Xavier Remongin / agriculture.gouv.fr

Today, agriculture finds itself facing a major challenge: Ensuring adequate production for a growing population all while protecting the environment. To limit the expansion of agricultural land, a possible path forward is increasing crop density (plants planted per square meter), which has the consequence of increasing competition of plants among themselves for resources (water, light, minerals, etc.), to the detriment of the yield. Researchers from INRAE, the French Institute of Agricultural and Environmental Research (which is the result, since January 1, 2020, of a fusion of INRA with another research institute in agriculture and food), and others, developed a selection model based on cooperation between plants planted in high density, which allows for an increase in yields while using existing agricultural lands.

Since the origin of agriculture, humans have practiced artificial selection to increase the quality and quantity of crops. Starting in the 20th century, agronomists have noted that resource competition between plants led to a decrease in crop yields. Plants that are more competitive are those that are able to develop specific necessary characteristics to capture and use these resources. Competition for sunlight is a good example: More competitive plants invest their resources in developing their stalks and leaves to capture more sunlight. This has two negative effects on crops: Competitive plants produce less seeds and take resources from their neighbors. From generation to generation, this selection process led to a decrease in the crop yields.

To improve plants, it has been suggested, starting in the 1960s, that selecting less competitive varieties could maximize crop production. However, the characteristics that make a plant less competitive remain poorly understood, and depend on the nature of resources that are limited. In animal species, evolutionary biology is interested in social behavior, and particularly in the existence of altruistic behavior. If they decrease the survival or reproductive rates of individuals that express them, they are beneficial for the group of individuals in interaction (decrease in aggression, for example). Kin selection theory thus offers a framework for improving crop yields. However, until now, no method had been formalized for cultivated species.

In this study, the researchers built a theoretical model of kin selection to formalize competitive relations between plants in response to different selection models. The researchers based their study on light competition, with the competitive trait of height, in a grain plant. They based their selection model of the production of seeds in a group of plants.

Through testing different selection strategies, the model allowed to highlight the most important parameters for encouraging cooperation between plants:

  • High density seeding (number of seeds planted per square meter)
  • High levels of relatedness between the plants
  • A strong contribution of the most productive groups to the next generation

By playing with these three parameters, only a handful of generations were needed to obtain groups of plants having cooperative characteristics (small size for light) and to increase the yield of the plot. While the researched based their study on competition for sunlight, the model is easily transposable to all types of resources.

The model proposed by the researchers allows to revisit the evolution of selection practices for sunlight in the theory of kin selection and to propose selection models that are more efficient for encouraging cooperation. The most important parameters for encouraging cooperation can easily be manipulated by a farmer. Additionally, the selection model described in this study does not require the use of plant phenotyping, a technique that can be very costly.