot, where the plants in each plot are grown from seed taken from a single ear per plot.

Other Characteristics

The above suggestions are really just examples, as breeding for drought resistance, disease/pest resistance, ease of harvest, etc can all be more or less important depending on your particular circumstances (though quantity and quality of yield may be the ultimate end goal anyway).

Personally, I do save some seed, but I don't have the time and energy to try to do it perfectly, so I just try to ensure the seed comes from a reasonably productive and healthy plant.

Also, if you do want to sort seeds by size, you could try putting all the seeds in a container and gently vibrating it. Larger ones will probably tend to rise to the top.

Answer 2

I would recommend some reading up on 'landrace' crops.

(from wikipedia with a source)

Landrace plants are grown from seeds which have not been systematically selected and marketed by seed companies, nor developed by plant breeders. The label landraces includes all those regional cultigens that are highly heterogeneous, but with enough characteristics in common to permit their recognition as a group.[citation needed]

This includes all cultigens cultivated without any specific nomenclature and value. A landrace identified with a unique feature, and selected for uniformity over a period of time for maintenance of the characteristic features of the population, can evolve into a "farmers' variety", or even a modern cultivar as in many crops (for example, Cajanus cajan 'Maruti' in the case of pigeon peas).[21]

Conversely, a modern cultivar grown over time can "evolve" into a landrace, especially when self-seeded and some human selection is applied.

Friis-Hansen, Esbern; Sthapit, Bhuwon (eds.) (2000). Participatory Approaches to the Conservation and Use of Plant Genetic Resources. Rome, Italy: International Plant Genetic Resources Institute. p. 199. ISBN 92-9043-444-9.

Landrace is typically a term applied to animals, but the principal remains the same here.

By allowing your crops to continually reproduce, and use all the seeds not just the ones you like, you develop genetic diversity, where the selective pressure is applied by the soil conditions and climate unique to your location. You might throw away a small seed, but that seed might be resistant to disease, or adapted for drought. I'm not familiar on the flavor aspects of this, perhaps someone else could weigh in.

Highly Irregular is correct, that you should consider more than just seed size in you selection.

This article may have some info http://gardenseeds.org/landrace/