But there are other more violent ways to change a genome. Early in the twentieth century it was discovered that x-rays or certain chemicals could make genes change at random. Almost all of these changes were harmful, but a few were not. Breeders were glad to get a few extra mutated kinds of genes to subject to the old fashioned crossbreeding. The United Nations Food and Agricultural Organization has identified over two thousand different varieties of crop which were developed using chemicals or radiation. Are they safe? Here we don't have centuries of experience to guide us. Each new mutation has its own story, not to mention all the ways it can be in combinations with other genes. Of course, breeders understand this and they do some elementary testing of their newly developed crops. They are motivated to be sure that the crops will be accepted in the market place. But the testing of mutated crops is not mandatory, and there are no standards.
Another violent way to change a genome is to double up its chromosomes. This can be accomplished by treating the fertilized ovum with a chemical called colchicine. Even one extra chromosome can make a big difference in the phenotype. The human genetic disease called Down's syndrome (formerly called mongolism) is the result of just one extra chromosome out of the normal human complement of 46. But again, breeders are glad to get the extra variation to drive their selections. Dozens of crops and many decorative plants were created this way.
The rules for testing of doubled chromosome varieties are the same as for mutated varieties -- there are none.
It has to be understood that when plants are developed by these violent techniques, we have no idea what genes are involved, what proteins they make, what chemical transformations these proteins facilitate. Suppose we wanted to recommend some tests for a particular mutated crop. Where would we start? When would be think we had tested enough, and why?
Contrast this with genetic engineering. It makes a change in, usually, one gene. That gene is understood in complete detail. The genetic engineer knows the ACGT sequence of his gene, and what turns it on or off. The engineer knows where the gene is placed in the chromosome. He or she knows what protein it synthesizes and knows that only that one extra protein is going to be made in the transformed crop. He knows what chemical reactions the protein controls.
This doesn't mean that the genetic engineer knows everything. Living things are just too complex to make perfectly reliable predictions. But the genetic engineer can easily plan a rational set of tests to assess the safety and success of the transformation. The safety of the new protein by itself can be assessed even before the transformation. Its effect on the rest of the life process of of the plant is assessed by comparing the chemical composition of the transformed plant with its precursor. It is just simply easier to test for the effects of a change when you know what you changed.
Why are genetic engineered crops tested so much more thoroughly than mutant crops? Because it's easier.