Isolation distances

[Joseph Lofthouse]

I'm inviting feedback on a new page I'm writing for my web site. What did I leave out? What is unclear?

Summary

Some books and articles have published data about "isolation distances" for seed saving. I believe that such recommendations are inappropriate for people growing open pollinated or landrace seeds.

The Theory of Isolation Distances

The theory goes something like this:

In order to keep a highly inbred variety "pure" we must isolate it by so many feet or miles from any other plant of that same species, because the variety would be ruined if pollen came in from somewhere else.

I believe the main failing of this theory is the idea that highly inbred crops are somehow desirable. Perhaps they have some utility to huge commercial operations, but "pure" crops are generally a bad choice for home gardeners and for small market growers. Highly inbred and genetically fragile crops are an especially bad choice for anyone that is interested in saving their own seeds, or developing their own varieties.

Inverse Square Law Inverse square law of pollination

Pollination approximately follows an inverse square law. What that means from a plant breeding perspective is that plants are extremely likely to be pollinated by their closest companions, and unlikely to be pollinated by plants from further away. As an example, supposing we plant a square foot of carrots as a seed crop, and our neighbor also plants a similar patch. If the distance between the two crops is 32 feet, there is about a 1 in 1000 chance that any particular seed will have been fertilized by pollen from the other garden. At a distance of 100 feet the odds are approaching 1 in ten thousand, and by a mile the odds are in the parts per million.

Pollination does not follow a perfect inverse square law, but it is a good first-approximation. This is basically a dilution process, the further away a pollen donor is, the less likely it is to contribute pollen to our seed crop. We can use this process in reverse, by producing an abundance of our own pollen it will overwhelm what little pollen manages to arrive from far away. Growing a large enough seed crop to maintain genetic diversity also produces enough pollen to substantially dilute what is coming in from elsewhere. To go back to the carrot example: If we increased the size of our carrot patch to 100 square feet, our neighbors patch could be expected to pollinate something like one seed in 100,000.

Open Pollinated Seeds

The term "open pollinated" implies at least the possibility of genetic diversity. It seems to me like a contradiction to speak of a "pure" strain of an open pollinated crop. I am only growing open pollinated crops, mostly landraces, so I don't care if I receive pollen from other sources. I propagate the best-growing most-suitable plants each year, so if stray pollen helps my crop grow better then I welcome it into my garden. If it causes something to grow worse then I don't select that plant for making seeds.

[mnjrutherford]

Passes spell check and the formatting is excellent. I do suggest centering the image, it's to the far right on my screen. That may be a Firefox issue though.

[spacecase0]

I love it,
very practical and easy to understand the point.
I bet everyone gets the idea
and the image is centered on my screen and I am also using fierfox

[12540dumont]

Joseph,
I think you need more than a square foot of carrot for diversity. I don't know if 40 carrots will fit in one square foot, and as carrots are outbreeders, you may want as many as 200.

We don't want your carrots to have inbreeding depression. I hate depressed vegetables, they're just too sad to be around.

It's very well written. Are you working on Joseph's Book of Breeding Landraces yet?

[MikeH]

Hmmm. While the inverse square relationship makes sense for insect pollinated plants, I'm not sure that it would apply to wind pollinated plants.

Regards,
Mike

[Joseph Lofthouse]

Sept 25, 2011 20:55:47 GMT -5 MikeH said:

Hmmm. While the inverse square relationship makes sense for insect pollinated plants, I'm not sure that it would apply to wind pollinated plants.

Any suggestions for a better model?

Ain't it the same general principal? Take a finite sample of pollen and spread it over a wide area and it gets really dilute really quick?

Eventually I'd like to incorporate a paragraph about the settling velocity of pollen. For example corn pollen falls at around 0.8 feet per second, so if it is released from a height of 6 feet it would take around 5 seconds to hit the ground on a calm day. In my garden in July the average wind speed is around 10 mph, so the average corn pollen grain could be expected to travel about 45 feet before hitting the ground, and after about 25 feet it would fall below the level where it could interact with silks. So in the case of wind pollination there are systems in place (gravity, plants acting as filters) that are actively removing pollen from the air and preventing it's dispersal.

If I could find a nice chart containing the settling velocities of carrot (or other) pollens I'd incorporate that into the page.

[12540dumont]

Joseph,
Leo wanted to say that carrots are not pollinated by bees, but small mouthed parted insects, so maybe another crop might be a better example?

Oh, we love this. And none of this criticism, but you know if you are teaching, you'd best use the best examples and give them the correct amount of square footage because, someone is going to follow your example to the letter

[Joseph Lofthouse]

You actually want me to do real math? Uuuugggh!!!! Calculus would handle these types of scenarios really well, but i never used it since I left school, and it's been decades already.

[Joseph Lofthouse]

An added paragraph:

Due the the rapid decline of pollen with distance, the pollen donors that have the most impact on a seed patch are those plants that are growing within the patch. If there is a rogue plant among the parents it will have a much larger impact on the descendents than a little bit of pollen coming in from elsewhere. The detrimental effects of non conforming parents can be reduced by careful selection of parents, by weeding out non-conforming plants during the growing season, and by planting larger numbers of parents. For example: I grow a radish landrace. I also have volunteer radishes in my garden. I can minimize the impact that those volunteers have on my seed crop by planting my parents in an area where I know that radishes did not go to seed the previous year. And during the radish bloom I can keep the area very close to the radish seed patch clear of blooming volunteers. For good results I don't have to keep the entire neighborhood free of feral radishes, just the area within and immediately around my radish bed.

