Indeed, in natural ecosystems, ecologists have long shown that increasing plant diversity usually results in increasing the productivity.
This has been used in agriculture in developing countries and was formerly used in European agriculture as well, prior to mechanization and intensification in the 1950s and 1960s. I am talking about intercropping systems, mixing different plant species in a given field, such as legume-cereal systems. These have proven very efficient at achieving stable yields and high protein content in cereals such as durum wheat. Systems that are even more diverse are agroforestry systems. However, such systems are difficult to manage, as they require special harvesting machines or pose some constraint related to their spatial heterogeneity.
An alternative approach is to increase the diversity within the crop one is using. Most of our arable lands are composed of single-species systems, in which every plant individual is almost identical as they share identical genetic characteristics: they are composed of a single genotype/variety.
Why not use mixtures of genotype/ varieties instead?
It has long been shown in species such as rice that such mixtures are more resistant against the spread of disease or pests. Although it is less documented, they are likely to be more tolerant against abiotic stresses, such as drought, or nutrient deficiencies, which is the reason why we aim to test these mixtures in the context of our SolACE research project.
The underlying concept is quite simple: Whatever genetic progress one can achieve, it is unlikely that it will be feasible to obtain all the desired traits in a given genotype. Instead, combining those traits by using adequate mixtures of genotypes is easier to achieve. However, the plasticity of some traits makes their expression more difficult to predict. For example, root depth or plant height depends on competition for resources by neighbouring plants.
The big challenge is that combining different genotypes or varieties offers so many possibilities of mixtures (depending on the number of genotypes to be mixed) that it is difficult to assess and requires extensive testing. You should design your own mix! Have a try. In the worst case scenario, you will get the same yield as you normally achieve with a single genotype!
What are you experiences? Questions? Let us know in the discussion forum below!