Service crops are allies to intensify rotations and improve ecosystem services. But how to include them without harming the yield of income crops?

With the aim of providing a decision tool, researchers from INTA Laboulaye, the National University of Villa María and producers from the Argentine Association of Direct Sowing Producers (Aapresid, by its acronym in Spanish) of the Justiniano Posse Farm presented a model that allows simulating the effects of different crop management variables of services in the yield of late corn.

The model developed is an algorithm capable of simulating the yield of corn based on changes in the sowing date and drying time of the service crop, as well as the probability of rain during the period between the drying of the service crop and the sowing of the crop crop.

As detailed by the technicians, the model is based on machine learning, that is, on algorithms trained to detect patterns in large data sets, which allows us to recreate various productive situations and make more informed decisions.

For its development, data obtained from field trials of the Aapresid-BASF Services Crop Network were incorporated, which cover a wide variety of crops, climatic and edaphic conditions and agronomic practices in numerous localities in the Pampas region.

After applying the model to various environmental scenarios and the usual management of producers, they revealed that yield is mainly influenced by factors such as: drying date of the service crop, initial useful water for planting the service crop, rainfall during the service crop cycle, duration and rainfall between the drying period of the service crop and the sowing of the crop crop, also called “fallow,” among others.

For example, the model showed that delaying the drying date can result in a significant decrease in corn yield, especially in dry years or without influence of the water table. On the contrary, advancing the drying date to September has a positive impact, especially in humid years or with the presence of napa.

On the other hand, they saw that the variable water accumulated at the beginning of the service crop significantly influences the yield of corn, especially in conditions of medium to high recharge.

The study also notes that different utility crops can have varying effects on corn yield, ranging from 80% losses to 65% gains. For example, rye pure or mixed with other species can cause significant drops in yield compared to pure vetch or mixed vetch (Vicia villosa+Persian clover+rye). For this combination, a lower negative impact (< 43% of cases) and water cost was recorded compared to a fallow situation.

The researchers state that "the predictive model developed allows us to evaluate feasible scenarios that could occur in the Pampas region and quantify their impact on corn yield, becoming a valuable tool to facilitate decision-making for producers and technicians."