[oxbowfarm]

Since carrots are the example here I'll pipe in with a twist on the question. I've never even attempted carrot saving due to the Queen Anne's Lace issue. My field is plowed out of a larger hayfield that is thick with QAL. I have QAL coming up in my lawn, you mow it and it will re-flower in two weeks. It's always been my impression that saving carrot seed would be a total waste of time due to the contamination of QAL pollen I would get. Under these condtions doesn't the inverse square analogy break down? If your square foot of carrots is surrounded literally by millions of individuals of QAL wouldn't they be swamped out even if I mowed a buffer strip of 100 feet around my carrot grow-out?

[Joseph Lofthouse]

God I hate math. But I love a stiff challenge, so I'll figure out a way to run the math on this problem. It might take me a few months though. I'm having a lot of fun harvesting right now.

I've grown carrot seed for years. I presume that Queen Ann's Lace grows everywhere I've gardened. I never really paid attention. I've never noticed any significant contamination problems.

[Joseph Lofthouse]

Well, it was easier than I had imagined... An extremely complicated and tedious calculation, but trivial to do with a spreadsheet. Based on a pure inverse square analysis, the data looks like this.

Queen Ann's Lace growing ubiquitously in the environment, and not weeded out of the carrot patch. The carrot is more likely than not to be pollinated by Queen Ann's lace.



Queen Ann's Lace weeded only out of the carrot patch. The rest of the garden not weeded. Within the center of the patch there is about a 15% chance that the carrot will be pollinated by QAL. On the edge of the patch the chances are about 50:50.



With Queen Ann's Lace weeded to a distance of 15 feet from the carrot patch there is almost no contribution of pollen from QAL.



We can play other scenarios: Two carrot patches 100 square feet growing 10 feet apart from each other. About a 1% chance that the flowers along the edge of the patch will be pollinated by the other patch.



One QAL weed growing in the middle of the carrot patch results in a contamination rate of about 1%.

[12540dumont]

Joseph, I love you. Calculus hurts before coffee.

How about rogue instead of rouge? You put rouge on a carrot and Canada Mike will have to step in.


Now that you've done all this work with carrots, we were thinking that maybe you should use squash. Make your example pepos and then you have an opportunity to clear up misinformation with Moschatas and Maximas at the same time.

Many new seed savers are scared of the "B" word (biennial). So you can use the carrot model and then do a corn/wind pollinated (chapter 2?) I know you have a book in you.

The following is the best I've read on corn and pollen dispersal. It really shows how topography and wind affects corn. The Swiss as Leo says never do anything by half measures. Some of the charts in here are really good. I forget who posted this to the board, but thanks! I saved a copy and have referred to it many times.

The Swiss were using contamination of a colored corn with non-colored corn to see what would happen with GMO corn

Simulation of transgenic pollen dispersal
by use of different grain colour maize by Michael Bannert
www.agrisite.de/

I found this reading his site: Now if I can just find something on jumping genes!
Breeding of varieties, that can transfer the blue grain colour trait by pollen, is quite difficult. In most blue varieties inheritance of the blue colour is only maternal and will not be transfered by pollen in a visible way. In some blue varieties the colour is also not reliable because of jumping genes or inhibitory allels in yellow varieties.

[Joseph Lofthouse]

Awe shucks... I was just trying to get Michael to notice me....

Great article: The take-home message being that by 30 to 60 feet the amount of stray corn pollen is negligible.

[johno]

Joseph, your mental bag of tricks never ceases to amaze. I have to say, however, that my experience with carrots being pollinated by Queen Anne's lace doesn't seem to follow in this case. Granted, I'm only going by small populations and not controlled circumstances and certainly not impeccable record keeping nor an impeccable memory.

My first attempt at carrot seed-saving was done before I had a grasp of the particulars. I just left a few good carrots in the ground in one spot and let them reproduce the following year. In retrospect, the population size was too small for an outbreeder, but I'm fairly certain there were at least two varieties, so that should have helped the vigor of the next generation. Anyway, the seeds turned out to be contaminated with QAL pollen. Again, no record, but I doubt I would have let QAL go unweeded in the vicinity (Which in this case would have been roughly 50 feet).

After that disaster, I educated myself a bit better and tried again. This time the population was also small by most standards (numbered around 100 plants), but I intentionally used two varieties in hopes that hybrid vigor would negate inbreeding depression. I made sure there was no QAL in the garden (about 80 feet distant), but weeding it out of existence beyond said borders was impossible. It is a predominant weed here. I got nearly a 100% contamination rate anyway. I believe the saturation of QAL pollen here is just too high to try carrot breeding in the open. Whether crossing occurred mostly by wind or insects, I don't know.

This year I have a genetically diverse carrot population (courtesy of Rick Machado) in the ground. I intend to replant them in a block this winter and cover with row cover when they begin flowering, then hand pollinate as outlined in Seed to Seed. I seriously hope to get better results!

What do you think was the problem here?