You are correct as regards the fruits of dicotyledons, your examples apple and pepper. In cereals (and other graminaceae), the fruit wall is thin and fused to the seed coat, and the grain is essentially comprised of a tiny embryo and the endosperm.

Anastasia is of course right on corn pollination. For more on this, see here . For grain color, here.

Your argument that “stray pollen doesn’t matter and GMO pollen can’t cause economic damage unless the farmer is saving seed for a future crop” has two parts: the word “stray”, and the rest. I’m afraid the anti-GMOs will not bite into the “stray” as a matter of principle (here, in Europe, they have fought hard to impose the detection limit as the trigger for labelling products as containing GMOs — for a robust piece of criticism of the ridiculous situation in Europe, see here). Nor are they biting into the main argument, not even for apples and peppers, since they get sick at the prospect of having to swallow a “GMO seed” or a “GMO pip” (let’s call them like that for convenience purpose).

To the extent that you have an interlocutor who is prepared to listen, your argument remains essentially valid for corn, however, unless the genetic modification is one which has a direct bearing on the grain characteristics.

A fruit like an apple or a pepper begins with an ovary (a part of the flower) containing ovules (eggs, haploid). A pollen grain fertilizes each ovule. Never mind whether one or two sperm are involved – it’s irrelevant. The fertilized ovule becomes a seed, which has genes from the ovule and from the pollen, e.g. diploid. In an apple or a pepper, the seed is a discrete little pip and in theory it can grow into a new plant, at least sometimes. The genetic traits of the seed come from both parents.

But the pips are found inside a container which develops only from the ovary. The ovary expands and becomes the largest part of the fruit. In the case of the apple or the pepper, that’s the part we eat. I’m claiming that the fleshy part of the apple or pepper is entirely genetically like the female parent. I’m claiming that the only genes originating from the pollen are in the seeds. Yes, the seeds consist of an embryo and a surrounding structure and two pollen sperms are involved in the seed, but the thing that matters for the quality of the fruit is not the seeds but the outer fleshy container which we eat.

I’ve used this argument to claim that stray pollen doesn’t matter and GMO pollen can’t cause economic damage unless the farmer is saving seed for a future crop. Your answer tells me that this argument is wrong for corn, that the source of the pollen contributes to the traits of the kernels. That’s disappointing, but inconvenient facts trump even the most satisfying hypotheses. I think my argument is still correct for apples and peppers.

With apples (and pears, peaches, plums, etc.) there’s another issue that the anti-biotech ignoramuses ignore. Just about all fruit trees are reproduced asexually (by grafting) because that’s the only way that a predictable variety can be reproduced. In many cases, the fruit tree has a root from one variety and fruit-bearing branches from a different variety. You can even buy, from garden catalogs, trees which produce different named varieties of fruit on different branches, because several different varieties have been grafted onto the same rootstock. The propagandists make a big deal about how GMOs are unnatural, and how farmers can’t save seeds. But where was their outcry when something was introduced so unnatural as a tree bearing three different kinds of fruit on a rootstock of a fourth variety. Is this something that could happen naturally? If the seed from one of these fruits were planted, would the resulting young tree bear any resemblance to the chimeric parent tree? Of course not.

Conventional breeding can’t produce one chromosome containing A and a, B and b, etc. But genetic engineering can. So it is possible to use genetic engineering to create a plant with two identical chromosomes with AaBbCcDd …, which will then breed true.

Genetic engineering thus far can’t do that either. Gene insertion is still pretty haphazard, genes may land anywhere in the genome – part of the selection process for any GM organism is to pick those which have a single insertion in a spot which is not deleterious – generally you’ll want to figure out exactly where in the genome the gene landed (figuring out the border is, if I recall correctly, one of the regulatory hurdles you need to jump to commercialize – so woe betide any gene that lands in a highly repetitive region (as it becomes nigh on impossible to figure out exactly what the surrounding code is)

As far as I am aware there are methods developed now for more targetted gene insertion (I think Pioneer has at least one patent/patent application in the works using zipper technology to stick genes into precise spots – and I believe that artificial chromosomes are another development which may see the light of day in the next 10-20 years)

You also have to keep in mind that creating two chromosomes with AaBb etc on each chromosome may not actually work at all – you’re essentially doubling the number of genes (Aa…Xx) which could play havoc with gene regulation etc particularly if one of the alleles you’re playing with is a master regulatory element (also recombination would prevent breeding true, after a few generations AaBbCcDdEe could well become AAbbCcDdEe – and with sufficiently homologous alleles crossing over could create some weird AaAaaABbDdEEEEEe type oddness (where Aa crosses over with Aa but the A from Xsome1a gets inserted into the a of Xsome1b to yield AAa)

Conventional breeding CAN produce specific combinations of alleles on a single chromosome because of recombination aka crossing over. Of course, for the most part, the alleles have to be on homologous chromosomes. While it is possible to get chromosomal rearrangements where parts of one chromosome end up tacked onto or replacing part of another chromosome, those events are more rare and can cause more trouble than they are worth. If some sort of genetic engineering was used to create the perfect chromosomes (what you speak of is extremely difficult if not impossible with current technology) those chromosomes would promptly mess themselves up with recombination.

There are other nuances of what can be done with both breeding and biotech, but I am off to the field. Thanks for some early morning food for thought!

