Do disease-resistant wheat varieties pay a price in yield?
For wheat growers, it’s a truism: Plant varieties that are resistant to prevalent diseases. But what if the wheat plant has to pay a price for resistance, possibly reducing its yield? Is the resistance worth it?
It’s not easy to detect any possible negative effects of resistance genes on yield. Any two wheat varieties you can imagine will very likely differ in their geographical adaptation, yield potential, and reaction to the range of Central Plains diseases. If, for example, you happen to compare a specific stripe rust-resistant variety with a susceptible variety, the resistant one might yield less than the susceptible one in a year without stripe rust. Depending on which ones you’ve picked, it could also yield more. However, that does not mean that resistance causes lower yield. The two varieties differ not only in that stripe-rust gene, but also carry contrasting genes throughout their genomes. Any of those genetic differences could contribute to a difference in yield potential.
Having said that, there is a legitimate question as to whether resistance genes themselves may have a yield effect. The theoretical basis for this comes from plant species in nature. Plant populations are highly variable when it comes to disease reaction, with some plants having more resistance genes and some fewer, and it’s a different situation for each disease. Therefore, ecologists and evolutionists have long reasoned that resistance to pathogens must come at a cost to a plant’s fitness; otherwise, natural selection over many generations would have driven populations toward accumulating every available resistance gene, and plants would be uniformly resistant. That’s not the case. So in crop species, if the goal of artificial selection — the plant breeder's full-time job — is to accumulate resistance genes in crop populations, will genetic yield potential suffer?
Answering that question requires field experiments to estimate the impact of resistance on yield. To be valid, such experiments must compare lines of wheat that are almost identical in their genetic makeup, except for the resistance gene(s) being tested. Either that or the experiment must compare two groups of lines from the same population, one with and one without the gene, so the genetic backgrounds would cancel out. Then these lines must be yield-tested in a replicated, randomized experiment in which the disease in question is totally absent or, better, in which the plots are chemically protected from disease.
Over the years, such controlled experiments in wheat and other crop species, all of them designed to answer the question, “Does the plant pay a price for resistance?” have provided us with a very clear, definitive answer: “Well, it depends.” A survey of these comparative studies, published in the 1990s, found that in exactly 44 out of 88 cases, covering a wide range of species and genes, resistant lines were less productive than susceptible ones in the absence of the relevant disease, insect, or herbicide. In the other 44 cases, there was no difference or, rarely, the resistant line was more productive. The results of studies done since that time have continued to vary in whether they find a yield effect and if so, how big that effect is. It depends on the specific resistance gene(s) involved as well as environmental conditions.
In wheat, as in other crops, some resistance genes reduce yield while others do not. Many of wheat’s genes for resistance have been transferred from related species. In the process, long stretches of DNA extending to either side of the resistance gene come along for the ride. Once in a wheat variety, some of those hitchhiking genes may hurt grain yield even if the resistance gene itself is benign.
This has happened in the past with a chromosome segment from Aegilops umbellulata carrying the Lr9 gene for leaf-rust resistance; it depressed yield by 5 to 14 percent. The Lr47 gene from Ae. speltoides was associated with a 4 percent yield reduction, but in some of the environments and genetic backgrounds examined, there was no effect. There are other interspecific genes such as Fhb1 for fusarium head blight that appear to have brought no yield-reducing hitchhikers with them, and the 1RS rye chromosome arm carrying genes for leaf, stem, and stripe rust resistance has actually had a positive effect on productivity, even when the resistance genes are superfluous or ineffective.
Sometimes the resistance gene itself appears to have a direct impact. The Lr34 gene, which confers adult-plant resistance to leaf rust, originated within common wheat, but its yield-depressing effect is well known. A much-studied powdery mildew gene in barley, Mlo, also reduces yield directly. But whether a yield reduction is caused directly by a resistance gene or indirectly by its bad neighbors, it is crucial to remember that these negative yield impacts have all been measured when there is no disease present. To a wheat grower, such an effect may be less important than the impact of the disease when it does strike.
For example, the experiments that demonstrated Lr34’s effect in spring wheat found that with no leaf rust infection, the line that carried the gene yielded 6 percent less than the one that did not. But with leaf rust infection, the disease hurt the yield of the line with Lr34 by just 15 percent, compared with 43 to 84 percent losses in the line without Lr34. At K-State in the 1990s, my colleagues and I found that a leaf-rust gene transferred from Ae. tauschii provided a 42 percent yield advantage under heavy leaf rust, while it had no yield-depressing effect when leaf rust was barred by fungicide.
You never know at planting time which diseases will be the biggest threats over the coming season, but for diseases common in your area, the possibility that a gene may have a modest negative effect on yield in the absence of disease is probably less important than the risk of taking a much bigger hit to yield, test weight, and quality that comes with sowing a susceptible variety.
A final note: Wheat breeding programs around the Great Plains, the country, and the world expend extraordinary efforts on building genetic resistance into the varieties they release. As a result, could diseases have had another, less obvious effect on wheat’s productivity? That could be the case if resistance breeding has occupied much of the time, effort, and funding that wheat breeders could otherwise have spent on increasing yield.
If that’s true, as some have suggested, maybe breeders should focus their efforts solely on yield and quality, develop susceptible varieties, and let the grower use fungicides to deal with diseases. But the negatives of spraying, such as heavy expense, environmental hazard, and difficulties of timing and effectiveness, easily outweigh the relatively small negative yield effects of some resistance genes — effects that can be detected only in a disease-free environment anyway. And in Kansas, you can never assume you’ll have a disease-free environment.
-- Stan Cox, The Land Institute (and former USDA wheat geneticist based in Manhattan)