The Decision Support Tool (DST) is a free, web-based software developed through the SoildiverAgro project to help end-users, including farmers and other stakeholders, optimize their wheat cropping systems and management practices.
The DST uses machine learning models developed from experimental data collected during the SoildiverAgro project to predict agricultural soil properties. It takes into account environmental, physicochemical, and land-use parameters, as well as biodiversity indices. The tool guides users through a step-by-step process, asking for key information such as pedoclimatic conditions, crop location, agricultural challenges, and current cropping systems. Based on this input, the DST provides tailored recommendations to enhance productivity, profitability, and sustainability.
Your tool to make more informed decisions and optimize your farming practices.
Download template
To get started, download the excel template provided. In this template, each row represents a single soil sample. Simply fill in the required data for each sample as specified in the template (there are three examples given, in case you need them).
Once you have filled it in, please uploaded it using the 'Browse' button.
Check results
Next, navigate to the "Check Results" section using the menu on the left. Here, you can verify that your data has been uploaded correctly.
The tool displays your data in tables and graphs, organized into three main categories: metadata, physicochemical variables and climate-related variables.
Obtain predictions
In the "Predictions" section, the DST shows estimated values for yield, bacterial biodiversity, fungal biodiversity, and nematode biodiversity based on your soil samples. These results appear in an interactive graph where you can hover over each sample to view the exact values.
By default, the predictions are compared to all samples from the SoildiverAgro project. Selecting the “By category?” option, the results will be grouped by metadata, allowing a more detailed comparation.
The tool also provides tables with numerical values of prediction: raw prediction values, percentiles values and percentiles per region values.