Double fertilization is a funny thing that causes unexpected changes in fruits (the fleshy sac around seeds). Pollen actually contains two sperm. One fertilizes the egg and the other fertilizes the fleshy sac which already has two polar nuclei. The genotype of the sperm fertilizing the egg affects the next generation but the genotype of the sperm fertilizing the fleshy sac affects this generation’s fruit (or endosperm as the case may be).

For corn, this means that field corn can mess up sweet corn and sweet corn can mess up popcorn, and so on whenever there are traits in the pollen that are dominant that will show up in the seed. That’s why we can get purple kernels resulting from a fertilization with pollen from a purple corn plant.

Unfortunately, my research focus on corn has made me less knowledgeable about other fruits, but I believe double fertilization still affects them

For apples, this means that each apple on a single tree could have different fleshy sac (fruit!) genotype and phenotype, depending on the pollen source.

I may be wrong, but your experience with the hot peppers seems to be a perfect example of double fertilization. Is the fruit of Capsicum a structure truly created only by the female plant or is it, like apples and corn endosperm, part of the developing ovule? I think it is the latter, though I couldn’t find a reference with a quick search.

Aren’t plants amazing?

We all know that many crops are almost exclusively hybrids and that second generation seeds from hybrids are worthless. Let’s see why?

If one parent line of a hybrid contains a chromosome with alleles ABCD … and the other parent line contains a chromosome with alleles abcd …, the hybrid results all contain A,a,B,b,C,c, D,d, … Presumably this is the genome that we want. Some trait of the plant is superior if it has both genes “A” and “a”, “B” and “b”, etc.

Conventional breeding can’t produce one chromosome containing A and a, B and b, etc. But genetic engineering can. So it is possible to use genetic engineering to create a plant with two identical chromosomes with AaBbCcDd …, which will then breed true.

I don’t see any company investing to create such a plant, because it wouldn’t be able to sell its seeds year after year. But that’s an economic issue, not a technology issue. A non-profit institute could create effectively pure reproducing seeds with the desirable traits of any hybrid.

If pollen from field corn fertilizes silks from sweet corn, the resulting kernels will develop a tiny embryo inside a very much larger packet of starchy endosperm. I believe that the latter’s composition is entirely dependent on the genetics of the parent plant. I think your sweet corn will taste just fine. Unless you save the kernels as seed for next year’s planting, I don’t think you’ve been harmed by cross-pollination.

I’m not sure about corn, but I can give a better assurance with some fruits. For example, I once planted sweet peppers in my backyard garden and when I ate one of the peppers it was hot, e.g. it contained capsaicin. My initial assumption was that it had been pollinated from someone else’s nearby hot peppers. But I was wrong. All the peppers from that one plant were spicy, and nobody nearby was growing hot peppers. What happened was that the seed from which the sweet pepper plant grew was the accidental result of a cross-pollination. The pepper plant fruit consists of a fleshy fruit entirely created by the parent plant, and seeds which contain some of the genes from the pollen. But of course, we don’t eat the seeds.

Similarly, if you have an apple tree, the apples are all genetically identical to the parent except for the seeds, which nobody eats. Named apple varieties are all clones, propagated by grafts.

I’ll add yet another nuance to Anastasia’s answer.

Many crop plants are not allowed to flower in production fields, so there can’t be any “cross contamination with organics”. Seed production for those crops is furthermore undertaken by specialist farmers (and I would also surmise, organic seed savers – if they were not, they would be rather incompetent) under very strict conditions which prevent cross-pollination not only between GMO and conventional (or organic), but also between the variety that is the subject of production and any other variety. So there is no “cross contamination” either. Examples are the beets and all root and leaf vegetables, e.g. the cabbages.

The original post has referred to the (seedless) banana, and you call for the exploration of alternatives to GMOs. Well, in the case of bananas and some other crops such as sugar cane, the alternatives are extremely complicated . The position you expose means that adepts of organic food, a minority in affluent countries, are in effect depriving millions of people who depend on bananas for their staple in developing countries of the benefit of genetic progress, even where that progress is a matter of life or death as in the case of the black sigatoka disease . They are also condemning workers in banana plantations, their families and people living nearby plantations to massive doses of pesticides. Is this ethical?

For one thing, the rate of outcrossing varies widely by species. Some plant species are natural selfers, others reproduce clonally. In these cases, there isn’t any cross pollination, so no cross contamination. Even if cross pollination was a good enough reason to ban GMOs, there’s no reason to ban genetic engineering in crops that don’t cross pollinate.

Another issue is contamination of one type of crop with another. Let’s say we’re neighbors. I grow sweet corn to sell at the farmer’s market and you grow feed corn to feed to your organic heirloom chickens. Both of our corn fields are certified organic and do not have any genetically engineered seed. Your pollen blows over to my field in the wind and pollinates my corn. Since field corn does not contain any of the mutations that makes corn sweet, any of my ears that are pollinated by your corn don’t taste good. Word goes around that my sweet corn isn’t sweet and my business is ruined. Is it your fault for contaminating my sweet corn? Maybe. If a farmer with GMOs should be blamed for “contaminating” a non-transgenic field, you should certainly be blamed for “contaminating” my sweet corn.

Finally, there has been quite a bit of research that shows mutagenesis, tissue culture, and wide crosses (crosses between a crop plant and a relative of a different species) cause more unintended changes in the genome than genetic engineering. None of these have to be tested for safety, and all of them are allowed in organic farming without question. Doesn’t it seem strange that the safest method is considered the least safe?