arrow right cross menu search arrow-left arrow-right menu open menu close plus minus

How Soil Sampling and Modeling Work Together

Soil health is a critical concern in regenerative agriculture.  Knowledge of soil health provides a basis for sound land management decisions and a starting point for regenerative practices. There are two common ways to understand soil health and analyze the nutrient management in fields – soil sampling and modeling. 

The Importance of Soil in Agriculture 

Some methods of agriculture, such as hydroponics, do not use soil.  For the vast majority of growers, however, soil is a primary concern. Agricultural soils vary widely in texture, depth, pH, and nutrient composition.  Understanding soil composition and soil health enables growers to make informed decisions to optimize yields, protect the environment, limit runoff, and improve the nutritional balance of an ecosystem. Knowledge of soil composition can provide an indication of potential nutrient deficiencies, pH imbalance or excess soluble salts.

Soil health can be monitored through soil sampling.  As we will see below, taking soil samples over time can provide the grower with an important mental map of how soil health is changing as the land is managed.  However, such soil samples are not only labor-intensive: they are also expensive and sometimes error-prone.   In recent years, some growers have turned to computer models to provide more insight into changes in soil in managed fields.  In this post, we will discuss the advantages and disadvantages of these two monitoring tools, and consider how combining these approaches may lead to the best overall results. 

What are the Methods of Soil Sampling? 

According to Michigan State University, there are two primary ways to conduct soil sampling. Dennis Pennington from Michigan State University explains, “Farmers today are sampling differently, attempting to better quantify the soil test levels and then prescribe fertilizer programs that maximize yield potential while preserving the environment.”

  • “Grid soil sampling is usually conducted on 2.5-acre grids. By collecting one sample for every 2.5 acres, farmers are better able to identify the variability in a field. But is 2.5-acre grids small enough? Some farmers will pull samples from 1-acre grids to try to identify problem areas of fields. This is costly upfront, but there may be cases where smaller grids may be a good diagnostic tool.”
  • “Management zone sampling is conducted in a different way. Rather than simply splitting a field up into exact 2.5-acre quadrants and pulling one sample from them, the farmer will create zones where soil types, topography, yield potential, organic matter, CEC and previous history shows similar responses to management. Each farm uses their own approach to creating management zones, but the end goal is to collect soil samples in a more informed way than sampling using GPS to put out points every 2.5 acres.”

Both of these methods have pros and cons. Farmers need to decide which method works best for their farm based on management decisions. Many farmers apply nutrients to their fields based on best performance zones which should be considered when soil sampling. 

How Modeling can Inform Soil Health

Another approach to monitoring soil changes over time is to use a simulation model.   Simulation models start with a known soil state, and then simulate the effects of agricultural practices on that starting state.  The starting state can be determined from an initial set of soil samples on a field, or obtained from a public soils database, such as SSURGO.

In some ways, a simulation model will be less accurate than a soil sample.  A model can take many factors into account, but it can never account for every factor that affects soil health.  But in other ways, a simulation model can provide additional insights that are difficult or impossible to achieve through soil sampling alone.  For example, a simulation model can easily be run at many different points throughout the field, to achieve whatever level of precision the grower may need.  In addition, a simulation model can be run under many different hypothetical scenarios.  This allows the grower to not only monitor the effects of past practices, but also to project forward the effects of new practices not yet adopted.